Embed Size (px)
Citation preview
DESCRIBING THE EFFECT OF MOTOR ABILITY ON
VISUAL-MOTOR SKILL ACQUISITION AND
TASK PERFORMANCE IN CHILDREN WITH
DEVELOPMENTAL COORDINATION DISORDER
By
Noémi Cantin
A thesis submitted in conformity with the requirements
for the degree of Doctor in Philosophy
Graduate Department of Rehabilitation Science
University of Toronto
© Copyright by Noémi Cantin (2012)
ii
DESCRIBING THE EFFECT OF MOTOR ABILITY ON VISUAL-MOTOR SKILL
ACQUISITION AND TASK PERFORMANCE IN CHILDREN WITH
DEVELOPMENTAL COORDINATION DISORDER
Noémi Cantin, Doctor in Philosophy (2012)
Graduate Department of Rehabilitation Science, University of Toronto
Abstract
Background: For children with developmental coordination disorder (DCD), the
acquisition and performance of everyday visual-motor activities such as buttoning,
shoe tying, cutting with scissors or writing, presents a major challenge. Regardless
of the activity considered, children with DCD are typically slower and less accurate
than their peers. Given the well-acknowledged difficulties of children with DCD, it is
surprising to find very few research studies systematically exploring visual-motor
skill acquisition and performance in children with DCD. Objective: The overall
objective of this study was to systematically describe visual-motor skill acquisition
and task performance in children with DCD.
Methods: Twenty-four children (8 years 11 months to 12 years 11 months) were
recruited for this study; 12 children with DCD, 12 children developing typically with
regards to their motor skills. A computer-based aiming task completed with three
different cursor controls of increasing levels of difficulty (mouse, joystick, novel
controller) was designed for this study. Mixed-effect modelling and visual graph
analyses were performed to describe the influence of motor ability and task difficulty
on visual-motor skill acquisition and task performance.
iii
Results: Motor ability modulated the impact of task difficulty on visual-motor skill
acquisition and task performance. Children with DCD were as fast and as accurate
as their peers in their initial performance of the simple, well-learned task (mouse).
However, they were slower and less accurate when performing the complex and
novel visual-motor task. Over repeated trials, the visual-motor task performance of
children with DCD improved on all tasks, even for the simple. With regard to the
complex, novel task, once children with DCD understood the features of the task,
their performance also improved and approached that of their peers.
Conclusion: While children with DCD can generally be characterized as less
accurate and slower than their peers, this characterization needs to be specified and
qualified; it is probably best not applied to a well-learned task.
iv
ACKNOWLEDGMENTS
Where to start! So many years have gone by since I embarked on this journey! I
have learned that during the course of a PhD, “life happens”, bringing with it its ups
and downs. Many people have come and gone during this PhD, each helping in their
different ways. They have enabled me to keep moving forward and to trust that one
day, I would be sitting at my computer writing this last page, these last few words.
I would first like to acknowledge the tremendous support I have received from my
supervisor, Helene Polatajko. Her generosity towards her graduate students is
remarkable. Knowing that, rain or shine, morning or night, she was always there for
me kept me motivated and kept me writing even in the wee hours of the night. I also
wish to acknowledge the support received from the members of my program
advisory committee: Heather Carnahan, Tom Chau, Michelle Keightley, and Jennifer
Ryan. I would like to thank Scott Young, then fellow PhD student in Biomedical
Engineering at the University of Toronto. Our many discussions inspired the design
of this study. I would like to recognize the work of Eric Wan who developed the
computer program that was used in this thesis. I would also like to acknowledge the
contribution of Malcolm Binns, statistician scientist at the Rotman Research Institute
in Toronto. His guidance while I learned to perform mixed modelling was invaluable.
I wish to acknowledge the numerous students that have contributed to this project
through the years: Freda Goh, Alison Firestone, Megan Henze, Melodie Lumague,
Jennifer Crouchman and He (Cherry) Ma. I also wish to sincerely thank the children
and families that participated in my study, and everyone who helped in the
recruitment process.
Indeed, “life happened” during the course of this PhD. Striving to keep my family a
priority throughout the process has helped me stay grounded and true to myself.
Thank you for your understanding when late nights were making it hard to play hide-
and-seek in the morning; merci grand-maman Diane for cutting and pasting for
hours with Violette and for playing ‘rescue, rescue’ with Philippe; and a heartfelt
thank you to my husband David for always being there for me and for believing that I
would one day finish!
v
TABLE OF CONTENTS
Abstract ....................................................................................................................ii
Acknowledgments ...................................................................................................iv
Table of Contents .................................................................................................... v
List of Tables............................................................................................................ix
List of Figures .......................................................................................................... x
List of Appendices .................................................................................................. xii
1 Introduction ...................................................................................................... 1
2 Background and Rationale ............................................................................... 6
2.1 Characterizing the Motor Abilities of Children with DCD ........................... 7
2.1.1 Motor Difficulties .......................................................................... 7
2.1.2 Visual-Motor Difficulties ............................................................... 9
2.2 Describing Visual-Motor Skill Acquisition and Task Performance in
Children with DCD ........................................................................................ 11
2.2.1 Writing Tasks............................................................................. 12
2.2.2 Trail-Drawing Tasks .................................................................. 15
2.2.3 Fine Motor Aiming Tasks ........................................................... 16
2.2.4 Summary ................................................................................... 21
2.3 The Nature of Visual-Motor Skill Acquisition and Task Performance ...... 24
2.3.1 Task-Specific Features .............................................................. 25
2.3.2 Motor Control Mechanisms........................................................ 31
2.4 Investigating Visual-Motor Task Acquisition and Visual-Motor Task
Performance in Children with DCD: Rationale and Research Objectives ..... 35
vi
3 Methods ......................................................................................................... 38
3.1 Participants.............................................................................................. 38
3.2 Descriptive Measures .............................................................................. 42
3.3 Experimental Measure ............................................................................ 43
3.3.1 Apparatus .................................................................................. 45
3.4 Procedure................................................................................................ 46
3.4.1 Experimental Visual-Motor Task................................................ 47
3.5 Variables & Data Handling ...................................................................... 48
3.6 Data Analyses ......................................................................................... 49
3.6.1 Description of Participants ......................................................... 51
3.6.2 Visual-Motor Task Performance: Modulating Variables ............. 51
3.6.3 Visual-Motor Skill Acquisition: Modulating Variables ................. 53
3.6.4 Visual-Motor Skill Acquisition: Patterns of Change.................... 56
4 Results ........................................................................................................... 59
4.1 Description of Participants....................................................................... 60
4.2 Visual-Motor Task Performance: Modulating Variables .......................... 61
4.2.1 The Effect of Task Difficulty on Initial Visual-Motor Task
Performance in TD Children ............................................................... 61
4.2.2 The Effect of Motor Ability on Initial Visual-Motor Task
Performance When Task Difficulty Is Also Considered ...................... 63
4.3 Visual-Motor Skill Acquisition: Modulating Variables ............................... 66
4.3.1 The Effect of Repeated Trials on Visual-Motor Task Performance
in TD Children When Task Difficulty Is Also Considered .................... 66
vii
4.3.2 The Effect of Motor Ability on Visual-Motor Task Performance
When Repeated Trials and Task Difficulty Are Also Considered ........ 68
4.4 Visual-Motor Skill Acquisition: Patterns of Change ................................. 69
4.4.1 Rates of Visual-Motor Skill Acquisition ...................................... 70
4.4.2 Patterns of Change in Visual-Motor Task Performance over
Repeated Trials .................................................................................. 74
5 Discussion ..................................................................................................... 83
5.1 Visual-Motor Task Performance: Modulating Variables ........................... 83
5.1.1 Describing the Effect of Task Difficulty on Initial Visual-Motor
Task Performance in TD Children ...................................................... 83
5.1.2 Describing the Effect of Motor Ability on Initial Visual-Motor
Performance When Task Difficulty Is Considered .............................. 85
5.2 Visual-Motor Skill Acquisition: Modulating Variables ............................... 87
5.2.1 Describing the Effect of Repeated Trials on Visual-Motor Task
Performance in TD Children ............................................................... 88
5.2.2 Describing the Effect of Motor Ability on Visual-Motor Task
Performance over Repeated Trials..................................................... 88
5.3 Visual-Motor Skill Acquisition: Patterns of Change ................................. 91
5.3.1 Patterns of Change in TD Children ............................................ 91
5.3.2 The Effect of Motor Ability on Patterns of Change..................... 92
5.4 Characterizing Visual-Motor Skill Acquisition and Task Performance in
Children with DCD ......................................................................................... 95
5.4.1 Children with DCD Are Less Accurate Than Their Peers .......... 95
5.4.2 Children with DCD Are Slower, Faster, “Same as” Their Peers 97
5.5 Limitations ............................................................................................... 98
viii
5.5.1 Exploratory Nature of This Study ............................................... 98
5.5.2 Measuring Motor Ability ............................................................. 99
5.5.3 Technical Issues and Missing Data ......................................... 100
5.6 Clinical Implications and Future Directions............................................ 100
5.6.1 A Consideration for Identifying Children with DCD .................. 100
5.6.2 A Consideration for Intervening with Children with DCD ......... 101
5.6.3 A Consideration for Future Research to Characterize Visual-
Motor Performance in DCD .............................................................. 101
5.7 Conclusions........................................................................................... 102
References ............................................................................................................ 104
Appendices ............................................................................................................ 114
ix
LIST OF TABLES
Table 3-1 Criteria used to describe participants as DCD or TD ............................... 39
Table 3-2 Independent variables investigated for their effect on visual-motor task
performance ............................................................................................................ 49
Table 4-1 Description of Participants (N = 24) ......................................................... 60
Table 4-2 Investigating the effect of task difficulty on initial visual-motor task
performance TD (n = 12) ......................................................................................... 61
Table 4-3 Investigating the effect of motor ability on initial visual-motor performance
when task difficulty is considered in all children (N = 24). ........................................ 63
Table 4-4 Investigating the effect of task difficulty on visual-motor task performance
over repeated trials in TD Children (n = 12). ............................................................ 67
Table 4-5 Investigating the effect of motor ability on visual-motor performance over
repeated trials when task difficulty is taken into account (N = 24). .......................... 69
x
LIST OF FIGURES
Figure 2-1 Relationship between task difficulty and expected performance as a
function of the individual's skill level. ....................................................................... 26
Figure 3-1 Computer game and novel cursor controller........................................... 43
Figure 4-1 Mean initial movement times and distances travelled for each level of
difficulty to illustrate the modulating effect of task difficulty on initial visual-motor task
performance for TD (n=12). ..................................................................................... 62
Figure 4-2 Illustrating the modulating effect of motor ability on initial visual-motor
performance for the simple, intermediate and complex tasks in all children (N = 24)
................................................................................................................................. 64
Figure 4-3 Illustrating the effect of motor ability on initial visual-motor task
performance in all children (N = 24) when task difficulty is considered. ................... 65
Figure 4-4 Illustrating the effect of motor ability on initial visual-motor task
performance in TD children (n = 12) when task difficulty is considered. .................. 67
Figure 4-5 Illustrating rates of skill acquisition and changes in visual-motor task
performance over repeated trials during the complex task for TD children (n = 12). 70
Figure 4-6 Illustrating rates of skill acquisition and changes in visual-motor task
performance over repeated trials during the simple task for all children (N = 24). ... 71
Figure 4-7 Illustrating changes in movement time over repeated trials .................... 72
Figure 4-8 Illustrating changes in visual-motor task performance over repeated trials
during the complex task for all children and comparing predicted visual-motor task
performance between TD children (n = 12) and all children (N = 24). ..................... 73
Figure 4-9 Illustrating patterns of change in visual-motor performance during
complex task TD children (n = 12) ........................................................................... 75
xi
Figure 4-10 Illustrating patterns of change in visual-motor task performance during
simple task for all children (N = 24) ......................................................................... 76
Figure 4-11 Illustrating patterns of change in visual-motor task performance during
simple task for all children (N = 24) and TD children (n = 12) to demonstrate the
impact of adding children with DCD. ........................................................................ 77
Figure 4-12 Illustrating patterns of change in movement time during intermediate
task for all children (N = 24). .................................................................................... 78
Figure 4-13 Illustrating patterns of change in movement time during intermediate
task for all children (N = 24) and TD children (n = 12) to demonstrate the impact of
adding children with DCD. ....................................................................................... 79
Figure 4-14 Illustrating patterns of change in visual-motor performance during
intermediate task for all children (N = 24). ............................................................... 80
Figure 4-15 Illustrating patterns of change in visual-motor performance during
complex task for all children (N = 24) and TD children (n = 12) to demonstrate the
impact of adding children with DCD. ........................................................................ 81
xii
LIST OF APPENDICES
Appendix A Distribution of M-ABC Scores (N = 24) ............................................... 115
Appendix B Missing Data and Outliers per Trial per Controller .............................. 116
Appendix C Description of Participants: Distributions and Group Differences ....... 119
Appendix D Parameters Estimates and Standard Errors ....................................... 120
Appendix E Visual Pattern Analysis of Changes in Performance of Children with
DCD (n = 12) during Complex Task ....................................................................... 121
Appendix F Methods for Calculating Special Causes ............................................ 122
Appendix G Regression Curves TD Children on Tasks for which Trial was not a
Modulating Variable ............................................................................................... 123
1
CHAPTER 1
INTRODUCTION
2
1 INTRODUCTION
Young children are expected to learn the skills to perform a variety of visual-motor
activities, such as buttoning, tying their shoes, drawing, cutting with scissors, or
writing. The term visual-motor activity typically refers to an activity for which the
involvement of the visual system is required to guide the movement (Schmidt &
Wrisberg, 2008). For most children, the acquisition of the visual-motor skills
necessary to perform such everyday activities occurs quickly and with relatively little
effort, following more or less predictable developmental patterns (Siegler, 2005;
Wickstrom, 1983). However, for some children, the acquisition of skills and the later
performance of such everyday activities present a major challenge.
Aaron’s Mom: When I was a little girl I was an excellent printer, was the best
in the class and that was just me. And so I thought Aaron doesn’t have to be
like me, he can be one of those guys with messy printing and that was my
initial reaction, but then I saw how it was impacting how he felt about himself.
(Mandich, Polatajko, & Rodger, 2003, p. 590)
While children with a variety of diagnoses report difficulties with the acquisition and
performance of visual-motor skills, difficulties with handwriting is the primary reason
for the referral of children with Developmental Coordination Disorder (DCD) to
occupational therapy services (Miller, Missiuna, Macnab, Malloy-Miller, & Polatajko,
2001; Missiuna, Moll, King, Stewart, & MacDonald, 2008). DCD is a
neurodevelopmental disorder characterized by motor abilities that fall below what
would be expected given a child’s age. Children with DCD experience motor
difficulties typically manifested by slower, less accurate, and more variable motor
performance (Elders et al., 2010; Estill, Ingvaldsen, & Whiting, 2002; Johnston,
Burns, Brauer, & Richardson, 2002; Smits-Engelsman, Bloem-van der Wel, &
Duysens, 2006).
The coordination difficulties of children with DCD affect their performance of many
motor-based childhood activities at home, at school, and in their community
3
(American Psychiatric Association [APA], 2000; Magalhães, Cardoso, & Missiuna,
2011; Miller et al., 2001). In addition to poor handwriting, performance problems
frequently identified in the literature include play activities such as bicycle riding,
catching, throwing balls or jumping rope, classroom activities such as using scissors
or drawing, and self-care activities such as buttoning, doing up zippers or using
utensils (Astill & Utley, 2008; Kennedy-Behr, Rodger, & Mickan, 2011; Magalhães et
al., 2011; Miller et al., 2001; Rodger et al., 2003; Stephenson & Chesson, 2008;
Summers, 2008; Wang, Tseng, Wilson, & Hu, 2009).
In the past two decades, numerous researchers have studied children with DCD in
an attempt to characterize their motor difficulties and to develop a better
understanding of the nature and mechanisms underlying these difficulties. More than
15 years ago, Wilson and McKenzie (1998) conducted a comprehensive meta-
analysis to identify the processes most strongly associated with the poor motor
abilities of children with DCD. They collected data from 50 studies that compared
control children to children meeting the diagnostic criteria for DCD on perceptual and
motor measures of information processing. After the analysis of effect sizes, Wilson
and McKenzie concluded that while the DCD group’s performance was inferior to
that of the control group on all measures, the largest of these differences occurred in
the complex visual-spatial category that involved a motor component. Considering
the activities reported to be affected in children with DCD, this finding is not
surprising. The performance problems of children with DCD in activities requiring the
interaction of the visual system and motor control processes is well documented
(e.g., Astill & Utley, 2008; Kennedy-Behr et al., 2011; Magalhães et al., 2011).
The ability to acquire visual-motor skills and perform visual-motor activities is often
considered a defining feature of human life, the central component of many daily
activities (Crawford, Medendorp, & Marotta, 2004). School-aged children spend
more than 50% of their time in school performing some type of visual-motor activity,
handwriting being the predominant one (McHale & Cermak, 1992). Numerous
descriptive studies attempting to capture the effect of the poor motor abilities
associated with DCD in children’s daily lives have identified that performance
4
problems related to visual-motor activities are characteristic of this group of children
(Magalhães et al., 2011).
To date, most researchers have focused on describing the end-product of visual-
motor tasks, most of them requiring writing instruments (e.g., Chang & Yu, 2010; Di
Brina, Niels, Overvelde, Levi, & Hulstijn, 2008; Kagerer, Contreras-Vidal, Bo, &
Clark, 2006; Rosenblum & Livneh-Zirinski, 2008; Smits-Engelsman et al., 2006;
Zwicker, Missiuna, Harris, & Boyd, 2011). Such studies have been useful in
describing visual-motor performance issues in children with DCD. However, they
have not generated the detailed characterization of these visual-motor skills that is
needed to develop a better understanding of the difficulties experienced by children
with DCD. Such an understanding would be invaluable to researchers interested in
exploring the nature of the motor difficulties of children with DCD. Indeed, very few
studies specifically exploring visual-motor skills acquisition were found (Chang & Yu,
2010; Missiuna, 1994; Zwicker et al., 2011). Given the continuum on which task
performance and skill acquisition lie, the paucity of research studies on this topic is
quite disconcerting. It is clear that if we are to develop effective intervention
approaches that will help children with DCD overcome their difficulties and acquire
the skills necessary to perform everyday activities, gaining a thorough understanding
of their skill acquisition is essential.
In light of the documented impact of DCD on children’s performance of visual-motor
activities, the first objective of this thesis was to describe the influence, or
modulating effect, of children’s motor abilities on their performance of a visual-motor
task with varying levels of difficulty. Considering the small number of studies
exploring visual-motor skill acquisition in children with DCD, the second objective
was to describe the impact of children’s motor abilities on their acquisition of a
visual-motor skill. Here it is important to make a distinction between skill acquisition
and motor learning. Skill acquisition refers to changes in performance occurring
during a practice session. Motor learning refers to the permanent changes in
performance that can be measured during a retention session (Schmidt & Lee,
2011). This thesis focused on skill acquisition. This was done by first exploring
5
modulating variables influencing skill acquisition, and then by describing the impact
of motor ability on patterns of changes in task performance occurring over repeated
trials.
This thesis was written following a traditional thesis style. Accordingly, in addition to
this chapter, there are four others: chapter 2, “Background and Rationale”; chapter
3, “Methods”; chapter 4, “Results”; and chapter 5, “Discussion and Conclusion.”
6
CHAPTER 2
BACKGROUND AND RATIONALE
7
2 BACKGROUND AND RATIONALE
This chapter is organized into four sections that will present background information
relevant to the study. Section 2.1 briefly describes the motor abilities and motor-
based performance difficulties of children with developmental coordination disorder
(DCD), highlighting the possibility of an underlying visual-motor deficit. Section 2.2
will follow with an overview of the current state of the evidence on visual-motor skill
acquisition and visual-motor task performance in children with DCD, clearly
demonstrating the paucity of research on this topic. This overview will also reveal
the methodological limitations associated with comparative study designs when
studying children with DCD. Section 2.3 will specify key elements of visual-motor
skill acquisition and visual-motor task performance, focusing particularly on the
influence of task-specific features. Finally, bringing together the information
presented, section 2.4 will outline the research objectives.
2.1 CHARACTERIZING THE MOTOR ABILITIES OF CHILDREN WITH DCD
2.1.1 MOTOR DIFFICULTIES
The motor abilities of children with DCD have been described at length in the
literature. Regardless of the motor-based activity under scrutiny, as a group, children
with DCD are typically reported to be less accurate and more variable in their trial-to-
trial performance when compared to their typically developing peers (e.g., Elders et
al., 2010; Estill et al., 2002; Johnston et al., 2002; Smits-Engelsman et al., 2006).
With regards to their movement speed, two characterizations seem to emerge. The
first one indicates that children with DCD have difficulty meeting accuracy demands
of tasks and move slower than their peers (e.g., Elders et al., 2010; Maruff, Wilson,
Trebilcock, & Currie, 1999; Smits-Engelsman et al., 2006; Johnston et al., 2002).
The second one suggests that children with DCD do not have an efficient strategy
emphasizing terminal accuracy and move faster than their peers (e.g., Elders et al.,
2010; Smits-Engelsman, Niemeijer, & van Galen, 2001; Smits-Engelsman, Wilson,
Westenberg, & Duysens, 2003). Although such general descriptions of motor
abilities make up the clinical portrait of children with DCD, it is also well
8
acknowledged that the specific difficulties that each child experiences are quite
varied (Dewey & Wilson, 2001; Szklut & Breath, 2001). The motor difficulties of a
child with DCD may predominantly affect gross motor skills, fine motor skills, or both
(Szklut & Breath, 2001; Polatajko, 1999). One child may have difficulties with
discrete finger movements, another with eye-hand coordination. One child may have
poor balance, while another may have reached developmental milestones later than
expected (Szklut & Breath, 2001; Polatajko, 1999). Accordingly, when considering
children with DCD as a group, within-group variability, or heterogeneity, is also
characteristic of their motor abilities (Szklut & Breath, 2001; Dewey & Wilson, 2001).
The poor motor abilities of children with DCD affect their performance of many
motor-based childhood tasks and activities (APA, 2000; Fox & Lent, 1996; Geuze,
Jongmans, Schoemaker, & Smits-Engelsman, 2001; Magalhães et al., 2011).
Recently, Magalhães and colleagues (2011) conducted a comprehensive systematic
review of studies describing the impact of DCD on children’s performance of daily
activities. They collected information from descriptive, intervention, and qualitative
studies published between 1995 and 2008. In all, 44 studies were included, and the
most frequently reported issue was difficulty with handwriting in the classroom
(which 70% of the articles, i.e., 31 of the 44, examined). This result echoes previous
reports that difficulty with handwriting is the primary reason for the referral of
children with DCD to health care professionals (Miller et al., 2001; Missiuna et al.,
2008). In addition to poor handwriting performance, other problematic activities
frequently identified in the literature relate to play activities such as bicycle riding,
catching, throwing balls or jumping rope, classroom activities such as using scissors
or drawing, and self-care activities such as buttoning, doing up zippers or using
cutlery (Astill & Utley, 2008; Kennedy-Behr et al., 2011; Magalhães et al., 2011;
Miller et al., 2001; Rodger et al., 2003; Stephenson & Chesson, 2008; Summers,
2008; Wang et al., 2009).
Together, the poor motor abilities and the motor-based performance difficulties of
children with DCD are the main defining characteristics differentiating them from
typically developing children (APA, 2000). However, contrary to children with
9
diagnoses such as cerebral palsy or muscular dystrophies, who are also
characterized by poor motor abilities and motor-based performance difficulties,
children with DCD experience difficulties that do not result from known general
medical conditions or diagnoses (APA, 2000). In fact, contrary to children with
specific diagnoses known to affect their motor abilities, research has shown that
children with DCD can overcome their difficulties through proper intervention and
learn to perform complex activities such as riding a bicycle, goaltending in hockey or
tying shoelaces (Mandich, Polatajko, Missiuna, & Miller, 2001; Miller et al., 2001;
Rodger et al., 2003). Not only can they acquire the skills necessary to perform such
activities but they can also go on to perform them in a way that is comparable to that
of their peers or can transfer and generalize the acquired skills to novel contexts and
activities (Mandich et al., 2001; Miller et al., 2001). The positive results reported in
intervention studies thus bring us to question whether the nature of the motor-based
performance difficulties of children with DCD actually lies within the realm of
performance or whether it lies within the realm of skill acquisition.
2.1.2 VISUAL MOTOR DIFFICULTIES
The difficulty of children with DCD in acquiring visual-motor skills and in performing
visual-motor tasks and activities is a predominant characteristic widely recognized
by clinicians and frequently reported in the literature (e.g., Dunford, Missiuna, Street,
& Sibert, 2005; Magalhães et al., 2011; Summers, 2008). In 1998, a now frequently
cited meta-analysis by Wilson and McKenzie was the first comprehensive synthesis
of the DCD literature to suggest the possibility of a visual-motor deficit in children
with DCD.
Wilson and McKenzie conducted a meta-analysis to identify processing deficits most
strongly associated with the motor difficulties of children with DCD. They examined
studies that compared control children to children meeting the diagnostic criteria for
DCD on perceptual and motor measures in studies published before 1996. In all, 50
studies were included in the meta-analysis. The studies were coded and placed in
one of five general categories: visual processing, other perceptual processing, motor
control, general intelligence and motor skill. Each category was further divided into
10
subcategories. Visual processing was comprised of ophthalmic, visuoperceptual
(non-motor), complex visuospatial (motor) and visuospatial memory factors. After
computation and analysis of effect sizes, Wilson and McKenzie concluded that while
the DCD groups’ performance was inferior to that of control groups in all categories
examined, the largest of these differences occurred in the complex visuospatial
(motor) skills (r = .55) category, which included what they named “visuospatial tasks”
that involved a motor component, such as graphic design or copying.
What Wilson and McKenzie defined as “complex visuospatial skills” is a concept
frequently named visual-motor skills in the literature (e.g., Martin, 2006; Wolpert &
Flanagan, 2010). While the term motor skill typically refers to a skill in which both the
movement and the outcome of the movement are emphasized (Newell &
Vaillancourt, 1991), the term visual-motor skill typically refers to a skill for which the
involvement of the visual system is required to guide the movement (Schmidt &
Wrisberg, 2008). Wilson and McKenzie (1998) proposed that children with DCD
have deficits in the processing of visual information related to the properties of tasks,
which would then be expected to lead to motor-based performance problems for
activities requiring the guidance of the visual system. Given the nature of their study,
the authors did not extrapolate their findings to further clarify how such a processing
deficit could also affect visual-motor skill acquisition.
Since the study by Wilson and McKenzie (1998), a number of researchers have
explored the performance of children with DCD on fine motor tasks requiring the
interaction of the visual and motor systems. The performance of children with DCD
has been primarily described for visual-motor tasks requiring handheld tools, more
specifically for writing (Chang & Yu, 2010; Di Brina et al., 2008; Rosenblum &
Livneh-Zirinski, 2008; Tseng, Howe, Chuang, & Hsieh, 2007) and for drawing
(Kagerer, Bo, Contreras-Vidal, & Clark, 2004; Kagerer et al., 2006; Smits-
Engelsman et al., 2006; Smits-Engelsman et al., 2001; Smits-Engelsman, Wilson,
Westenberg, & Duysens, 2003; Zwicker et al., 2011). The next section summarizes
the literature on visual-motor skill acquisition and visual-motor task performance in
children with DCD with regard to fine motor tasks. The studies presented will bring to
11
light the accumulating evidence supporting the findings of Wilson and McKenzie
(1998), suggesting the presence of a specific visual-motor deficit in children with
DCD.
2.2 DESCRIBING VISUAL-MOTOR SKILL ACQUISITION AND TASK
PERFORMANCE IN CHILDREN WITH DCD
Before continuing, it is useful to make an aparté here, or to digress for a moment, to
draw attention to the terminology used to describe different levels of performance.
Until now, task performance in this thesis has implicitly described a set of
purposeful, observable movements that have a product or outcome and that may
involve tools or materials (Polatajko et al., 2007). It has also been inferred that task
performance forms part of a bigger picture, one that occupational therapists would
call occupational performance (Polatajko et al., 2007).
To clarify further, let us consider writing as an example. Looking at a child writing a
story in his classroom would be exploring the child’s occupational performance.
Similarly, section 2.1 summarized the occupational performance of children with
DCD when discussing their motor-based performance difficulties in many childhood
activities. Describing the process and end result of how a child sharpens his pencil,
writes words and letters or erases mistakes would be characterizing the child’s task
performance. Finally, continuing with the example of writing, reaching, grasping,
flexing or remembering are voluntary movements and mental processes involved in
tasks related to writing.
A review of the literature attempting to characterize visual-motor skill acquisition and
task performance in children with DCD could thus target any, or all, of these levels of
performance. Nevertheless, the objectives of this thesis, which will be described in
section 2.4, are situated within the realm of task performance; therefore, section 2.2
will focus on describing visual-motor skill acquisition and task performance in
children with DCD. Since no synthesis of this evidence exists at the current time, it is
considered an essential first step before going further in the systematic investigation
of this topic. Furthermore, it is imperative that it be done here to provide the current
12
state of the evidence and to offer the context for the research objectives addressed
in this thesis.
To organize the literature in this section, studies are categorized based on the goal
of the visual-motor task under study. In writing and drawing, the goal of the task is to
follow a specific and planned trajectory, which calls upon different control
mechanisms than a fine motor aiming task (e.g., moving a computer cursor to a
target using a computer mouse), which calls upon end-point accuracy, regardless of
trajectory accuracy.
2.2.1 WRITING TASKS
Three studies compared the writing process and product of children with DCD to that
of their typically developing peers (Chang & Yu, 2010; Di Brina et al., 2008;
Rosenblum & Livneh-Zirinski, 2008). Di Brina and colleagues (2008) conducted a
two-group comparison study to explore the usefulness of dynamic time warping as a
new method for describing the writing process and resulting written product of
children with DCD (n = 20; mean age = 8 years 0 months + 10 months) and without
DCD (n = 20; mean age = 8 years 7 months + 8 months) from Dutch schools.
Children were asked to write the cursive letter “a” 20 times, under three conditions:
(1) normal condition, forming each letter detached; (2) fast condition, going as fast
as possible; and (3) accurate condition, staying between two solid horizontal lines.
Variables analyzed were writing time, trajectory length, velocity, duration of pauses
and pen pressure. Variables were entered into a repeated-measures general linear
model for analysis. Overall, their results revealed much higher variability in letter
forms, higher pen pressure and faster pen movements for children with DCD.
However, it would seem that the different writing conditions influenced, or
modulated, writing performance differently for the two groups. For example, during
the second condition (writing fast), no group differences were detected in variability
of letter forms; children without DCD showed increased variability in letter forms in
the first condition (normal), but children with DCD showed decreased variability in
letter form from the first condition (normal). Because the main objective of the study
was a descriptive one, the authors did not offer further explanations for the
13
differential influence that the speed requirements of the task had on the writing
performance of children with and without DCD.
Somewhat different results were obtained by Rosenblum and Livneh-Zirinski (2008),
who investigated the handwriting process and resulting written end-product
characteristics of Hebrew-speaking children with DCD (n = 20; mean age = 8 years
0 months) and without DCD (n = 20; mean age = 7 years 9 months) to explore which
variables best differentiate children with and without DCD. For that purpose, they
used the Computerized Penmanship Evaluation Tool (CPET; Rosenblum et al.,
2003) to describe processes related to handwriting (name writing, alphabet
sequence, and paragraph copying tasks), and the Hebrew Handwriting Evaluation
(HHE; Erez & Parush, 1999) to evaluate handwriting product. For each of the tasks
of the CPET, the variables analyzed were on-paper and in-air time per stroke, stroke
width and height, pen pressure and tilt. For the HHE, the variables analyzed were
global legibility, number of letters erased and spatial arrangements. Student t-tests
(with Bonferoni’s adjustment) and Mann-Whitney U tests were used to compare
handwriting process and product of the two groups. Discriminant analyses were
conducted to explore predictors of group membership. With regards to the
handwriting process, the authors found that children with DCD spent more in-air time
and more on-paper time for each stroke. Contrary to the results obtained by Di Brina
and colleagues (2008), children with DCD in this study applied less pressure on the
page than their peers when writing. For the handwriting product characteristics,
children with DCD erased or overwrote letters and numbers more frequently and had
more difficulty with spatial arrangement and global legibility than their non-DCD
peers. However, contrary to the results obtained by Di Brina and colleagues (2008),
children with DCD wrote at a much slower pace than their peers. Discriminant
analyses revealed that the highest predictor of group membership was the number
of letters erased, followed by legibility and pen pressure. An important limitation of
this study was the data analysis procedure. As a group, children with DCD are
recognized for their pattern of high within-group variability. Although the authors did
not specifically discuss the standard deviations of the variables collected, closer
examination of the results revealed that, sometimes, the standard deviations of the
14
DCD group for the mean value of a given variable were more than four times higher
than those of the non-DCD group (Rosenblum & Livneh-Zirinski, 2008). Moreover,
on certain variables, the standard deviation was almost as high as the mean.
Performing t-tests in such conditions fails to acknowledge the large variation within
the data and its potential impact on the mean values of the variables and on the
results of the statistical tests (Portney & Watkins, 1993).
Using computerized movement analyses, Chang and Yu (2010) aimed to
characterize handwriting deficits in young Chinese-speaking children identified as
having a handwriting deficit with DCD (n = 33; mean age = 7 years 5 months + 8
months) and without DCD (n = 39; mean age = 6 years 11 months + 6 month). A
group of children without handwriting deficits and without DCD (n = 22; mean age =
6 years 10 months + 7 months) was also recruited. Children had to write three
simple Chinese characters, followed by three difficult characters that involved more
strokes and turning points. Interestingly, in addition to exploring the control of
handwriting movements through variables derived from stroke velocity and axial pen
pressure, Chang and Yu also explored aspects of children’s skill acquisition by
looking at changes in task performance over repeated trials. To characterize skill
acquisition, they modelled directional changes of velocity over repeated trials using
an exponential decay function. Accordingly, skill acquisition was defined as a
63.21% (or 1-e-1) reduction in the number of directional changes of velocity. One-
way and two-way repeated measures analyses of variance were used to compare
the three groups. With regards to movement time, the authors reported that the
performance of children with DCD was modulated by the difficulty of the writing task.
When the task involved simple Chinese characters, children with DCD had faster
stroke velocity than their non-DCD peers. Conversely, when the writing task involved
more difficult Chinese characters, children with DCD had slower stroke velocity than
their peers. With regards to skill acquisition, the authors (Chang & Yu, 2010) report
that the acquisition rates of children with DCD and handwriting deficits (14.64 trials +
7.92) were statistically different from those of the non-DCD children with (7.64 trials
+ 5.11) and without (6.18 trials + 5.23) handwriting deficits. It is interesting to note,
however, that children with DCD did eventually reach the same level of performance
15
than the non-DCD children. No statistical differences were detected in the
acquisition rates of the non-DCD children with and without handwriting deficits.
2.2.2 TRAIL-DRAWING TASKS
Two studies compared their trail-drawing ability (Smits-Engelsman et al., 2001;
Zwicker et al., 2011). Smits-Engelsman and colleagues (2001) collected kinematic
measures of drawing movements in Dutch children with handwriting difficulties and
DCD (n = 12; mean age = 8 years 4 months) and without handwriting difficulties or
DCD (n = 12; mean age = 8 years 6 months) during the trail-drawing item of the
Movement Assessment Battery for Children (M-ABC) (Henderson & Sugden, 1992).
The M-ABC is a standardized, norm-referenced assessment tool used to evaluate
motor ability in children. It is frequently used in clinical settings and research studies
to identify children with DCD (Geuze et al., 2001). In addition to drawing errors (out-
of-bounds), other variables collected were trajectory length, movement time,
velocity, in-air time, and pen pressure. The variables were compared between the
two groups using Mann-Whitney U tests. The authors reported that when compared
to good writers, children with DCD made more drawing errors and moved faster than
their peers during task completion. It is important to note that, although the authors
did not speak to it, the standard deviation of the DCD group for movement velocity
was twice as large as that of the control group. As mentioned earlier, within-group
variability is frequently reported in studies of children with DCD. Such lack of
homogeneity of variance between groups could have threatened the validity of more
robust parametric statistical analyses if they had been selected. Nevertheless, the
description of the performance of children with DCD on an item of a test of motor
ability offered by Smits-Engelsman and colleagues (2001) is not surprising given the
characteristic motor difficulties of children with DCD. However, it is important to note
that a large number of children in the DCD group scored within the norm on the M-
ABC, thus bringing into question whether the group described here was actually a
group of children with DCD.
Zwicker and colleagues (2011) also used a trail-drawing item of the M-ABC
(Henderson & Sugden, 1992). They conducted a two-group comparison study to
16
explore the influence of practice on patterns of brain activity, as measured by
functional MRI (fMRI), in children with DCD (n = 7; mean age = 10 years 10 months
+ 2 months) and without DCD (n = 7; mean age = 10 years 11 months + 2 months).
Given the limited movements allowed in an MRI, they used a joystick strapped to
children’s fingers and held like a pen to perform the task. The task was practiced, in
four 2-minute blocks of practice per day, over 5 days. In addition to fMRI-related
variables, visual-motor task performance variables measured included the number
of traces completed and the time per trace, as well as the number of errors.
Repeated measures analyses of variance were conducted to explore between-group
and within-group differences before and after practice. According to the authors,
their results suggested that visual-motor task performance of children with DCD was
as good as that of their non-DCD peers initially and after repeated practice.
Furthermore, neither group showed a statistically significant increase in accuracy or
time per trace in a comparison of pre- and post-practice performance. In other
words, statistical analyses revealed no skill acquisition despite repeated practice.
The authors suggested that power issues related to a small sample size could partly
explain these findings. To further investigate their results, they looked at effect sizes
within and between groups. This analysis suggested a trend in which children with
DCD took less time for each trace after practice but made more errors, while
typically developing children took the same amount of time per trace but made fewer
errors. Additional testing would be needed to confirm whether such group trends
hold with a larger number of participants. Finally, despite the seemingly absent skill
acquisition measured, statistically significant differences in patterns of brain
activation were detected pre- and post-practice. Further examination of the data
demonstrates that, here again, within-group variability was quite large. In a number
of cases, standard deviations were more than twice as large as those of the non-
DCD children, and also sometimes almost as large as the mean value of the
variable measured. Again, such within-group variability could threaten the validity of
the information gained from the statistical analyses conducted.
2.2.3 FINE MOTOR AIMING TASKS
17
In a two-group comparative study, Missiuna (1994) investigated the influence of
practice on the skill acquisition process of children with DCD (n = 24; mean age = 7
years 5 months + 9 months) and without DCD (n = 24; mean age = 7 years 6
months + 10 months). She also explored the effect of changing task parameters on
task performance. The visual-motor task practiced to investigate skill acquisition was
a computer task in which children were required to move a cursor (shuttle) to a
target (planet) as fast as they could using a computer mouse. The variables
measured were movement time and reaction time. Children performed the task in
blocks of 16 trials until their movement time reached a plateau, at which time the
skill was considered acquired. Then children were presented with a similar task but
with altered parametric features: different target sizes, movement amplitudes,
movement directions and graphic layout (baseball and baseball glove). According to
the author, the results suggest that while visual-motor task performance was
different in children with DCD when compared to that of their peers, the pattern of
visual-motor skill acquisition (i.e., number of blocks required for skill acquisition) was
similar. With regards to the impact of changing task parametric features on
performance, children with DCD were apparently more affected than their peers
when accuracy requirements were increased. Closer examination of the results
showed that after the first block of 16 trials, a statistical group comparison revealed
no differences in movement time. With practice, children in both groups became
faster, reportedly following a similar pattern of change and acquisition rate.
Nevertheless, after the sixth block of trials (i.e., 96 trials), the two groups were
statistically different on movement time, with children with DCD being slower than
their peers. Such results seem incoherent, since the absence of an initial group
difference but presence of a final group difference should necessarily lead to a
difference in pattern or rate of acquisition. A possible explanation could be linked to
limitations of the statistical analyses chosen in the presence of large within-group
variability. The author proposed that such high within-group variability could have
masked an initial between-group difference. But even then, the fact that, as a group,
children with DCD were not different from their peers initially, or that their pattern of
skill acquisition was similar to that of their peers, is inconsistent with findings
18
commonly reported in the literature. Different explanations for such results are
possible. First, the simplicity of the task could be to blame for the lack of group
differences initially measured and for the skill acquisition pattern of children with
DCD. It is not the simple visual-motor tasks that seem to be problematic for children
with DCD, but, rather, the more difficult ones. Children with DCD can acquire the
skills necessary to control their pens and to draw circles, triangles, or squares. It
seems that it is when they must apply such pen control to writing words and
sentences that they struggle (Kirby, 2011). Another possible explanation could be
the method chosen for data analysis. In addition to the already mentioned limitations
of performing group comparisons when high within-group variability exists, blocking
performance in groups of 16 data points greatly reduces the amount of information
that can be extracted during data analysis. Task performance during the acquisition
of a simple skill is likely to change fairly quickly. Chang and Yu (2010), as discussed
earlier, reported skill acquisition rates between 6 and 14 trials. It is possible that the
blocks compared were too large to allow differences in patterns of skill acquisition
between children with and without DCD to emerge.
Smits-Engelsman and colleagues (2003) used Fitts’s law (Fitts, 1954) to explore
whether children with DCD are sensitive to the accuracy demands of visual-motor
tasks. Fitts’s law postulates that aiming to targets of various sizes will lead to
predictable speed-accuracy trade-offs (Fitts, 1954). In their study, Smits-Engelsman
and colleagues (2003) investigated the modulating effect of target size and
movement type on visual-motor task performance (movement time, trajectory length,
end-point accuracy and end-point spread) in children with DCD and learning
disabilities (n = 32; mean age = 11 years 4 months) and in typically developing
children (n = 32; mean age = 11 years 2 months). Children with DCD and learning
disabilities (LD) were recruited from a school for children with LD; their motor
abilities were confirmed either through a standardized motor assessment or an
evaluation of their handwriting. Typically developing children were recruited from
normal primary schools. Data were recorded while performing a reciprocal pen-
aiming task on a digitizing tablet. The task was modified by changing target size
(0.22, 0.44 and 0.88 cm) and the type of movement required to perform the task
19
(from discrete to cyclical movements). As predicted by Fitts’s law and as
demonstrated in typically developing children, the results reported suggest that the
task performance of children with LD/DCD was modulated by target size. Also, and
as anticipated, the type of movement required affected task performance both in
children with LD/DCD and in their peers, cyclical movements leading to decreased
visual-motor performance. However, the authors report that the influence of cyclical
movements on task performance was larger in children with LD/DCD than in their
peers; during cyclical movements, children with LD/DCD moved faster, made more
end-point errors and were less sensitive to target size. According to the authors,
their results suggest that children with LD/DCD may have difficulty with the
predictive and ongoing online control of hand movements. While authors are often
tempted to speculate on specific deficits when analyzing their results, it is important
to recognize that the main conclusion that can be drawn from this study is that the
task performance of children with LD/DCD is modulated by the accuracy demands of
visual-motor tasks in the same way as it is for their peers. Another conclusion that
can be reached is that the modulating effect of movement type on task performance
was larger for children with LD/DCD than for their peers. Specifically, fast, repetitive
aiming drawing movements presented a greater challenge for children with LD/DCD.
However, before applying such findings to children with DCD, it is important to
recognize that the sample selected for this study came from a school for children
with LD and that motor abilities were not assessed in 38% of the sample, other than
through a handwriting assessment. This limits the generalizability of the findings to
children with DCD.
In a further study, Smits-Engelsman and colleagues (2006) explored the modulating
effect of target position on the visual-motor performance of children with DCD (n =
48; mean age = 7 years 11 months) and without DCD (n = 48; mean age = age
matched) by asking children to use a pen to reach targets situated either at the
midline, the contralateral, or the ipsilateral side of the midline of the body. Because
of the neuronal circuits involved in such movements, the authors proposed that the
effect of target position on task performance would be larger for children with DCD
than for their peers. Overall, children with DCD moved slower than their peers, but
20
the effect of target position on their visual-motor task performance was not different
than its effect on the performance of their peers. In other words, while the task of
reaching to the contralateral side of the body affected accuracy to a greater extent
than when reaching to the ipsilateral side of the body, this effect was not greater in
children with DCD.
Finally, Kagerer and colleagues (Kagerer et al., 2004; Kagerer et al., 2006) carried
out two studies exploring the influence of abrupt and gradual rotations of visual
feedback on task performance in children with DCD (2006: n = 10; mean age = 8
years 2 months + 18 months) and without DCD (2006: n = 10; mean age = 8 years 6
months + 13 months) during a centre-out aiming task. In both studies, children were
seated in front of a computer monitor, with one of their arms resting on a digitizing
tablet. Vision of the hand/arm was occluded, and visual feedback of hand movement
on the tablet was presented on the computer monitor. The task required children to
draw a line from a centre position to one of three targets appearing randomly at one
of three positions (80°, 200° and 320°), moving as fast and as straight as possible.
During the exposure condition, feedback of pen movements on the tablet was
distorted by either applying a gradual (increments of 10° for 21 trials, up to 60°) or
abrupt (60°) clockwise rotation. Afterwards, the distortion was removed and children
were required to perform the centre-out aiming task again to evaluate whether the
distorted visual feedback had had a lasting effect on the children’s visual-motor
control. Results reported confirm that the visual-motor task performance of both
groups of children was influenced by the rotated visual feedback. Furthermore,
despite the rotated feedback, during the exposure condition, task performance
improved in both groups. With regards to the influence of visual feedback distortion
on the task performance of both groups post-exposure, some discrepancies
between the two studies are evident. In the first study, only an abrupt distortion of
visual feedback was applied. Results demonstrated that such a distortion did not
influence the post-exposure performance of any of the groups (Kagerer et al., 2004).
In the second study, both abrupt and gradual distortions of visual feedback were
applied. This time, results demonstrated that an abrupt distortion did influence the
post-exposure performance of both groups. However, the effect of a gradual
21
distortion of visual feedback on task performance was not the same for both groups.
While it did affect post-exposure performance for typically developing children, it did
not affect the performance of children with DCD (Kagerer et al., 2006). The authors
report that the main reason for such a difference in the results of the two studies is
that twice as many trials were allotted to the exposure condition (126 trials) in the
second study, thus likely allowing the error signals received from the visual feedback
to be better integrated than in the first study, giving more time for skill acquisition to
occur.
2.2.4 SUMMARY
Considering the findings summarized in section 2.2, it is evident that the current
literature on visual-motor skill acquisition and visual-motor task performance in
children with DCD provides a rather incomplete and inconsistent picture. Results
from most studies support the clinical observation that the visual-motor task
performance of children with DCD differs from that of their peers. Performance is
characterized by poor end-point or ongoing accuracy (Chang & Yu, 2010; Di Brina,
2008; Rosenblum & Livneh-Zirinski, 2008; Smits-Engelsman et al., 2001, 2003;
Tseng et al., 2007). However, just as in the earlier description of the motor abilities
of children with DCD in section 2.1, divergent characterizations emerge with regards
to their movement times: in a number of studies, children with DCD moved slower
than their peers (Chang & Yu, 2010; Rosenblum & Livneh-Zirinski, 2008; Smits-
Engelsman et al., 2006), while in others, they moved faster (Chang & Yu, 2010; Di
Brina et al., 2008; Smits-Engelsman et al., 2001, 2003). Such a description of the
visual-motor abilities of children with DCD falls short of a detailed characterization
that would lead to a better understanding of the nature of their difficulties. As already
mentioned, developing a better understanding of the nature of the difficulties of
children with DCD has critical implications for researchers and clinicians working
with these youngsters.
Moving beyond the summarized findings discussed above, it is useful here to
consider the studies that did attempt to better understand the nature of the
performance difficulties of children with DCD by exploring which task-specific
22
features actually had a modulating effect on children’s task performance, more
specifically on their movement time and movement accuracy. The effect of target
position and target size on task performance was found to be the same for children
with DCD and their peers (Smits-Engelsman et al., 2001, 2003). The effects of task
difficulty (Chang & Yu, 2010), speed requirements of a task (Di Brina et al., 2008)
and type of movement (cyclical versus discrete) (Smits-Engelsman et al., 2003)
were found to modulate visual-motor task performance differently for children with
DCD when compared to their peers. At this time, the paucity of studies
systematically exploring the modulating effect of task-specific features on
performance limits our ability to extrapolate the findings summarized here and to
attempt a better understanding of the nature of the visual-motor task performance
difficulties in children with DCD.
Conspicuous by its absence in this section is a systematic investigation of visual-
motor skill acquisition in children with DCD. Given the continuum on which visual-
motor skill acquisition and visual-motor task performance lie, and given the known
performance-based difficulties of children with DCD in tasks and activities that have
a visual-motor component, it is striking to find only three studies specifically
exploring visual-motor skill acquisition in children with DCD (Chang & Yu, 2010;
Missiuna, 1994; Zwicker et al., 2010). As mentioned earlier, evidence from
intervention studies would suggest that the nature of the observed motor-based
difficulties of children with DCD could very well reside within the realm of skill
acquisition. Yet two of the three studies of visual-motor skill acquisition suggest that
the pattern of skill acquisition in children with DCD is similar to that of their peers. It
is clear from this section that a systematic exploration of visual-motor skill
acquisition in children with DCD is missing from the literature but is urgently needed
to develop a better understanding of the nature of their visual-motor task
performance difficulties through a better understanding of their skill acquisition.
Before discussing visual-motor skill acquisition and task performance in more detail,
it is important to highlight here a recurring limitation of most of the studies discussed
in section 2.2. The studies described earlier have all used comparative study
23
designs, using motor ability to define groups of children to be compared. Such a
study design and approach to data analysis fails to acknowledge the well-known
within-group variability characteristic of children with DCD and leads to major
limitations. Within-group variability is the norm in studies of children with DCD (King,
Harring, Oliveira, & Clark, 2011; Geuze et al., 2001). Regardless of the task under
study, most authors who have looked at individual differences within groups of
children with DCD often report that a few children end up performing better than
their peers, while others are as good as they are and others still are worse (Cantin,
Polatajko, Thach, & Jaglal, 2004; Green, Chambers, & Sugden, 2008). As a
consequence, when group means are calculated, large standard deviations are
typically reported, thus threatening the assumption of the homogeneity of variances
required by most statistical tests used in comparative study designs (Portney &
Watkins, 1993), Furthermore, the information that could have been gained from such
individual differences is lost. It then becomes difficult to know whether reported
group differences are due to a few outliers or whether they truly represent the
performance of the group. In addition, children with DCD are also often reported to
be variable in their own trial-to-trial performance (Cantin et al., 2004; King et al.,
2011). Thus, when task performance over a certain number of trials is averaged, this
trial-to-trial characteristic variation is lost and can again be greatly influenced by
outliers (Motulsky & Brown, 2006). Averaging performance across trials also fails to
account for the possibility of performance changes across trials; in fact, such
changes could even be different for children with DCD than for their peers. Failing to
capture this information constitutes a major shortfall of the choice of statistical
analyses and study designs reported in section 2.2. While numerous pitfalls are
associated with conducting group comparisons when studying children with DCD,
most studies summarized in this section still used such an approach. An alternative
methodology is thus urgently needed to adequately take into account the within-
group and within-child variability characteristic of children with DCD.
Before continuing to propose study objectives that would allow a systematic
description of visual-motor skill acquisition and task performance in children with
24
DCD, the next section offers a general overview of the current understanding of the
nature of visual-motor skill acquisition and resulting task performance.
2.3 THE NATURE OF VISUAL-MOTOR SKILL ACQUISITION AND TASK
PERFORMANCE
The ability to acquire visual-motor skills and to perform visual-motor tasks is often
considered a defining feature of human life, the central component of many daily
activities (Crawford et al., 2004). Although the link between skill acquisition and task
performance is not often clarified in studies exploring one or the other (Newell &
Vaillancourt, 1991), it is typically agreed that skilful visual-motor task performance
results from skill acquisition and motor learning (Gentile, 1998). Skilful visual-motor
task performance is characterized by increased efficiency in performance (Gentile,
1998), which is often operationalized to imply decreased movement times,
decreased end-point errors, and increased efficiency of movements (such as
straighter movement paths) (Schmidt & Lee, 2011).
In adults, skill acquisition and opportunity for practice is usually considered the most
influential factor predicting improvement in task performance (Guadagnoli & Lee,
2004; Thomas & Thomas, 2008; Wolpert & Flanagan, 2010). Opportunity for
practice is also considered to be influential in children, although growth and
maturation are also necessary in explaining changes in performance (Bard, Hay, &
Fleury, 1990; Thomas & Thomas, 2008).
It is generally well accepted that skilful motor task performance depends on a
number of interacting components including, but not limited to, acquiring an
understanding of the impact of task-specific features on task performance and being
able to generate appropriate motor commands resulting from the combination of
different motor control mechanisms (Wolpert & Flanagan, 2010). The influence of
both task-specific features and motor control mechanisms on visual-motor skill
acquisition and task performance will be discussed in turn in the next two
subsections.
25
2.3.1 TASK-SPECIFIC FEATURES
Task-specific features are features that are intrinsic to the task (Wolpert & Flanagan,
2010). As such, task-specific features are not determined or influenced by the
learner, although they do affect skill acquisition and task performance. The impact of
task difficulty, structural task features and parametric task features on visual-motor
skill acquisition and task performance is discussed in this section.
Task Difficulty
Intuitively, the difficulty of a task is probably one of the most influential features that
can affect skill acquisition and task performance. Guadagnoli and Lee (2004) have
proposed two broad categories to conceptualize task difficulty: (1) nominal difficulty
and (2) functional difficulty. Nominal difficulty is task specific. It reflects the inherent
perceptual and motor requirements of a task, regardless of the skill level of the
individual performing it or the context of performance. For example, writing the
number 1 is of lower nominal difficulty than writing the number 5 because of its lower
perceptual and motor requirements.
Conversely, functional difficulty takes into account the level of expertise of the
individual performing a task and the context within which performance occurs.
Accordingly, functional difficulty is not conceptualized as a task-specific feature but
rather as the connection that exists among inherent task difficulty, the individual and
the environment. Thus, functional difficulty attempts to capture the fact that task
performance will differ in relation to individual expertise (Figure 2-1)1. For a task of
low nominal difficulty such as writing the number 1, the performance of a junior
kindergartener is expected to be very close to that of a first grader. However, for a
task of higher nominal difficulty such as writing the number 5, the performance of the
junior kindergartener would be expected to be slower and less accurate than that of
the first grader. The concept of functional difficulty seems so intuitive that its
importance to the understanding of visual-motor task performance and skill
acquisition could easily be overlooked. Nevertheless, according to Guadagnoli and
1 Taylor & Francis, the publisher of this figure, offers reuse of its content for a thesis or dissertation
free of charge contingent on resubmission of permission request if work is published.
26
Lee, the concept of functional difficulty is essential to consider within the context of
task performance, especially when exploring skill acquisition.
Figure 2-1 Relationship between task difficulty and expected performance as a function of the
individual's skill level.
27
Guadagnoli and Lee (2004) have proposed the Challenge Point Model to integrate
the concept of functional difficulty to skill acquisition. This model is built on the
premise that whenever an individual performs a task, the feedback received from
task performance—for example, visual feedback about end-point accuracy or
proprioceptive feedback about the efficiency of the movement—is information that
can be used by the individual to acquire skills and improve task performance. This
premise views skill acquisition as a problem-solving process that can be elucidated
through the repeated performance of the task to be acquired. Each time an
individual performs a task, the individual gains essential information about the task
that can be used to solve the problem at hand. However, Guadagnoli and Lee also
contend that information is only useful to an individual when it reduces uncertainty.
They offer an example with the statement, “It is dark outside”. While such a
statement carries little meaning and does not reduce uncertainty if spoken at night, it
would provide useful information when spoken during the day.
The Challenge Point Model proposes that the usefulness of the information gained
from task performance is in fact dependent on an individual’s skill level (Guadagnoli
& Lee, 2004). In other words, functional difficulty influences, or modulates, skill
acquisition. For example, let us consider two children: Julia, who is just starting to
ride her bicycle without training wheels, and Sophia, who enjoys mountain biking as
a hobby. For both children, riding a bicycle with training wheels on a cycle path is a
fairly easy task that represents a low level of nominal difficulty; their performance is
expected to be comparable. The information gained, or feedback received, from
performing this task is not useful to either of them as it does not provide information
that could be used to improve their performance.
When the training wheels are removed, the expertise that Sophia brings to the task
is evident. Sophia can ride her bicycle in a straight line, keep her balance, avoid
obstacles and maintain a constant speed, seemingly without much effort. Julia,
however, struggles significantly. She falls down repeatedly and has difficulty getting
enough momentum to get her bicycle going and to maintain her balance. For
Sophia, the information gained from her performance would not be useful and would
28
not improve her performance. Sophia can ride her bicycle under these conditions
almost without thinking about it; her body adjusts to the conditions of the cycle path
almost automatically. However, for Julia, the information gained from her
performance is very useful and should lead to, or promote, skill acquisition and
improved task performance. After falling down a few times, Julia should gain the
information necessary to solve the problem of balancing on her bicycle. By trying to
pedal to get her bicycle going, she will learn that to keep going, she has to keep
pedaling until she has gained enough momentum, and that balancing is easier to
maintain when the bicycle travels fast enough.
For Sophia to acquire new skills and to improve her bicycle riding performance, the
nominal difficulty of the task has to be increased significantly. For Sophia, it is when
she is in the woods, going down a trail with logs, rocks and other obstacles to avoid
or to use as props for acrobatics that she is truly challenged. It is then that the
information or feedback gained from her performance is useful to her. However, if
Julia were placed under the same conditions, the information received from her
performance would likely be overwhelming and would not contribute to reducing
uncertainty. As such, the information received neither would promote skill acquisition
nor would it lead to improved task performance. Indeed, Guadagnoli and Lee’s
Challenge Point Model (2004) proposes that high nominal difficulty promotes task
acquisition in an expert performer but provides too much information for a beginning
performer, impeding the task acquisition process. Conversely, low nominal difficulty
provides just enough information for a novice performer to acquire new skills, but not
enough for the expert performer to improve task performance. Thus, according to
the model, different patterns of skill acquisition and task performance are
conceivable when considering the concept of functional difficulty; the interaction
between an individual’s skills and abilities and a task’s nominal difficulty is expected
to modulate, or influence, skill acquisition and task performance.
The Challenge Point Model has not been discussed in relation to motor skill
acquisition and task performance in children, although it is reasonable to consider
that concepts of difficulty and skill levels are likely to apply to children as well as to
29
adults. The model offers an interesting theoretical framework to guide the systematic
exploration of motor skill acquisition and task performance in children with differing
levels of motor skills and abilities. As discussed earlier, Chang and Yu (2010) did
explore the modulating effect of task difficulty on task performance. However, their
exploration did not look into its impact on skill acquisition.
Structural Task Features
Structural task features are the overall movement shape structures required for
successful task performance (Gentile, 1998; Wolpert & Flanagan, 2010). These will
affect task performance and must be acquired through practice (Gentile, 1998;
Wolpert & Flanagan, 2010). When learning structural task features, the learner
develops a greater understanding of the rules linking motor actions to their sensory
consequences in relation to the specific task at hand. In the case of visual-motor
tasks, visual-motor transformations are what is learned (Wolpert & Flanagan, 2010).
When a handheld tool is involved, learning the way motor commands interact with
the tool and how the tool interacts with the environment is also part of structural
learning. For example, one need only watch a toddler trying to master the activity of
cutting with scissors by simply pushing the scissors through the paper to rip to
understand that the acquisition of the general shape of the movement is required
before any improvement in task performance can occur.
Parametric Task Features
Parametric task features are the task-specific force and timing components of
movements required for successful task performance. Similar to structural task
features, parametric task features must also be learned through practice (Gentile,
1998; Wolpert & Flanagan, 2010). Once the overall structural features of a task are
learned, then parametric features are learned or adapted from previously learned
visual-motor transformations (Wolpert & Flanagan, 2010). Parametric learning and
adaptation involves the acquisition of timing and force-generation processes
underlying movements, ultimately leading to the refinement of force dynamics
through the efficient control of internal forces and the prediction of external forces
(such as inertia) (Gentile, 1998). For example, once the structure of cutting with
30
school-sized scissors is learned, then that structure can be adapted to a structurally
similar activity featuring different parameter settings, such as cutting with larger
kitchen scissors, or stiffer paper. At that point, it is the parameters of that specific
task which must be learned or, as in the example of the scissors, which must be
adapted from the previously learned task of cutting with school-sized scissors.
During skill acquisition, a differential effect on the trial-by-trial changes in visual-
motor task performance has been reported depending on whether structural or
parametric task features must be acquired (Wolpert & Flanagan, 2010). A rapid
improvement in performance is typically reported when simple visual-motor tasks
with known structural features but novel parametric features are used, supposedly
because structure has already been acquired and only the acquisition or adaptation
of parametric task features is needed to meet the new task demands (Kagerer,
Contreras-Vidal, & Stelmach, 1997). For example, Jansen-Osmann, Ritcher,
Konczak and Kalveram (2002) investigated the adaptation of the visual-motor
transformation involved in a simple, visually guided hand movement in 6- to 10-year-
olds by exploring the children’s ability to adapt to the novel parametric features of
external damping forces. During the acquisition period, when damping forces were
applied, all groups of children demonstrated improved performance; the younger
group quickly adapted to the new parametric features within five trials. During post-
exposure trials, all groups demonstrated aftereffects, confirming that they had
successfully updated their visual-motor transformation of this simple, visually guided
hand movement to include the new parametric features.
However, when novel, complex visual-motor tasks are used, no initial improvement
in performance is typically reported, confirming that the acquisition of structural task
features is necessary before any performance improvement can happen (Sailer,
Flanagan, & Johansson, 2005). For example, for adults, Sailer and colleagues
(2005) designed a complex visual-motor task in an attempt to explore the
relationship between gaze behaviour and the acquisition of complex visual-motor
skills. A control using torques and longitudinal forces was used to move a cursor on
a computer screen and to successively hit displayed targets. Three distinct
31
acquisition phases were observed in regards to visual-motor task performance in the
majority of their participants. At first, during an exploratory stage, performance was
poor: the rate at which subjects hit targets was low and the cursor trajectory was
variable and inconsistent. During the second stage, the skill-acquisition stage,
performance increased quickly: hit rate increased steadily, as did the accuracy of
cursor trajectory. Finally, during the last stage, the skill-refinement stage,
performance gradually levelled off: hit rate and cursor trajectory gradually reached a
plateau.
To summarize, the impact of three task-specific features on visual-motor skill
acquisition and task performance was discussed in this section. First, the Challenge
Point Model (Guadagnoli & Lee, 2004) was presented to introduce the concept of
task difficulty and to discuss it in relation to level of expertise, skill acquisition and
task performance. Then, the influence of structural and parametric task features on
task performance and skill acquisition was introduced, along with the notion of
visual-motor transformations. The concept of visual-motor transformations is further
discussed in the next subsection (2.3.2), as it is a construct that relates mainly to
motor control mechanisms.
2.3.2 MOTOR CONTROL MECHANISMS
Within the context of task performance, motor control can be defined as the process
of transforming sensory inputs into motor outputs; in other words, transforming the
sensory feedback from descending efference copy of motor commands and from
sensory organs into motor commands that will affect muscles necessary for skilful
performance (Wolpert & Flanagan, 2010). Different motor control mechanisms have
been proposed to coexist and combine in different ways to enable skill acquisition
and task performance. Accordingly, in this section, feedback and forward motor
control mechanisms are discussed to illustrate how they combine to enable visual-
motor skill acquisition and task performance. Finally, particulars of motor control
mechanisms in children will be discussed.
32
Feedback Motor Control
Sensory feedback loops provide essential knowledge about one’s body position in
space and about the context within which task performance is occurring (Flanagan,
Vetter, Johansson, & Wolpert, 2003; Wolpert, Ghahramani, & Flanagan, 2001).
Sensory feedback plays a critical role in acquiring sensory-to-motor transformations
by enabling the learner to develop a greater understanding of the rules linking motor
actions to their sensory consequences (Wolpert & Flanagan, 2010).
Considering visual-motor task performance, Milner and Goodale have proposed a
model of cortical visual processing that delineates the distinction between vision for
perception and vision for action. The role of vision for perception is to transform
visual inputs into perceptual representations that correspond to the permanent
characteristics of objects. The role of vision for action is to transform visual
information about a task, on a moment-to-moment basis, into appropriate
coordinates for the effector being used (Milner & Goodale, 2006). While vision for
perception serves to gather information required to form a visual percept that retains
details about the relative spatial relationship of an object to others within a visual
arrangement, vision for action plays a critical role in rapid (and automatic) goal-
directed movements by providing the detailed specification and online control of the
movements that form the action (Milner & Goodale). Vision for action and vision for
perception have complementary roles that enable successful task performance.
Sensory organs other than the eye also provide valuable sensory feedback to the
cortex about one’s body position and movement in space. However, for most natural
movements, simple reliance on sensory feedback would not enable smooth,
coordinated task performance (Wolpert et al., 2001). Sensory signals transmitted in
feedback mechanisms are open to significant interference and delays and, as such,
present a major obstacle to skilful task performance (Wolpert et al., 2001). Forward
motor control mechanisms are thought to solve that problem.
Forward Motor Control
To solve the problem of interference and delay from feedback mechanisms, forward
motor control mechanisms use efferent copies of outgoing motor commands to
33
predict the sensory consequences of movement. For tasks requiring the use of tools,
forward motor control mechanisms predict the mechanical consequences of tools
based on the efferent copies of the outgoing motor commands (Wolpert & Flanagan,
2010). Such mechanisms thus capture the forward relationship involved in the
dynamic of movements (Schmidt & Lee, 2011; Wolpert et al., 2001).
The ability to predict the consequences of outgoing motor commands is not inherent,
it must be acquired. In this case, forward transformations linking outgoing motor
commands to predicted consequences are the ones learned (Davidson & Wolpert,
2005; Flanagan et al., 2003; Mehta & Schaal, 2002; Pearson, 2000; Wolpert &
Flanagan, 2010). In the context of the acquisition of forward transformations, the
goal is to acquire the rules linking consequent sensory inputs to generated motor
outputs (Wolpert et al., 2001). During the acquisition of forward transformations, a
discrepancy signal is produced by contrasting information gained from the actual
sensory feedback received through task performance against the sensory output
predicted (Wolpert et al., 2001). This discrepancy signal, or error, then serves as
information that can be used to adapt or update task-specific forward
transformations (Gentile, 1998; Wolpert et al., 2001).
Feedback control mechanisms thus take on an important role during skill acquisition,
mostly during the early stages (Davidson & Wolpert, 2005; Flanagan & Johansson,
2002). With repeated task performance, forward transformations become more
accurate (Flanagan et al., 2003), leading in turn to greater accuracy in task
performance, along with a more rapid movement execution (Arbib, Bonaiuto,
Jacobs, & Frey, 2009; Frey, 2007; Jensen & Korff, 2004; Sailer et al., 2005).
Motor Control Mechanisms in Children
During childhood, motor control mechanisms and the acquisition of forward
transformations are further mediated by growth, neural development and children’s
limited repertoire of visual-motor and forward transformations (Ferrel-Chapus, Hay,
Oliver, Bard, & Fleury, 2002; Zipp & Gentile, 2010). Studies have shown
developmental trends in the weight given to the different motor control mechanisms
during the performance of visual-motor tasks in children (Bard et al., 1990; Fayt,
34
Minet, & Schepens, 1993; Ferrel-Chapus et al., 2002). There appears to be a critical
period around 8 years of age when the online integration of visual feedback
throughout activity performance seems to emerge as a motor control strategy in
children. During that critical period, movement times are typically reported to be
longer than prior and after that critical period. Prior to that critical period, movements
are often characterized by a preprogrammed rapid acceleration, followed by a pause
to visually assess movement progress, followed by another preprogrammed
movement, and so on, until the end goal is reached. In this case, visual feedback is
not used throughout movement but is rather used in between movement segments.
Imagine a young child doing a dot-to-dot activity, first looking at the targeted dot,
then moving towards the target, stopping to look at the target again, and then
moving closer to the target, and so on until the dot is reached. After that critical
period, predictive motor control is used to accurately program initial movements, and
the online monitoring of visual feedback is used during the final honing phase to
ensure accuracy (Bard et al., 1990; Fayt et al., 1993; Ferrel-Chapus et al., 2002).
Thinking back to the dot-to-dot activity, an older child would perform the activity
fluidly, looking ahead at upcoming target dots while still moving the pen, likely
gazing at the target dot during the final phase of the movement. Looking back at the
DCD studies discussed in section 2.2, it is interesting to note that most of them
explored visual-motor performance in children aged 8 years or younger.
To summarize, in this section, two modes of motor control mechanisms were
discussed: feedback control and forward control. Although discussed in turn, it was
made evident that both mechanisms are essential and complement each other
during task performance and skill acquisition. The importance of development and
age on motor control was highlighted. Now that the nature of visual-motor skill
acquisition and task performance has been presented, the information will be
brought together to revisit the rationale for this study and to present the research
objectives.
35
2.4 INVESTIGATING VISUAL-MOTOR TASK ACQUISITION AND VISUAL-MOTOR
TASK PERFORMANCE IN CHILDREN WITH DCD: RATIONALE AND
RESEARCH OBJECTIVES
Section 2.1 highlighted the need to systematically explore and characterize visual-
motor skill acquisition and task performance in children with DCD to develop a better
understanding of their everyday difficulties, which would in turn greatly benefit
researchers and clinicians. However, it is evident from the few available research
findings presented in section 2.2 that the current DCD literature does not allow such
a characterization. It is especially clear that a systematic exploration of visual-motor
task acquisition in children with DCD is missing from the literature. Considering the
link between visual-motor task acquisition and later skilful performance, the paucity
of research in this area is disconcerting. For both researchers and clinicians, better
understanding the impact of children’s motor difficulty on their visual-motor task
performance and skill acquisition would provide invaluable information to explore the
nature of DCD and to intervene to overcome the children’s difficulties. Accordingly,
the general objective of this study was to describe visual-motor skill acquisition and
task performance in children with DCD.
In section 2.2 it was also noted that the impact of within-DCD group and within-DCD
child variability on the validity of the data analyses selected by most studies
reviewed has been largely ignored. Greater rigour in study designs and methods of
data analysis is needed to eventually develop a greater understanding of the nature
of the difficulties of children with DCD. Taking into account the expected within-
group and within-child variability in performance, mixed-effect modelling would be an
appropriate analysis strategy to use (Singer, 1998). While the particulars of such
data analysis will be further explained in section 3.4, it is important to recognize that
the selection of this analysis strategy shapes the formulation of the research
objectives. In mixed modelling, the influence and modulating effects of predictor
variables on specific dependent variables are explored. For example, in this case,
motor ability would be recognized as a child-specific variable that could potentially
predict or modulate visual-motor performance, rather than being a way to create two
36
groups to compare. As a result, specific research objectives were formulated to
adopt the language and methodology associated with such a strategy to data
analysis.
The formulation of specific research objectives was further guided by the information
presented in section 2.3. Indeed, the notion that task difficulty interacts with skill
level to modulate skill acquisition and task performance offered a promising avenue
for the systematic exploration and description of visual-motor skill acquisition and
task performance in children with DCD.
Consequently, considering children with and without DCD, the specific research
objectives were to describe:
1. the modulating effect of task difficulty and motor ability on children’s
performance of a visual-motor task with increasing levels of difficulty;
2. the modulating effect of task difficulty and motor ability on children’s
acquisition of visual-motor skills;
3. the impact of motor ability on patterns of changes in task performance during
the acquisition of visual-motor skills.
37
CHAPTER 3
METHODS
38
3 METHODS
An exploratory study was conducted to describe visual-motor skill acquisition and
task performance in children with DCD. To meet this general objective, performance
of a visual-motor task with increasing levels of difficulty was measured and
described when children with and without DCD first performed the task, and over
repeated trials.
This study was approved by the Health Sciences II Ethics Review Board at the
University of Toronto. Prior to participation in the study, informed written assent and
consent was obtained from children and parents.
3.1 PARTICIPANTS
To recruit participants with a range of motor abilities (Appendix A), children with and
without DCD were recruited from two main sources: public media and occupational
therapy practices in a large urban area in Canada. Acknowledging the critical
developmental period of motor control mechanisms guiding the performance of
visual-motor tasks (Ferrel-Chapus et al., 2002), children recruited were 8 years 11
months to 12 years 11 months of age.
As this was an exploratory study, no data were available to estimate the number of
children necessary to recruit to have adequate power for the statistical calculations.
However, after consulting published exploratory studies of children with DCD (Di
Brina et al., 2008; Rosenblum & Livneh-Zirinski, 2008; Smits-Engelsman et al.,
2001) and consulting a statistician, it was decided that the total number of children
targeted would be 20.
While the intent was to recruit children with a DCD diagnosis, it is widely
acknowledged in the literature that this group of children is largely undiagnosed
(Missiuna et al., 2008). In fact, and as described in section 2.2, a large number of
published studies on children with DCD actually recruit children with motor abilities
characteristic of DCD as opposed to children diagnosed with DCD. In such studies,
recommendations from the literature (e.g. European Consensus Conference, 2010;
Geuze et al., 2001) are adopted to ensure that the children recruited would be
39
diagnosed as DCD if a qualified health care professional was available to do so.
Children in this study were thus recruited either as children diagnosed with DCD or
as children with motor abilities characteristic of DCD. As recommended in the DCD
literature (European Consensus Conference, 2010; Geuze et al., 2001; Sugden,
2006), the four DSM IV-TR criteria for DCD diagnosis (APA, 2000) were then used
to describe the children recruited and discriminate children ‘typically developing’ with
regards to their motor abilities (TD), from children with motor abilities characteristic
of DCD2 (criteria summarized in Table 3-1).
Table 3-1 Criteria used to describe participants as DCD or TD
Criteria for identification of Developmental Coordination Disorder (DCD) (APA, 2000)
Criterion A Motor abilities substantially below expectations given chronological age
A score < 15th percentile on the Movement Assessment Battery for Children (Henderson & Sugden, 1992)
OR
A score < 5th percentile on one of the 3 domains assessed by the Movement Assessment Battery for Children
Criterion B Disturbance in Criterion A significantly interferes with academic achievement or activities of daily living
A score within the “suspect DCD” category on the Developmental Coordination Disorder Questionnaire (Wilson, 1998)
OR
Significant functional impact of coordination difficulties reported by parents or occupational therapist
Exclusion Criteria for the study
Criterion C Disturbance is not due to a medical condition or pervasive developmental disorder
Children with a known diagnosis such as cerebral palsy, spina bifida, pervasive developmental disorder, brain injury, tremors or any other neurological disorder
Criterion D Motor difficulties are in excess of those associated with intellectual delays
Children with a verbal IQ score < 70 on the Kaufman Brief Intelligence Test (Kaufman & Kaufman, 1990)
Additional Exclusion Criterion
Children with attention deficit hyperactivity disorder
2 These will be referred as children with DCD from now on to lighten the text.
40
The first two criteria of the DSM IV-TR are concerned with a child’s motor abilities.
Criterion A specifies that a child’s motor abilities must fall substantially below
expectations for chronological age. The Movement Assessment Battery for Children
(M-ABC) (Henderson & Sugden, 1992) was the normative measure used to quantify
a child’s motor abilities in comparison to his or her age group. The 2nd edition of the
M-ABC was not available when data collection started, and as such, the 1st edition
was used. The measure objectively assesses, identifies, and describes a child’s
motor abilities in three domains: fine motor skills, ball skills, and balance. It is the
standardized measure most often used to assess motor abilities in studies targeting
children with DCD (Geuze et al., 2001). Good psychometric properties are reported
(Henderson & Sugden, 1992). An overall M-ABC score that falls below the 15th
percentile is considered to indicate borderline motor abilities. A score that falls below
the 5th percentile is considered to indicate definitive motor difficulties. However, the
sensitivity of the M-ABC to the DCD diagnosis is unknown at present and some
have suggested that the M-ABC may under-identify children with DCD (e.g. Rodger
et al., 2007). Accordingly, and also because of the requirement of the additional
DSM IV-TR DCD diagnostic criteria, the 2010 European Consensus Conference on
DCD (Brussels) recommended that either of the following two options be used when
assessing DSM IV-TR criterion A for DCD diagnosis:
1) an overall M-ABC score below the 15th percentile, or
2) a score below the 5th percentile for at least one of the three domains
assessed.
Those recommendations were adopted in this study.
Criterion B specifies that the impact of a child’s motor difficulties measured in
criterion A must be evident in his or her daily life. In the DCD literature, source of
referral to the study, parent/professional reports, and/or standardized questionnaires
are typically used to assess this diagnostic criterion (Geuze et al., 2001; European
Consensus Conference, 2010). In line with other studies of children with DCD
(Geuze et al., 2001), the DCD Questionnaire [DCDQ] (Wilson, Dewey & Campbell,
1998) was used in this study to capture the impact of a child’s motor abilities on his
41
or her performance of everyday activities. The DCDQ is a research tool that was
designed to quantify parents’ perceptions of their child’s motor abilities, asking them
to consider their child’s performance on motor-based activities and compare it to that
of their peers. The DCDQ has demonstrated strong concurrent and construct validity
(Wilson, Kaplan, Crawford, Campbell, & Dewey, 2000). Overall scores on the DCDQ
can be used to classify a child as possibly (raw score < 57; < 25th percentile) or
definitely (raw score < 49; < 10th percentile) having DCD. During the course of this
study, a number of parents reported finding it difficult to judge how their child’s motor
abilities compared to that of their peers, even when they were aware of their child’s
difficulties in completing motor-based activities. Accordingly, either
parents/occupational therapists’ report or children’s score on the DCDQ were used
to identify the impact of a child’s motor abilities in his or her daily life. Such an
approach has been used in the literature as an effective way of assessing DSM IV-
TR criterion B (Zwicker et al., 2011) and meets the recommendations brought
forward by the European Consensus Conference in Brussels (2010).
The remaining two DSM IV-TR criteria for DCD diagnosis were used as exclusion
criteria for this study. Criterion C specifies that the motor difficulties identified in
criterion A should not be due to a medical condition or pervasive developmental
disorder. In line with other studies of children with DCD (Geuze et al., 2001), parent
report was used to ascertain that none of the children referred had a medical
condition, such as muscular dystrophy, cerebral palsy, a history of traumatic brain
injury, or other neurological symptoms that could potentially lead to motor difficulties.
Furthermore, given the nature of the experiment to be carried out and the focused
attention required of the children, children with attention deficit hyperactivity disorder
were also excluded from this study.
Criterion D specifies that if intellectual delays are present, the motor difficulties
identified in criterion A should be in excess of those associated with intellectual
delays. It is widely acknowledged in the literature that this criterion is difficult to
operationalize (European Consensus Conference, 2010; Geuze et al., 2001).
Ultimately, in research, the issue is typically avoided altogether by excluding children
presenting with measured Verbal Intelligence Quotient (IQ) scores falling below the
42
70th percentile (Geuze et al., 2001). Only verbal IQ scores are typically used due to
the potential confounding impact of the known performance difficulties of children
with DCD on the non-verbal intelligence subtest (Geuze et al., 2001). The same
approach was used in this study. The Kaufman Brief Intelligence Test (K-BIT)
(Kaufman & Kaufman, 1990) was administered to all participants to assess their
verbal IQ. The K-BIT is a standardized assessment used as a screening instrument
to obtain an estimate of the overall intelligence quotient, as well as verbal and
nonverbal intelligence. The authors report good validity and good correlation with the
Wechsler Intelligence Scale for Children—Third Edition (WISC-III) (Wechsler, 1991).
In the end, a total of 24 children were recruited: 12 children were classified as
children with DCD, and 12 children were classified as TD children, developing
typically with regards to their motor skills.
3.2 DESCRIPTIVE MEASURES
Two types of measures were used in this study: descriptive measures and
experimental measures. In addition to the descriptive measures used to identify
children with DCD in section 3.1, (i.e. the M-ABC, DCDQ, and K-BIT), the
Developmental Test of Visual-Motor Integration (VMI) 5th Ed (Beery & Buktenica,
2004) was also administered to children. The VMI is a standardized, norm-
referenced measure widely used by health care professionals working with children
with DCD (Miller et al., 2001; Rodger et al., 2003). It is administered to evaluate a
child’s visual-motor integration abilities. The VMI requires a child to copy a sequence
of forms arranged in order of increasing complexity. A total raw score that can be
converted to a norm-referenced standard score is obtained by adding the number of
correctly drawn shapes up to the ceiling level of three consecutive failures. The
authors report good psychometric properties with high overall reliability, moderate
concurrent validity and moderate to good construct and content validity (Beery &
Buktenica, 2004).
43
3.3 EXPERIMENTAL MEASURE
A computer-based visual-motor task was designed specifically for this study. To
complete the task, children were required to move a cursor (appearing as a 3 x 5
mm shuttle) to hit successively displayed targets (a 14 mm diameter planet)
appearing at random locations on a computer screen (Figure 3-1), equidistant from
each other (120 mm).
Figure 3-1 Computer game and novel cursor controller
To meet the general objective of this study and describe the impact of task difficulty
on visual-motor skill acquisition and task performance in children with and without
DCD, the task was designed to span a range of difficulty. The task was manipulated
so that initially it could be simple enough for all children to perform it skilfully, and so
that later it could be challenging enough to promote skill acquisition in skilled
children but possibly not in less skilled ones. Three levels of difficulty were targeted:
simple, intermediate, and complex.
To increase task difficulty gradually, structural and parametric features of the visual-
motor task were manipulated. As mentioned earlier, initial performance of a visual-
motor task depends on whether the structural and the parametric features of the
task have already been acquired and whether they are part of an existing visual-
motor transformation (Kagerer et al., 1997; Wolpert & Flanagan, 2010). When a
visual-motor transformation exists, then inverse and forward processes of motor
control combine and yield an accurate and efficient performance (Arbib et al., 2009;
44
Davidson & Wolpert, 2005; Frey, 2007). A visual-motor task for which the visual-
motor transformation should already exist would thus be of low difficulty. While the
computer mouse was a novel hand-held tool for many children at the time Missiuna
(1994) completed her study, it is now a very common tool used by many children to
access computers (Donker & Reitsma, 2007). It was proposed that children recruited
for this study would have had extensive experience using a computer mouse and
would thus have already acquired the structural and parametric features of the task.
This assumption was confirmed when children were asked about their computer
usage. All children in this study confirmed using a computer mouse (as opposed to a
touchpad) to access the computer. They also all confirmed using the computer
regularly, at least 2 times per week. Most children used the computer to send
emails, chat, do homework and play computer games. Accordingly, the simple task
was to land a shuttle on successively presented planets using a computer mouse to
control the shuttle.
To increase task difficulty to an intermediate level, evidence from two studies of
visual-motor task performance was used (Kagerer et al., 1997; Wolpert & Flanagan,
2010). It has been demonstrated that when the structural features of a task remain
constant but parametric features must be adapted, initial performance quickly
improves, followed by a gradual refinement of force controls and dynamics of the
movement (Kagerer et al., 1997; Wolpert & Flanagan, 2010). That is to say that the
general rules linking hand movements to their visual consequences already exist but
must be adapted to the new parameters of the task before skilful performance can
occur. In this study, it was proposed that using a cursor controller that kept the same
structural rules of movement of the mouse (i.e. away from the body = up on the
computer screen, and so on) but used different parameters would represent an
intermediate level of task difficulty. Accordingly, the intermediate task was to land
the shuttle on successively presented planets using a regular gaming joystick to
control the shuttle. All children in this study reported having used a gaming joystick
once or twice in the last year.
45
Finally, to increase task difficulty to a complex level, a task designed by Sailer and
colleagues (2005) where both structural and parametric features of the task had to
be acquired was used. Sailer’s task (2005) required adults to use a novel cursor
controller using forces and torques applied along the longitudinal axis of the
controller to reach successively appearing targets. As presented in section 2.3,
Sailer and colleagues (2005) demonstrated that adults’ initial task performance was
characterized by a period of exploration during which participants acquired the
structural features of the task, followed by a rapid increase in performance accuracy
and efficiency, and then by a more gradual improvement. As discussed earlier, when
a novel visual-motor task is performed and the structural features of the task must
be learned, no previously acquired visual-motor transformation can be used in motor
control processes (Sailer et al., 2005; Wolpert & Flanagan, 2010). As such, no initial
improvement in performance is typically reported until the rules governing the visual-
motor transformation become better defined and task structure is acquired. Then,
parametric features can be acquired and used to refine force controls and
movement dynamics. For this study, it was proposed that using a novel cursor
control that did not rely on previously learned movement structure or task
parameters would present a high level of difficulty. Accordingly, the complex task
was to land the shuttle on successively presented planets using a novel controller
similar to the one used by Sailer and colleagues (2005). None of the children in this
study had prior experience using the novel controller.
3.3.1 APPARATUS
Three controllers were used to target the levels of difficulty of the 1) simple, 2)
intermediate, and 3) complex tasks.
1) Simple task: Computer mouse (Microsoft) – A computer mouse was used to
measure the performance of a well-learned task. Default computer settings were
used. To land the shuttle, children pressed the left button, as typically set on
computers.
2) Intermediate task: Gaming joystick (Logitech) – A joystick was used to measure
the performance of a task structurally similar to the mouse, but presenting
46
different parametric features. The directionality of the visual-motor transformation
required to control the shuttle was the same as the mouse. To land the shuttle,
children pressed the button located at the index finger, as typically set in
computer games.
3) Complex task: Novel controller - A novel controller was designed to resemble the
one used by Sailer et al. (2005). It consisted of a rectangular box to which two
handles were attached (Figure 2-2). The controller used a force/torque
transducer attached to one of the handles to move the cursor. Push/pull forces
applied along the horizontal plane (forward/backward) moved the cursor left and
right along the Y axis on the computer screen. Torque forces applied around the
longitudinal axis moved the cursor up and down. Moving the cursor diagonally
thus required a combination of push/pull and torque forces applied
simultaneously. To land the shuttle, children pressed a button added where the
2nd, 3rd, 4th, and 5th fingers rested. The other handle was inactive and was used
to stabilize the controller. The children were allowed to use which ever hand
they chose as the active hand, the other hand being their stabilizing hand.
3.4 PROCEDURE
During an initial phone interview, participants received further information about the
study and preliminary questions were asked of parents to ensure that children
meeting the exclusion criteria did not continue further with the study. Once informed
consent from a parent and assent from the child were obtained, 3 sessions were
scheduled in the following order: 1) measure of visual motor integration (VMI), 2)
experimental visual-motor task, and 3) descriptive measures aimed at classifying
children with and without DCD, (i.e. measures of motor ability (M-ABC & DCDQ) and
exclusion measure (K-BIT)). By administering the measures of motor ability last,
everyone involved in this study was kept blind to the motor ability of the children.
The principal investigator administered the experimental visual-motor task while the
remaining measures of visual-motor integration, of verbal intelligence, and of motor
ability were administered by either the principal investigator or trained 2nd year
occupational therapy students. The three sessions could be scheduled back to back
47
or spread over 2 weeks. Most participants chose to schedule the sessions back to
back, with an adequate amount of rest provided in-between to prevent fatigue. The 3
sessions were usually completed within 2 hours.
Testing was conducted at the University of Toronto, Baycrest Geriatric Centre,
and/or the child’s school or home. Regardless of the setting selected, children were
comfortably seated in an environment without distractions.
3.4.1 EXPERIMENTAL VISUAL-MOTOR TASK
Children were seated approximately 40 cm away from the computer screen with
their eye-level corresponding to the middle of the computer screen. Children were
instructed to move the shuttle to the planet as fast and as straight as they could, and
to land the shuttle using a button press. Auditory signals were emitted when the
shuttle successfully or unsuccessfully landed on the planet to provide feedback to
children. A shuttle successfully landed on a planet when at least 80% of the shuttle
overlapped the planet on button press. Following a successful landing, the planet
disappeared and the next one appeared at a fixed distance of 120 mm from the
previous planet.
As mentioned above, task difficulty was modified by changing the controller used to
move the shuttle on the computer screen. All three controllers were utilized by all
children, in fixed order: the computer mouse, the gaming joystick, and the novel
controller. All children used their preferred hand to manipulate the controllers. Vision
of the controller and hands was permitted to preserve the ecological validity of the
task and to allow children to use vision of their limb in action to update the visual-
motor rules linking their manual actions to the visual consequences of their
movements.
For each controller, children were presented an initial block of 10 trials, followed by 2
blocks of 20 trials. Not all children completed the same number of trials. During initial
data collection on the first 10 children, only 20 trials were presented for the mouse,
30 for the joystick, and 40 for the novel controller in an attempt to prevent fatigue.
After initial inspection of the data, it was decided to attempt to increase the number
of trials to 50 for each controller to ensure that visual-motor task performance
48
levelled-off during the session for most children. However, with 50 trials, fatigue and
frustration became an issue for struggling children and not all of them were able to
complete the 50 trials for each controller. Furthermore, technical difficulties with the
computer program and/or novel controller sometimes led to the loss of data during
data collection. Given the intent of looking at skill acquisition by exploring changes in
task performance over repeated trials, missing or lost trials were treated as missing
data and were not repeated (see Appendix B for documentation of number of trials
completed per child per controller).
3.5 VARIABLES & DATA HANDLING
Visual-motor task performance, the main behaviour under study, was captured by
two dependent variables: movement time (MT) and distance travelled by the cursor
(DIST). MT is operationalized as the difference between movement onset and target
hit. DIST is operationalized as the distance travelled by the cursor. MT and DIST are
related since the longer the distance travelled, the more time it is likely to take.
However, the relationship correlations may not be perfect as one could increase or
decrease movement time without increasing distance travelled (e.g. if a child goes
slowly to ensure that he or she follows a straight path). As such, both variables were
included to characterize visual-motor task performance but were considered in turn
in the data analysis.
To measure and describe performance of the visual-motor task when it was first
performed and reduce the influence of skill acquisition, and thus changes in visual-
motor performance, potentially occurring over repeated trials, initial task
performance was also calculated by computing MTi and DISTi. Because no studies
specifically looking at initial task performance in children could be found, no
guidance was available from the literature to operationalize these variables.
Nevertheless, findings from a study by Jansen-Osmann and colleagues (2010)
demonstrated that, during exposure to new visual-motor parameters, a younger
group of 6 year-olds acquired the new task parameters within five trials. Accordingly,
it was decided that MTi and DISTi would be captured by the average of MT and
DIST, respectively, over the first five trials for each of the three controllers. Visual
49
inspection of the data confirmed that averaging task performance over the first five
trials would accurately describe initial visual-motor performance.
As introduced in the study objectives, the main constructs of interest and
independent variables being investigated for their modulating effect on visual-motor
performance in this study were task difficulty and motor ability. Furthermore, since
one of the objectives aimed to describe skill acquisition over repeated trials, trial
number was also of interest. Finally, because this study was conducted with
children, the possible effect of age on task performance was also considered (Table
3-2).
Table 3-2 Independent variables investigated for their effect on visual-motor task performance
Independent Variable
Abbreviation Description
Task Difficulty DIFF Categorical variable representing the controller used to complete the visual-motor task. Three categories were included: simple (mouse), intermediate (joystick), and complex (novel).
Motor Ability M-ABC Independent variable captured by a child’s overall score on the M-ABC.
Trial TRIAL Independent variable captured by the trial number.
Age AGE Independent variable captured by a child’s age and expressed in months
3.6 DATA ANALYSES
All statistical calculations were performed using the computer software SAS 9.2
(SAS Institute, 2008). The level of significance was set at α=0.05.
As a preamble to data analysis, participants’ characteristics and results on the
descriptive measures administered were first summarized using descriptive
statistics. Then, the data analysis was constructed so as to proceed in logical steps,
according to the research objectives of this study.
The analysis strategy chosen to investigate the modulating effect of independent
variables on visual-motor skill acquisition and task performance was mixed-effects
modelling: a linear individual growth model, as well as a variance-covariance model
to take into account the repeated nature of the data. More and more, statisticians
have come to use mixed-effects models to analyze multilevel longitudinal data
50
because of the many advantages they offer over traditional data analyses (Moser,
2004; Raudenbush, 1993; Singer, 1998).
Mixed-effects models include both fixed effects and random effects. As such,
they generalize the classical linear regression model (fixed effects) by also
including terms that describe the correlation structure of the residuals (random
effects).
In addition, mixed-models are more flexible as they can incorporate continuous
and classification predictors that may be fixed or may vary with time, with a
potentially different design matrix for each subject.
Mixed-effects models accommodate repeated measures data by taking into
account and modelling the covariation between observations on the same
participant at different times.
Finally, and of great importance in this study, mixed-effect models are fitted by
maximum likelihood, and as such can handle missing observations and unbalanced
designs more efficiently, leading to more reliable conclusions.
Mixed-effects models are built progressively as independent variables are tested for
their effect on the dependent variables (Singer, 1998). When adding an independent
variable to a model, covariance parameter estimates and goodness of fit statistics
are examined along with solutions for fixed effects to ascertain whether a newly
added independent variable contributes to explaining variability of the dependent
variables.
With the intent of adopting a systematic approach to data analysis, and since motor
ability (M-ABC) was an independent variable that was believed likely to influence
visual-motor task performance (DIST and MT), only the data from TD children (n =
12) were analyzed at first to gain a better understanding of the modulating effect of
the other independent variables on DIST and MT. Given the objectives and design
of this study, results from these analyses cannot be considered normative data
representing ‘typical’ performance in children. Nevertheless, results from these
analyses were informative and helpful in meeting the objectives of this study and
describing the effect of motor ability on task performance and skill acquisition.
51
Accordingly, for each objective, the analyses performed on the data from TD
children (n = 12) are discussed first, followed by the analyses performed on the data
from all children (N = 24).
3.6.1 DESCRIPTION OF PARTICIPANTS
For descriptive purposes, group differences were assessed using Independent t-
tests for Age, KBIT and VMI. Acknowledging the expected skewed distribution of M-
ABC and DCDQ scores, Mann-Whitney U tests were used to test for group
differences. Significant group differences were visually presented using box plots
and group distributions (Appendix C).
3.6.2 VISUAL-MOTOR TASK PERFORMANCE: MODULATING VARIABLES
The first objective of this study was to describe the modulating effect of task difficulty
and motor ability on visual-motor task performance in children with and without
DCD. Two analyses were performed to meet this objective. First, the modulating
effect of task difficulty on visual-motor task performance was investigated in TD
children. Then, the modulating effect of motor ability on the effect of task difficulty on
visual-motor performance was investigated in all children. To reduce the possible
influence of skill acquisition over repeated trials, MT i and DISTi were used as
dependent variables.
(i) Investigating the effect of task difficulty on initial visual-motor task
performance.
The first analysis was performed on the data from TD children (n = 12). Mixed-
effects modelling using the PROC MIXED procedure was selected to explore if MTi
and DISTi were modulated by DIFF. DIFF was the predictor variable modeled in the
fixed-effect model. MTi and DISTi were the dependent variables. To increase the
linearity of the data and facilitate convergence of the models, here and for the other
analyses, a logarithmic transformation was applied to the dependent variables prior
to formal analysis (Draper & Smith, 1998).
A RANDOM effect statement was added to allow the intercept and slope of the
model to vary across children over DIFF (Singer, 1998). To take into account the
52
repeated nature of the data collected over DIFF on the same child, the structure of
the within-participant error covariance matrix was specified. Different error structures
were investigated to determine which one best fits the data using covariance
parameter estimates and goodness of fit statistics provided in the SAS output
(Singer, 1998). The first-order auto-regressive error structure was selected.
Solutions for fixed effects and Type 3 tests of fixed effects were used to test for the
null hypothesis modeled. The Type 3 test for DIFF was examined to determine
whether the overall curve accounting for possible heterogeneous slopes was
significantly different from zero. Finally, solutions for fixed effects for the three
parameter estimates of DIFF and corresponding p values were examined to test for
evidence supporting the null hypothesis that the slopes for each of the level of DIFF
are equal to zero. The standard error related to the parameter estimates are
reported in Appendix D.
Taking into account the potential confounding effect of age on visual-motor
performance, AGE was added as a predictor variable to explore whether it should
also be included in the model. Covariance parameter estimates and goodness of fit
statistics provided in the SAS output were examined when AGE was added to the
model. Solutions for fixed effects revealed that AGE was not a statistically significant
predictor in the model and covariance parameter estimates and goodness of fit
statistics confirmed that AGE did not contribute to explaining variability in the data
when added to the model; it was thus excluded.
(ii) Investigating the effect of motor ability on initial visual-motor task
performance when the influence of task difficulty is also considered.
The second analysis was performed on data from all children (N = 24). Mixed-effect
modelling using the PROC MIXED procedure was selected to explore if MT i and
DISTi were modulated by M-ABC, and to determine if that effect was modulated by
DIFF. As such, DIFF, M-ABC, and the interaction M-ABC*DIFF were the predictor
variables modeled in the fixed-effect model. MTi and DISTi were the dependent
variables. As before, a RANDOM effect statement was added to allow the intercept
and slope of the model to vary across children over DIFF (Singer, 1998). The first-
53
order auto-regressive error structure was selected. Finally, similarly to above, AGE
was examined for its significance as a predictor and, again, it was determined that
AGE would not be included in the model.
Solutions for fixed effects and Type 3 tests of fixed effects were used to test for the
null hypotheses modeled. First, the Type 3 test for the higher-order term M-
ABC*DIFF was examined to determine if the null hypothesis of equality of slopes for
the three levels of task difficulty could be rejected. Then, the Type 3 test for M-ABC
was examined to determine whether the overall curve accounting for possible
heterogeneous slopes was significantly different from zero. Finally, solutions for
fixed effects for the three parameter estimates M-ABC*DIFF and corresponding p
values were examined to explore the results of the Type 3 test for evidence
supporting the null hypothesis that the slopes for each of the level of DIFF is equal
to zero. The standard error related to the parameter estimates are reported in
Appendix D.
3.6.3 VISUAL-MOTOR SKILL ACQUISITION: MODULATING VARIABLES
The second objective of this study was to describe visual-motor skill acquisition in
children with and without DCD. Since visual-motor skill acquisition is characterized
by changes in visual-motor task performance, the modulating effect of repeated
trials on task performance was explored. Two analyses were performed. First, the
modulating effect of repeated trials on visual-motor task performance was
investigated in TD children when the influence of task difficulty was also considered.
Then, the modulating effect of motor ability on skill acquisition in all children was
investigated when task difficulty and repeated trials were taken into account.
(i) Investigating the effect of repeated trials on visual-motor task performance
when the influence of task difficulty is also considered.
The first analysis was performed on the data from TD children (n = 12). The PROC
MIXED procedure in SAS was selected to investigate if MT and DIST are modulated
by TRIAL, and to determine if that effect is modulated by DIFF. As such, DIFF,
TRIAL, and the interaction DIFF*TRIAL were the predictor variables modeled in the
fixed-effect model. MT and DIST were the dependent variables.
54
A RANDOM effect statement was added to allow the intercept and slope of the
model to vary across participants over trials (Singer, 1998). To take into account the
repeated nature of the data collected over trials on the same participants, the
structure of the within-participant error covariance matrix was specified. Different
error structures were investigated to determine which one best fits the data using
covariance parameter estimates and goodness of fit statistics provided in the SAS
output (Singer, 1998). The first-order auto-regressive error structure was selected.
Again, taking into account the potential confounding effect of age on visual-motor
performance, AGE was added as a predictor variable to explore whether it should
also be included in the model. Solutions for fixed effects revealed that AGE was not
a statistically significant predictor in the model and covariance parameter estimates
and goodness of fit statistics confirmed that AGE did not contribute to explaining
variability in the data when added to the model; it was thus excluded.
Solutions for fixed effects and Type 3 tests of fixed effects were used to test for the
null hypotheses modeled. First, the Type 3 test for the higher-order term
DIFF*TRIAL was examined to determine if the null hypothesis of equality of slopes
for the three levels of task difficulty could be rejected. Then, the Type 3 test for
TRIAL was examined to determine whether the overall growth curve accounting for
possible heterogeneous slopes was significantly different from zero. Finally,
solutions for fixed effects for the three parameter estimates DIFF*TRIAL and
corresponding p values were examined to explore the results of the Type 3 test for
evidence supporting the null hypothesis that the slopes for each of the level of DIFF
is equal to zero. The standard error related to the parameter estimates are reported
in Appendix D.
(ii) Investigating the effect of motor ability on visual-motor task performance.
when the influence of repeated trials and task difficulty are also considered
The second analysis was performed on data from all children (N = 24). Mixed-effect
modelling using the PROC MIXED procedure was selected to explore if MT and
DIST are modulated by M-ABC and TRIAL, and to determine if that affect is
modulated by DIFF. Recognizing the limit of the sample size, each level of DIFF was
55
modeled separately to reduce the degrees of freedom involved in the analysis. As
such, for each level of DIFF, the predictor variables modeled in the fixed-effect
model were M-ABC and TRIAL. MT and DIST were the dependent variables. As
before, a RANDOM effect statement was added to allow the intercept and slope of
the model to vary across participants over trials (Singer, 1998). The first-order auto-
regressive error structure was selected.
Here again, AGE was added as a predictor variable to explore whether it should also
be included in the model. Covariance parameter estimates and goodness of fit
statistics provided in the SAS output were examined when AGE was added to the
model. Solutions for fixed effects revealed that AGE was not a statistically significant
predictor in the model for the simple and intermediate visual-motor task, but was a
significant predictor for the complex task. Furthermore, covariance parameter
estimates and goodness of fit statistics confirmed that although AGE did not
contribute to explaining variability in the data when added to the model for the
simple and intermediate visual-motor task, it did when it was added to the model for
the complex task. Accordingly, AGE was added to the model for all three tasks, to
ensure that the variability present in the data because of AGE when all children were
considered would be taken into account for all three tasks, thus allowing comparison
of results across the three levels of DIFF.
Solutions for fixed effects and Type 3 tests of fixed effects were used to test for the
null hypotheses modeled. First, the Type 3 test for the higher-order term M-
ABC*DIFF was examined to determine if the null hypothesis of equality of slopes for
the three levels of task difficulty could be rejected. Then, the Type 3 test for M-ABC
was examined to determine whether the overall curve accounting for possible
heterogeneous slopes was significantly different from zero. Finally, solutions for
fixed effects for the three parameter estimates M-ABC*DIFF and corresponding p
values were examined to explore the results of the Type 3 test for evidence
supporting the null hypothesis that the slopes for each of the level of DIFF is equal
to zero. The standard error related to the parameter estimates are reported in
Appendix D.
56
3.6.4 VISUAL-MOTOR SKILL ACQUISITION: PATTERNS OF CHANGE
The third objective of this study was to further describe skill acquisition in children
with and without DCD. Two analyses were thus performed to describe patterns of
change in visual-motor task performance over repeated trials in children. First, for
tasks where skill acquisition was evident across trials (as determined in 3.6.3),
visual-motor skill acquisition rates were estimated using non-linear regression
analyses. Non-linear regression analyses are frequently used in studies that
investigate phenomenon that are not expected to follow a linear trend, such as in
skill acquisition (Cantin, 2004; Chang & Yu, 2010; Martin, Keating, Goodkin, Bastian,
& Thach, 1996). Then, patterns of change in visual-motor task performance across
trials were described by performing visual graph analyses.
To echo the analyses performed in section 3.6.2 and 3.6.3, the proposed analyses
were first done on the data from TD children only (n = 12), followed by analyses on
the data from all children (N = 24). In the same way as the mixed-effects modelling
performed above, this systematic approach provided an appreciation of how
patterns of visual-motor task performance changed over repeated trials when
children with DCD were added to the analysis. For comparative purposes, the
analyses were also performed on the data from children with DCD and are
presented visually in Appendix E.
(i) Investigating visual-motor skill acquisition rates.
Before a non-linear regression analysis could be done, outliers were removed from
the data set to eliminate data points that did not follow the trend from the points
around them. While the choice to remove outliers can sometimes be controversial
(Motulsky & Brown, 2006), it becomes important when attempting to fit data to
models, such as the non-linear regression model used here, as one outlier can
greatly influence the shape of a regression line. While outlier elimination is often
done in an ad hoc manner (Motulsky & Brown, 2006), the fourth spread method was
used here (Breckenridge, Tallia, & Like, 1988) to objectively set upper and lower
boundaries to guide the process. The fourth spread (or quartile) method identifies
the boundaries of each of the quartiles in the data set, measures the distance
57
between the lower and upper quartiles (or fourth spread (fs)), and sets the upper
and lower boundaries equal to median + 1.5*fs. However, given that an outlier is
defined to be an observation that deviates enough from surrounding observations so
as to raise suspicions that it might be generated by a different mechanism, it is also
important to consider the particular features of a data set when removing outliers
(Hawkins, 1980). For example, in this data set, for the complex task, visual-motor
performance was expected to change over trial and variability was expected to
characterize initial visual-motor performance. Visual analysis of the data suggested
that, to be conservative, the upper boundary should not be applied to the first 15
trials of the novel task to allow expected variability in performance to emerge.
Once outliers were removed, DIST and MT for each trial were averaged across
children and plotted. A non-linear regression analysis was conducted to estimate the
visual-motor skill acquisition rate of the resulting curve. An exponential decay curve
was fitted to the averaged group data (Y = a- b * e-x/c), where Y is the dependent
variable DIST or MT, x is the trial number, a represents the minimum value that the
function approaches, b is the magnitude of the acquisition required from the first
trials to the value a, and c is the acquisition rate (Cantin et al., 2004; Chang & Yu,
2010; Martin et al., 1996). Thus, the number of trials taken to reach a point of (1-e-1)
or 63.2% of the way through acquisition is represented by c. The curve’s coefficients
were obtained by conducting a non-linear regression analysis in SAS. Predicted
curves were plotted along with actual MT and DIST across children for each task
where visual-motor acquisition was evident (as determined in section 3.6.2). To
allow visual comparison of patterns of change when considering TD children only
and all children, predicted curves for the TD children (n = 12) were plotted along
those of all children (N = 24).
(ii) Qualifying patterns of change in visual-motor performance over repeated
trials.
Then, to qualify patterns of change in visual-motor performance over repeated trials,
visual graph analyses were conducted. For this purpose, principles from statistical
process control using an average moving range (XmR) control chart (Callahan &
58
Barisa, 2005) were applied to this data set to offer a systematic strategy to this
visual graph analysis. Similar to the approach used to identify outliers above, upper
and lower boundaries were calculated for the MT and DIST XmR control chart.
Three upper and three lower boundaries were calculated, creating a confidence
interval around mean MT or DIST (methods described in Appendix F). Then, the
graph was visually analyzed to identify special causes, which in control charts can
be: one point falling outside the upper or lower control limit, seven or more
consecutive points falling above or below the centre mean line (called a run), or
seven or more consecutive points moving up or down, bisecting the centre mean
line (called a trend) (Callahan & Barisa, 2005). In addition to special causes, a
description of the curve and data points in relation to the various boundary lines of
the control chart was performed. The boundaries determined for TD children were
used to visually analyze the graphs of the complete sample and identify special
causes. TD curves were again plotted along those of the entire sample to allow
visual comparison of patterns of change.
59
CHAPTER 4
RESULTS
Legend for all figures To differentiate groups:
Shades of black: All children (N = 24) MT Dist
Shades of blue: TD children (n=12) MT Dist
Shades of red: DCD children (n=12) MT Dist To identify visual-motor tasks
Simple task
Intermediate task
Complex task
60
4 RESULTS
4.1 DESCRIPTION OF PARTICIPANTS
As summarized in Table 4-1, 24 participants (10 girls and 14 boys) completed
the study. The median age of the sample was 10 years 9 months, ranging from 8
years 11 months to 12 years 11 months. In total, 12 children were classified as
TD (7 girls and 5 boys), and 12 children were classified as DCD (3 girls and 9
boys). When compared on the various standardized assessments administered,
both groups differed on the motor measures (i.e. M-ABC and DCDQ). In the
group of children with DCD, nine children scored at or below the fifth percentile
on the M-ABC. The remaining three children scored at the sixth, eighth and 11th
percentile, respectively. The two groups differed on the VMI, although the
average standard score of both groups fell within the established norms and only
one child with DCD scored below the 15th percentile on the measure. No
significant age differences between the two groups were found.
Table 4-1 Description of Participants (N = 24)
Measure
Entire Sample (N = 24)
TD (n = 12)
DCD (n = 12)
Mean (s.d.)
Median (range)
Mean (s.d.)
Median (range)
Mean (s.d.)
Median (range)
Age (months) 129 (14.5)
128.5 (107-155)
133 (17.5)
137 (107-155)
128 (12.5)
127 (111-147)
*VMI standard
score
100.4 (12.0)
98.5 (77-131)
105.2 (13.3)
104 (88-131)
95.6 (8.6)
98 (77-107)
*M-ABC overall
score
11.4 (7.9)
10.5 (0-29.5)
5.0 (3.4)
4.25 (0-10)
17.8 (5.4)
17.25 (11-29.5)
*DCDQ overall
score
65.8 (15.9)
72.5 (21-85)
76.7 (7.4)
78.5 (59-85)
54.7 (14.7)
56.5 (21-75)
KBIT standard score
112.1 (16.7)
113 (76-137)
112.0 (13.7)
114 (86-128)
112.2 (19.9)
113 (76-137)
Abbreviations: s.d.: standard deviation; *: statistically significant group difference at α = .05;
VMI: Developmental Test of Visual-Motor Integration; M-ABC: Movement Assessment Battery for Children
DCDQ: Developmental Coordination Disorder Questionnaire; K-BIT: Kaufman Brief Intelligence Test
61
4.2 VISUAL-MOTOR TASK PERFORMANCE: MODULATING VARIABLES
The first objective of this study was to describe visual-motor task performance in
children with and without DCD. Two analyses were performed to meet this
objective and investigate the modulating effect of two independent variables,
task difficulty (DIFF) and motor ability (M-ABC), on initial visual-motor
performance (MTi and DISTi).
4.2.1 THE EFFECT OF TASK DIFFICULTY ON INITIAL VISUAL-MOTOR TASK
PERFORMANCE IN TD CHILDREN
As summarized in Table 4-2, Type 3 tests of fixed effects for MTi and DISTi
detected a statistically significant main effect for DIFF (p < .0001 for both
variables). In other words, there is evidence to suggest that DIFF modulates
initial visual-motor performance in TD children.
For MTi, solutions for fixed effects found a statistically significant main effect for
DIFF for the intermediate and complex tasks (p < .0001 for both), but did not
detect a statistically significant main effect for the simple task (p = .114). For
DISTi, solutions for fixed effects found a statistically significant main effect for
DIFF for all three tasks (p < .0001). In other words, in TD children, visual-motor
performance can be explained, in part, by the difficulty of the task, except for MTi
during the simple task.
The effect of task difficulty on visual-motor performance is illustrated in Figure 4-
1; Note that mean MTi was more than 7 times larger for the complex task than
for the intermediate task, and mean DISTi was almost 4 times larger.
Table 4-2 Investigating the effect of task difficulty on initial visual-motor task performance TD (n =
12)
Analysis Ln MTi (TD) Ln DISTi (TD)
Test value; df (33) p value Test value; df (33) p value
Effect of DIFF F = 150.89 <.0001* F = 6173.16 <.0001*
DIFF (simple) t = 1.62 .1143 t = 96.69 <.0001*
DIFF (intermediate) t = 7.45 <.0001* t = 41.10 <.0001*
DIFF (complex) t = 20.89 <.0001* t = 34.24 <.0001*
Abbreviations: *: statistically significant at α = .05; MTi: Initial movement time; DISTi: Initial distance
travelled; DIFF: Task difficulty; Ln: natural log transformation applied to the data
62
Figure 4-1 Mean initial movement times and distances travelled for each level of difficulty to
illustrate the modulating effect of task difficulty on initial visual-motor task performance for TD
(n=12).
Legend Shades of blue: TD children (n = 12) MTi DISTi MTi: Initial movement time (s); DISTi: Initial distance travelled (mm)
0
5
10
15
20
25
30 M
Ti (
s)
Task Difficulty
Effect of Task Difficulty on Initial Movement Time in TD Children
Simple Intermediate Complex
0
200
400
600
800
1000
1200
DIS
Ti (
mm
)
Task Difficulty
Effect of Task Difficulty on Initial Distance Travelled for TD Children
Simple Intermediate Complex
1.35 (0.58)
3.42 (2.49)
26.13 (16.41)
150.02 (29.82)
254.39 (168.79)
1001.99 (745.48)
63
4.2.2 THE EFFECT OF MOTOR ABILITY ON INITIAL VISUAL-MOTOR
PERFORMANCE WHEN TASK DIFFICULTY IS CONSIDERED
Type 3 tests of fixed effects detected a statistically significant main effect for
DIFF (p < .001 for both MTi and DISTi) and M-ABC (p = .032 and p = .049, for
MTi and DISTi, respectively). A statistically significant interaction M-ABC*DIFF
for MTi and DISTi (p = .025 and p = .008, respectively) was also detected
Further exploration of the interaction M-ABC*DIFF is summarized in Table 4-3.
No statistically significant interactions for MTi and DISTi for the simple task (p =
.758 and p = .683, respectively) and the intermediate task (p = .551 and p =
.958) tasks were noted. However, a statistically significant interaction was
detected for the complex task (p = .001 and p < .001, respectively). In other
words, motor ability was found to modulate the effect of increased difficulty on
visual-motor performance only for the complex task. To illustrate the analysis
performed to understand the interaction between M-ABC and DIFF, M-ABC
scores are graphed along with corresponding movement times and distance
travelled in Figure 4-2. Furthermore, to echo Figure 4-1 illustrating the effect of
task difficulty on visual-motor performance in TD children, the same graph was
drawn again in Figure 4-3 but with each child’s individual curve drawn and colour
coded based on their motor abilities.
Table 4-3 Investigating the effect of motor ability on initial visual-motor performance when task
difficulty is considered in all children (N = 24).
Analysis Ln (MTi) Ln (Disti)
Test value; df (42) p value Test value p value
Effect of M-ABC*DIFF F = 4.03 .013* F = 4.51 .008*
M-ABC*DIFF (simple) t = 0.31 .758 t = -.19 .683
M-ABC*DIFF (intermediate) t = 0.60 .551 t = -1.04 .958
M-ABC*DIFF (complex) t = 3.46 .001* t = -9.12 < .001*
Abbreviations: MTi: Initial movement time; DISTi: Initial distance travelled; DIFF: Task difficulty; M-ABC:
Motor ability; Ln: natural log transformation applied to the data.
*: statistically significant at α = .05
64
0.00
0.75
1.50
2.25
3.00
1 5 9 13 17 21 25
Ln
(M
Ti)
MABC Scores
MTi Intermediate Task
3.00
3.75
4.50
5.25
6.00
6.75
7.50
1 5 9 13 17 21 25
Ln
(D
IST
i)
MABC Scores
DISTi IntermediateTask
-0.40
0.00
0.40
0.80
1.20
1 5 9 13 17 21 25
Ln
(M
Ti)
MABC Scores
MTi Simple Task
4.00
4.40
4.80
5.20
5.60
6.00
1 5 9 13 17 21 25
Ln
(D
IST
i)
MABC Scores
DISTi Simple Task
0.00
1.00
2.00
3.00
4.00
5.00
6.00
1 5 9 13 17 21 25
Ln
(M
Ti)
MABC Scores
MTi Complex Task
4.00
5.00
6.00
7.00
8.00
9.00
1 5 9 13 17 21 25
Ln
(D
IST
i)
MABC Scores
DISTi Complex Task
Figure 4-2 Illustrating the modulating effect of motor ability on initial visual-motor performance for the
simple, intermediate and complex tasks in all children (N = 24). Comparing the graphs for each task, it
becomes clear that M-ABC only modulates; Abbreviations: MTi: Initial movement time; DISTi: Initial
distance travelled; M-ABC: Motor ability (higher scores = DCD)
Legend
Simple Task *
Intermediate Task Complex Task MT: movement time DIST: distance travelled
65
Abbreviations: MTi: Initial movement time; DISTi: Initial distance travelled;
Legend
Shades of blue: TD children (n = 12) MTi DISTi
Shades of red: children with DCD (n = 12) MTi DISTi
-
1,000
2,000
3,000
4,000
5,000
6,000
DIS
Ti (m
m)
Task Difficulty
Effect of Motor Ability on Initial Distance Travelled (N = 24)
Simple Intermediate Complex
-
20
40
60
80
100
120
MT
i (s)
Task Difficulty
Effect of Motor Ability on Initial Movement Time (N = 24)
Simple Intermediate Complex
Figure 4-3 Illustrating the effect of motor ability on initial visual-motor task
performance in all children (N = 24) when task difficulty is considered.
66
4.3 VISUAL-MOTOR SKILL ACQUISITION: MODULATING VARIABLES
The second objective of this study was to describe visual-motor skill acquisition in
children with and without DCD. Two analyses were performed to meet this objective.
First, the modulating effect of repeated trials on task performance in TD children was
investigated when task difficulty was also taken into account. Then, the modulating
effect of motor ability on skill acquisition in all children was investigated when task
difficulty and repeated trials were taken into account.
4.3.1 THE EFFECT OF REPEATED TRIALS ON VISUAL-MOTOR TASK
PERFORMANCE IN TD CHILDREN WHEN TASK DIFFICULTY IS ALSO
CONSIDERED
Type 3 tests of fixed effects for MT and DIST detected a statistically significant
interaction TRIAL*DIFF (both p < .001) and significant main effect for TRIAL and
DIFF (both p < .001 for both variables). In other words, the impact of TRIAL on
visual-motor performance was different for all three levels of DIFF (Table 4-4).
Furthermore, when the interaction term TRIAL*DIFF and TRIAL were added to the
model, both variables still had an overall modulating effect on visual-motor
performance.
Solutions for fixed effects for TRIAL*DIFF (simple, intermediate and complex) did
not detect statistically significant interactions for MT and DIST for the simple (p =
.793 and p = .852, respectively) and the intermediate (p = .09 and p = .300) tasks.
However, a statistically significant interaction was detected for the complex task (p <
.001 for both). This indicates that for TD children, MT and DIST were stable over
repeated trials for the simple and intermediate tasks, but changed during the
complex task (Figure 4-4). For TD children, the increased difficulty of the complex
task did lead to skill acquisition and a decrease in MT and DIST.
67
Table 4-4 Investigating the effect of task difficulty on visual-motor task performance over repeated trials
in TD Children (n = 12).
Analysis
Ln MT (TD) Ln DIST (TD)
Test value df (1020)
p value Test value df (1020)
p value
Effect of TRIAL*DIFF
F = 54.29 <.001*
F = 50.23
<.001*
TRIAL*DIFF (simple) t = 0.26 .793 t = -.19 .852
TRIAL*DIFF (intermediate) t = -1.70 .090 t = -1.04 .300
TRIAL*DIFF (complex) t = -9.87 .001* t = -9.12 .001*
Abbreviations: MT: movement time; DIST: distance travelled; DIFF: Task difficulty; TRIAL: Trial number; Ln:
natural log transformation applied to the data.
*: statistically significant at α = .05;
Legend
Simple Task *
Intermediate Task Complex Task MT: movement time DIST: distance travelled
4.5
4.8
5.1
5.4
5.7
6.0
6.3
1 50
Ln
(D
IST
)
Trial Number
Effect of Task Difficulty on Distance Travelled over Repeated Trials in TD children
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1 50
Ln
(M
T)
Trial Number
Effect of Task Difficulty on Movement Time over Repeated Trials in TD children
Figure 4-4 Illustrating the effect of motor ability on initial visual-motor task
performance in TD children (n = 12) when task difficulty is considered.
68
4.3.2 THE EFFECT OF MOTOR ABILITY ON VISUAL-MOTOR TASK PERFORMANCE
WHEN REPEATED TRIALS AND TASK DIFFICULTY ARE ALSO CONSIDERED
For this analysis, three sub-analyses for each DIFF were conducted to reduce the
number of degree of freedoms that would have otherwise been involved. Since AGE
was not a statistically significant predictor when only TD children were considered,
but a main effect was found for the complex task when all children were considered
(as mentioned in section 3.6.3), post hoc analyses of the variable AGE were
performed to explore if its influence on visual-motor performance was due to an
interaction with the variable M-ABC already included in the model, or if it was simply
jointly related to visual-motor performance when M-ABC was also considered. In
other words, even though no statistically significant differences in AGE was detected
between children with and without DCD earlier (Table 4-1), it may still be a
confounder that needs to be included in the model. No correlation between AGE
and-M-ABC was found (p > .05), thus both parameters were left in the model for all
three tasks.
Considering the simple task, for MT, a statistically significant main effect was
detected for TRIAL (p = .003) but not for M-ABC (p = .069) or AGE (p = .638). For
DIST, a statistically significant main effect was detected for TRIAL (p = .001) and M-
ABC (p = .010), but not AGE (p = .730).
Considering the intermediate task, for MT, a statistically significant main effect was
detected for TRIAL (p = .007), but not for M-ABC (p = .174) or AGE (p = .156). For
DIST, no statistically significant main effect were detected for TRIAL (p = .074), M-
ABC (p = .142) or AGE (p = .280).
Considering the complex task, for MT, a statistically significant main effect was
detected for TRIAL (p < .001), M-ABC (p = .028), and AGE (p < .001). For DIST, a
statistically significant main effect was also detected for TRIAL (p < .001), M-ABC (p
= .041), and AGE (p = .002).
69
Thus, motor ability was found to modulate the effect of increased difficulty on visual-
motor performance over repeated trials. When all children were considered, both MT
and DIST decreased over repeated trials for the simple task and the complex task.
For the intermediate task, only MT was found to decrease over repeated trials.
Motor ability was a statistically significant predictor of MT and DIST for the complex
task and of DIST for the simple task. Age was a statistically significant predictor of
MT and DIST for the complex task only. Findings are summarized in Table 4-5.
Table 4-5 Investigating the effect of motor ability on visual-motor performance over repeated trials when
task difficulty is taken into account (N = 24).
Analysis
Ln MT Ln DIST
Test value (df) p value Test value (df)
p value
Simple Task
TRIAL t(977) = -2.98 .003* t(972) = -3.23 .001*
AGE t(21) = -.48 .638 t(21) = -.35 .730
M-ABC t(21) = 1.92 .069 t(21) = 2.82 .010*
Intermediate Task
TRIAL t(1155) = -2.72 .007* t(1145) = -1.79 .074
AGE t(21) = -1.46 .156 t(21) = -1.11 .280
M-ABC t(21) = 1.41 .174 t(21) = 1.53 .142
Complex Task
TRIAL t(1182) = -8.33 <.001* t(1187) = -6.70 <.001*
AGE t(21) = -3.82 .001* t(21) = -3.48 .002*
M-ABC t(21) = 1.70 .028* t(21) = 2.18 .041*
Abbreviations: MT: movement time; DIST: distance travelled; DIFF: Task difficulty; TRIAL: Trial number; M-ABC:
motor ability; Ln: natural log transformation applied to the data.
*: statistically significant at α = .05
4.4 VISUAL-MOTOR SKILL ACQUISITION: PATTERNS OF CHANGE
The third objective of this study was to further describe skill acquisition in children
with and without DCD by investigating patterns of change in visual-motor task
70
performance over repeated trials. Two analyses were performed to meet this
objective and investigate the impact of task difficulty and motor ability on patterns of
change in visual-motor task performance over repeated trials.
4.4.1 RATES OF VISUAL-MOTOR SKILL ACQUISITION
In TD children (n = 12), only the complex task led to skill acquisition over repeated
trials. Accordingly, only rates of visual-motor skill acquisition on this task were
examined. Non-linear regression analysis revealed that the minimum value the
modeled function approached for MT was a = 7.21 s, and for DIST was a = 269 mm.
The magnitude of the acquisition required for MT was b = 32.37 s, and for DIST was
b = 1168.3 mm. The rate of acquisition for MT was c = 6.25 trials and for DIST was c
= 6.73 trials (Figure 4-5). In other words, by trial 7, children had reached more than
63.2% of the magnitude of the acquisition required for MT (i.e. 20.46 s) and for DIST
(i.e. 738.37 mm).
Legend
Shades of blue: TD children (n = 12)
Predicted MT Predicted DIST Actual MT Actual DIST Abbreviations: MT: movement time; DIST: distance travelled
0
500
1000
1500
2000
2500
3000
3500
0 10 20 30 40 50
DIS
T (
mm
)
Trial Number
Changes in Distance Travelled during Complex Task in TD Children
DIST = 269 + 1168.3*e(-trial/6.73)
Trial 6
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50
MT
(s)
Trial Number
Changes in Movement Time during Complex Task in TD Children
MT = 7.21 + 32.37*e(-trial/6.25)
Trial 6
Figure 4-5 Illustrating rates of skill acquisition and changes in visual-motor task performance over repeated
trials during the complex task for TD children (n = 12).
71
0
1
2
3
4
5
0 10 20 30 40 50
MT
(s)
Trial Number
Changes in Movement Time during Simple Task in all Children
0
100
200
300
400
0 10 20 30 40 50
DIS
T (
mm
)
Trial Number
Changes in Distance Travelled during Simple Task in all Children
When all children were considered together, all three tasks led to statistically
significant changes in some aspects of visual-motor performance over trials (as
evidenced by the statistically significant effect for TRIAL). Skill acquisition was noted
for all three tasks for MT, but only for the simple and complex task for DIST.
Accordingly, rates of visual-motor skill acquisition were examined for both variables
and for all three tasks, except for DIST for the intermediate task.
For the simple task, when considering all children (N = 24), the non-linear regression
analysis revealed that the minimum value the modeled function approached for MT
was a = 1.21 s, and for DIST was a = 145.8 mm. The magnitude of the acquisition
required for MT was b = 3.16 s, and for DIST was b = 130.6 mm. The rate of
acquisition for MT was c = 0.76 trials and for DIST was c = 0.87 trials (Figure 4-6).
Accordingly, after the first trial, children had already reached 63.2% of the
magnitude of the acquisition required for MT and DIST. For comparative purposes,
the regression was also performed for TD (n = 12); parameters estimates and
regression curves of TD children are presented in Appendix G.
Legend
Shades of black: all children (N = 24)
Predicted MT Predicted DIST Actual MT Actual DIST Abbreviations: MT: movement time; DIST: distance travelled
MT = 1.21 + 3.16*e(-trial/0.76)
DIST = 145.8 + 130.6*e(-trial/0.87)
Figure 4-6 Illustrating rates of skill acquisition and changes in visual-motor task performance over repeated
trials during the simple task for all children (N = 24).
72
For the intermediate task, when considering all children (N = 24), the non-linear
regression analysis revealed that the minimum value the modeled function
approached for MT was a = 2.50 s. The magnitude of the acquisition required for MT
was b = 1.83 s and the rate of acquisition was c = 5.17 trials (Figure 4-7).
Accordingly, children reached 63.2% of the magnitude of the acquisition required for
MT by the 6th trial. In other words, by trial 6, MT was less than 3.17 s. For
comparative purposes, the regression was also performed for TD (n = 12) children;
parameters estimates and regression curves of TD children are presented in
Appendix G.
Figure 4-7 Illustrating changes in movement time over repeated trials
Finally, for the complex task, when considering all children (N = 24), the non-linear
regression analysis revealed that the minimum value the modeled function
approached for MT was a = 9.14 s, and for DIST was a = 345.6 mm. The magnitude
of the acquisition required for MT was b = 70.24 s, and for DIST was b = 3,750.6
mm. The rate of acquisition for MT was c = 3.54 trials and for DIST was c = 2.82
trials (Figure 4-8). Accordingly, by trial 4, children had reached 63.2% of the
magnitude of the acquisition required for MT. By trial 3, children had reached 63.2%
of the magnitude of the acquisition required for DIST. In other words, by trial 4 MT
was 34.99 s, and by trial 3 DIST was 1725.82 mm. Thus, when children with DCD
were added to the group, the magnitude of the acquisition required for MT and DIST,
0
1
2
3
4
5
0 10 20 30 40 50
MT
(s)
Trial Number
Changes in Movement Time during Intermediate Task in all Children
Legend Predicted MT Actual MT Abbreviation: MT: movement time
MT = 2.50 + 1.83*e(-trial/5.17)
Trial 5
73
0
500
1000
1500
2000
2500
3000
3500
0 10 20 30 40 50
DIS
T (
mm
)
Trial Number
Changes in Distance Travelled during Complex Task in all Children
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50
MT
(s)
Trial Number
Changes in Movement Time during Complex Task in all Children
as well as the minimum values approached by both variables increased, but the rate
of acquisition to reach 63.2% of the acquisition decreased. For comparative
purposes, the regression curves of TD children (n = 12) were drawn alongside those
of all children.
Figure 4-8 Illustrating changes in visual-motor task performance over repeated trials during the complex
task for all children and comparing predicted visual-motor task performance between TD children (n =
12) and all children (N = 24).
Legend Shades of black: all children (N = 24) Predicted MT Predicted DIST Actual MT Actual DIST Shades of blue: TD children (n = 12) Predicted MT Predicted DIST Abbreviations: MT: movement time; DIST: distance travelled
MT = 9.14 + 70.24*e(-trial/3.54)
DIST = 345.6 + 3750.6*e
(-trial/2.82)
Trial 3 Trial 2
0
500
1000
1500
2000
2500
3000
3500
0 10 20 30 40 50
DIS
T (
mm
)
Trial Number
Comparing Changes in Distance Travelled during Complex Task
0
10
20
30
40
50
60
70
0 10 20 30 40 50
MT
(s)
Trial Number
Comparing Changes in Movement Time during Complex Task
74
4.4.2 PATTERNS OF CHANGE IN VISUAL-MOTOR TASK PERFORMANCE OVER
REPEATED TRIALS
In TD children (n = 12), skill acquisition was only evident for the complex task.
Accordingly, only control charts for this task were examined. Boundaries and special
causes for MT and DIST control charts of TD children (n = 12) for the complex task
are illustrated in Figure 4-9. For MT, 5 data points falling outside the upper control
limit were identified within the first 8 trials, as well as 3 consecutive runs starting at
trial 29. No trends of 7 successive data points were found. The largest trend found
was an upward trend of 5 data points between trials 32-36. With regards to the
boundaries of the control charts, all data points fell below the + 2 sigma after 8 trials,
below +1 sigma after 11 trials, and below the centre line after 29 trials. Thus, for TD
children, the first 8 trials were characterized by extreme values and large MT,
followed by a rapid decrease in MT until trial 11, and a more gradual decrease
afterwards.
For DIST, 6 data points falling outside the upper control limit were identified within
the first 8 trials, as well as 5 consecutive runs starting at trial 16. No trends of 7
successive data points were found. Two 4-point downward trends found between
trials 14-17 and trials 29-32, and two upward trends between trials 5-8 and 45-48.
With regards to the limits of the control charts, all points fell below the + 2 sigma
after 8 trials, below +1 sigma after 11 trials, below the centre line after 15 trials, and
below -1 sigma after 42 trials. Thus, for TD children, the first 8 trials were
characterized by extreme values and large DIST, followed by a rapid decrease in
DIST until trial 15, and a more gradual decrease afterwards.
75
Figure 4-9 Illustrating patterns of change in visual-motor performance during complex task TD children (n = 12)
Legend
Shades of blue: TD children (n = 12) MT DIST
Trends: upward trend downward trend extreme value Abbreviations: MT: movement time; DIST: distance travelled; UCL: upper control limit; LCL: lower control limit
-
5
10
15
20
25
30
35
40
0 5 10 15 20 25 30 35 40 45 50
MT(
mm
)
Trial Number
Patterns of Change in MT during Complex Task in TD
MT (TD)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
-
200
400
600
800
1,000
1,200
1,400
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Patterns of Change in DIST during Complex Task in TD
DIST (TD)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
76
When all children (N = 24) were considered together, all three tasks led to
statistically significant changes in some aspects of visual-motor performance over
trials. Skill acquisition was noted for all three tasks for MT, but only for the simple
and complex task for DIST. Accordingly, control charts were examined for both
variables and for all three tasks, except for DIST for the intermediate task.
Considering the simple task in all children (N = 24), for MT, 2 data points falling
outside the upper control limit were identified within the first 2 trials, and no runs
were found. No trends of 7 successive data points were identified. Only one
downward trend of 5 data points was identified, between trials 16-20. With regards
to the limits of the control charts, all points fell below the + 2 sigma after 41 trials, but
never consistently fell below +1 sigma (Figure 4-10).
Legend
Shades of black: All children (N = 24)
MT DIST
Trends:
upward trend
downward trend
extreme value Abbreviations: MT: movement time; DIST: distance travelled; UCL: upper control limit; LCL: lower control limit
0.75
1.00
1.25
1.50
1.75
2.00
2.25
0 5 10 15 20 25 30 35 40 45 50
MT(
mm
)
Trial Number
Patterns of Change in MT during Simple Task (N=24) with TD Chart Limits
MT (N=24)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
100
110
120
130
140
150
160
170
180
190
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Patterns of Change in DIST during Simple Task (N=24) with TD Chart Limits
DIST (N=24)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
Figure 4-10 Illustrating patterns of change in visual-motor task performance during
simple task for all children (N = 24)
77
0.75
1.00
1.25
1.50
1.75
2.00
2.25
0 5 10 15 20 25 30 35 40 45 50
MT
(mm
)
Trial Number
Patterns of Change in MT during Simple TaskComparing the modulating effect of motor ability
MT (N=24)
MT (TD)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
Figure 4-11 Illustrating patterns of change in visual-motor task performance during simple task for all children (N
= 24) and TD children (n = 12) to demonstrate the impact of adding children with DCD.
For DIST, 2 data points falling outside the upper control limit were identified within
the first 2 trials, and no runs were found. No trends of 7 successive data points were
identified. Two 5-point downward trends were found between trials 14-18 and 22-26.
With regards to the limits of the control charts, all points fell below the + 2 sigma
after 29 trials but never fell consistently below + 1 sigma (Figure 4-10). To visually
compare the modulating effect of adding children with lower motor abilities on
patterns of change of visual-motor performance, MT and DIST control charts of all
children were drawn on the control charts of TD children (Figure 4-11).
Legend TD children (n = 12) MT DIST All children (N = 24) MT DIST Abbreviations: MT: movement time; DIST: distance travelled; UCL: upper control limit; LCL: lower control limit
100
120
140
160
180
200
220
240
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Patterns of Change in DIST during Simple TaskComparing the modulating effect of motor ability
DIST (N=24)
DIST (TD)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
78
Considering the intermediate task in all children (N = 24), for MT, 2 data points
falling outside the upper control limit were identified within the first 9 trials, and no
runs were found (Figure 4-12). No trends of 7 successive data points were identified.
Only one downward trend of 4 data points was identified, between trials 40-43, and
one upward trend between trials 46-49. With regards to the limits of the control
charts, all points fell below the + 2 sigma after 9 trials, below +1 sigma after 40 trials,
but never consistently fell below the centre line. To visually compare the modulating
effect of adding children with lower motor abilities on patterns of change of visual-
motor performance, MT and DIST control charts of all children were drawn on the
control charts of TD children (Figure 4-13).
Figure 4-12 Illustrating patterns of change in movement time during intermediate task for all children (N
= 24).
Legend
Shades of black: all children (N = 24) MT
Trends: Upward trend Downward trend Extreme value Abbreviations: MT: movement time; UCL: upper control limit; LCL: lower control limit
1.25
1.75
2.25
2.75
3.25
3.75
4.25
4.75
5.25
0 5 10 15 20 25 30 35 40 45 50
MT(
mm
)
Trial Number
Patterns of Change in MT during Intermediate Task (N=24)with TD Chart Limits
MT (N=24)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
79
Figure 4-13 Illustrating patterns of change in movement time during intermediate task for all children (N
= 24) and TD children (n = 12) to demonstrate the impact of adding children with DCD.
Legend
Shades of blue: TD children (n = 12) MT Shades of black: all children (N = 24) MT
Abbreviations: MT: movement time; UCL: upper control limit; LCL: lower control limit
Considering the complex task in all children (N = 24), for MT, 6 data points falling
outside the upper control limit were identified within the first 8 trials, as well as 2
consecutive runs after trial 29. No trends of 7 successive data points were found.
The largest trend found was a downward trend of 5 data points between trials 1-5.
With regards to the limits of the control charts, all points fell below the + 2 sigma
after 8 trials, below +1 sigma after 16 trials, and below the centre line after 45 trials
(Figure 4-14).
For DIST, 6 data points falling outside the upper control limit were identified within
the first 8 trials, as well as 4 runs after trial 16. No trends of 7 successive data points
were found. Five downward trends were found, the largest ones being between trials
1-5 and 8-12. With regards to the limits of the control charts, all points fell below the
+ 2 after 9 trials, below + 1 sigma after 11 trials, and below the centre line after 16
1.25
1.75
2.25
2.75
3.25
3.75
4.25
4.75
5.25
0 5 10 15 20 25 30 35 40 45 50
MT(
mm
)
Trial Number
Patterns of Change in MT during Intermediate TaskComparing the modulating effect of motor ability
MT (N=24)
MT (TD)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
80
-
5
10
15
20
25
30
35
40
45
50
0 5 10 15 20 25 30 35 40 45 50
MT(
mm
)
Trial Number
Patterns of Change in MT during Complex Task (N=24)with TD Chart Limits
MT (N=24)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
68.76.
trials. To visually compare the modulating effect of adding children with lower motor
abilities on patterns of change of visual-motor performance, MT and DIST control
charts of all children were drawn on the control charts of TD children (Figure 4-15).
Legend
Shades of back: all children (N = 24) MT DIST
Trends: Upward trend Downward trend Extreme value Abbreviations: MT: movement time; DIST: distance travelled; UCL: upper control limit; LCL: lower control limit
Figure 4-14 Illustrating patterns of change in visual-motor performance during intermediate task for
all children (N = 24).
81
-
200
400
600
800
1,000
1,200
1,400
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Patterns of Change in DIST during Complex TaskComparing the modulating effect of motor ability
DIST(N=24)
DIST(TD)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
Legend
Shades of blue: TD children (n = 12) MT DIST
Shades of black: all children (N = 24) MT DIST Abbreviations: MT: movement time; DIST: distance travelled; UCL: upper control limit; LCL: lower control limit
Figure 4-15 Illustrating patterns of change in visual-motor performance during complex task for
all children (N = 24) and TD children (n = 12) to demonstrate the impact of adding children with
DCD.
-
5
10
15
20
25
30
35
40
0 5 10 15 20 25 30 35 40 45 50
MT(
mm
)
Trial Number
Patterns of Change in MT during Complex TaskComparing the modulating effect of motor ability
MT(N=24)
MT(TD)
UCL
+2 sigma
+1 sigma
Centre line
-1 sigma
-2 sigma
LCL
82
CHAPTER 5
DISCUSSION
83
5 DISCUSSION
This study was conducted to describe visual-motor skill acquisition and task
performance in children with DCD. To that end, the performance of a visual-motor
task with increasing levels of difficulty was measured and described when children
with and without DCD first performed the task, and over repeated trials. Three
analyses were conducted to describe the modulating effect of task difficulty and
motor ability on visual-motor task performance and skill acquisition, first in TD
children (n = 12), and then in all children (N = 24) recruited for this study, with and
without DCD. Finally, the impact of motor ability on patterns of change in task
performance over repeated trials during skill acquisition was also described. The
discussion will revisit the analyses conducted to provide an interpretation of the
findings. This discussion will also contextualize the findings within the current
literature on DCD to add to the characterization of visual-motor task performance
and skill acquisition in children with DCD.
5.1 VISUAL-MOTOR TASK PERFORMANCE: MODULATING VARIABLES
The first objective of this study was to describe visual-motor task performance in
children with and without DCD. The results from the two analyses performed served
to describe the modulating effect of task difficulty on initial visual-motor task
performance in TD children, and to describe the modulating effect of motor ability on
initial visual-motor task performance in all children.
5.1.1 DESCRIBING THE EFFECT OF TASK DIFFICULTY ON INITIAL VISUAL-MOTOR
TASK PERFORMANCE IN TD CHILDREN
For TD children, the findings suggest that task difficulty did modulate initial
movement time and distance travelled, although it did not modulate initial movement
time during the simple task. Variables other than task difficulty likely modulate, or
explain, movement time for a task as simple as using a computer mouse.
Nevertheless, the results obtained in this study are among the first ones to
demonstrate that the association between task difficulty and task performance
proposed by Guadagnoli and Lee (2004) also holds in children, which was important
84
to ascertain before exploring the modulating effect of motor ability on task
performance.
The procedure used to study the effect of task difficulty on visual-motor task
performance in children is itself novel. Few studies systematically exploring the
effect of task difficulty on visual-motor task performance in children were found in
the literature. Bo, Contreras-Vidal, Kagerer, and Clark (2006) investigated the effect
of a task’s increased visual-motor complexity on arm movements of children 4-8
years of age. However, their conceptualization of visual-motor complexity was
significantly different than the notion of task difficulty used in this study, limiting the
possibility to compare their results with the ones discussed here. Bo and colleagues
(2006) used an aiming task similar to the one used in this study. However, they
increased task complexity by keeping the structural and parametric features of their
task constant but changing the way visual feedback was presented (horizontally or
vertically on a computer screen). Furthermore, they occluded vision of the hand/arm
during task performance. Arguably, by preventing vision of the hand/arm during
performance, the transformation used for motor control had to rely on proprioceptive
feedback of the hand, as opposed to the more natural combination of visual and
proprioceptive feedback. In addition to their conceptualization of visual-motor
complexity differing from the one used in this study, their findings also differ. Overall,
their results suggested that visual-motor complexity did not differentiate task
performance for intermediate and complex tasks. In contrast, findings from this study
did suggest that task difficulty differentiated task performance for intermediate and
complex tasks.
Indeed, the findings obtained from TD children confirm that by specifically
manipulating the structural and parametric features of the task, the visual-motor task
designed successfully offered incremental levels of difficulty. As detailed in section
3.3, the simple task was designed so that both the task’s structural and parametric
features would have been previously learned by children, thus calling upon an
already acquired visual-motor transformation. The intermediate task was designed
so that most of the structural features of the simple task were kept but combined to
85
different parametric features. The complex task was designed so that both structural
and parametric features would have to be acquired before the task could be
performed with efficiency. When a visual-motor task is performed and the structural
and parametric features of the task must be learned, no previously acquired visual-
motor transformation can be used in motor control processes (Sailer et al., 2005;
Wolpert & Flanagan, 2010). As such, no initial improvement in performance is
typically reported until the rules governing the visual-motor transformation become
better defined and task structure is acquired. In that regard, it was important to
confirm the impact of task difficulty on initial visual-motor task performance to make
sure that the task designed would be challenging enough to offer children a chance
to acquire new visual-motor skills over repeated trials.
It is also interesting to note that the focus on initial visual-motor task performance in
this first part of the analysis is novel and goes further than the description of visual-
motor task performance largely discussed in the literature. For example, while Bo
and colleagues (2006) did take into account the repeated nature of the data
collected in their analyses, they did not explore the initial impact of increased visual-
motor complexity on visual-motor task performance. Accordingly, their inconclusive
results could potentially be explained by rapid changes in performance over
repeated trials, rather than by complexity. By first attempting to limit the influence of
repeated trials on visual-motor task performance and by focusing solely on initial
visual-motor performance, the systematic analysis performed in this thesis ensures
that a clear description of the modulating effect of motor ability on visual-motor task
performance emerges separate from skill acquisition.
5.1.2 DESCRIBING THE EFFECT OF MOTOR ABILITY ON INITIAL VISUAL-MOTOR
PERFORMANCE WHEN TASK DIFFICULTY IS CONSIDERED
Once the impact of task difficulty on initial visual-motor task performance was
understood in typically developing children, the same impact was explored in all
children, with and without DCD (N = 24). This time, the modulation effect of motor
ability on the effect of increased task difficulty on children’s initial visual-motor
performance was explored.
86
The work of Guadagnoli and Lee (2004) presented earlier shaped this analysis.
Guadagnoli and Lee have proposed that the impact of task difficulty on task
performance is influenced by skill level. In this study, because visual-motor task
performance was under investigation, a parallel was suggested between children’s
motor ability and skill level: the effect of task difficulty on visual-motor task
performance was proposed to be modulated by motor ability.
Findings suggest that motor ability did modulate the effect of task difficulty on initial
visual-motor performance, although the modulation effect was limited to the complex
task. Indeed, with regards to the simple and intermediate tasks, task difficulty did not
interact with motor ability to predict initial visual-motor performance. According to
Guadagnoli and Lee, this could either suggest that either both the simple and the
intermediate tasks were of the same level of difficulty, or that the intermediate task
was not difficult enough to significantly challenge children with lower motor ability
and thus to affect their performance. As already discussed above, since the levels of
difficulty of the visual-motor tasks in this study were specifically modified by
changing their structural and parametric features, and since task difficulty modulated
initial visual-motor performance in TD children, it is unlikely that both tasks were of
the same level of difficulty. The alternative explanation is thus likely: the intermediate
task was not difficult enough to challenge children with DCD, and thus when added
to the group of TD children, motor ability did not contribute to further explaining initial
visual-motor performance. For the intermediate task, the results obtained in this
study suggest that children with DCD were as able as TD children to call upon an
already learned visual-motor transformation to adapt to the new parametric features
of the task. However, a statistically significant interaction was detected between task
difficulty and motor ability for the complex task. Thus, while motor ability did not
affect initial visual-motor performance for the simple and intermediate tasks, it did
affect initial performance for the complex task. When adding children with DCD to
the group, initial movement times and travelled distances became longer than when
only TD children were considered. These results are in line with the premise of the
Challenge Point Model (Guadagnoli & Lee, 2004), which states that the impact of
task difficulty on task performance is influenced by skill level. In this study motor
87
ability thus modulated the impact of task difficulty for the complex task where both
structural and parametric task features had to be acquired.
No studies that systematically compared the modulating influence of motor ability on
the effect of task difficulty on initial visual-motor performance were found in the
literature. However, the effect of task difficulty on the visual-motor performance of
children with DCD has been explored through comparative study designs using
Fitts’s law. As discussed earlier, principles from Fitts’s law have been applied to
increase the difficulty of a visual-motor task by changing its parametric features,
such as varying the size of the target to reach. Task difficulty has also been explored
by varying the structural features of tasks, such as varying the structure of the
movement required (e.g., cyclical versus discrete movements) (Smits-Engelsman et
al., 2003, 2006). Nevertheless, the modulation effect of motor ability on the initial
effect of increased difficulty has not been explored. The results obtained from this
study are thus the first to demonstrate that for visual-motor tasks that are well
learned, or tasks requiring an adaptation to novel parametric task features, motor
ability does not modulate children’s visual-motor task performance. However, the
results also demonstrate that when a task is more complex and both structural and
parametric task features must be acquired, the initial effect of the increased task
difficulty is greater for children with DCD than for TD children.
5.2 VISUAL-MOTOR SKILL ACQUISITION: MODULATING VARIABLES
The second objective of this study was to describe visual-motor skill acquisition in
children with and without DCD. Guadagnoli and Lee (2004) propose that, in addition
to modulating task performance, task difficulty is also essential to skill acquisition. It
is through increased task difficulty that information about task performance can be
gained, and in turn used to improve performance. Accordingly, to meet the second
objective, results from the analyses performed to describe the modulating effect of
repeated trials on task performance in TD children, and the modulating effect of
motor ability on skill acquisition, are discussed here.
88
5.2.1 DESCRIBING THE EFFECT OF REPEATED TRIALS ON VISUAL-MOTOR TASK
PERFORMANCE IN TD CHILDREN
For children developing typically, findings indicate that trial number had a modulating
effect only for the complex task. Thus, skill acquisition only occurred for the complex
task, as it would seem that the complex task was the only one offering enough
information to promote skill acquisition over repeated trials in TD children.
For the simple and intermediate tasks using the computer mouse or the joystick,
results suggest that visual-motor task performance was stable throughout. For the
simple task, mouse-use was well learned and, congruent with what was presented
earlier when discussing the Challenge Point Model (Guadagnoli & Lee, 2004), it is
possible that the information gained from performing the task repeatedly did not offer
additional information that could be used by children to improve their performance.
For the intermediate task, although the task was not well learned, because children
had less experience using a joystick, the structural features of the task were similar
to those of the simple task. As discussed above, in such a case, adaptation to new
parametric features occurs quickly, so it is not surprising that results suggest stable
visual-motor task performance over repeated trials.
5.2.2 DESCRIBING THE EFFECT OF MOTOR ABILITY ON VISUAL-MOTOR TASK
PERFORMANCE OVER REPEATED TRIALS
Results indicated that when the visual-motor task performance of all children was
analysed over repeated trials, improvements in task performance were evident for all
three tasks, except for distance travelled during the intermediate task. In other
words, when children with DCD were added to the group, skill acquisition was
evident even for the simple, well-learned task.
In her study of visual-motor skill acquisition in children with and without DCD,
Missiuna (1994) had also noted a significant decrease in movement time over
repeated trials on her visual-motor task using a computer mouse. However, in her
study, both TD and children with DCD demonstrated visual-motor skill acquisition
89
over repeated trials, while in this study TD children only demonstrated skill
acquisition for the complex task.
As mentioned earlier, according to Guadagnoli and Lee’s Challenge Point Model
(2004), increased difficulty is essential to skill acquisition, as it is through increased
difficulty that useful information about task performance can be gained.
Nevertheless, the information that can be gained from task performance is
influenced by one’s skill level. For the children in this study, findings suggest that
motor ability may have influenced the level of information that could be extracted
from task performance and then used to promote skill acquisition. The fact that no
changes in visual-motor task performance for the simple and intermediate tasks over
repeated trials were evident when only TD children were considered, but that
changes were detected when adding children with DCD, supports this argument.
These findings further support the need to visually analyse patterns of change in
visual-motor performance to better understand the effect of motor ability over
repeated trials.
Considering independent variables added to the model to explore factors potentially
influencing visual-motor task performance, findings suggested that motor ability
influenced distance travelled in the simple task, and both movement time and
distance travelled for the complex task. For the complex task, the result is not
surprising; as already discussed above, motor ability influenced initial visual-motor
task performance here. However, for the simple task, it is interesting to note that
while motor ability did not influence initial distance travelled, it did influence
performance over repeated trials. This finding suggests that it is not the initial
distance travelled that was modulated by motor ability but the changes in distance
travelled that occurred over time. While this finding is in line with the current
characterization of children’s visual-motor task performance, which proposes that
children with DCD are less accurate than their peers (Elders et al., 2010; Maruff et
al., 1999; Smits-Engelsman et al., 2006), it goes beyond such a general description
by considering task difficulty and differentiating the modulating effect of motor ability
on initial task performance from task performance over repeated trials.
90
Movement time over repeated trials for the simple task was not influenced by motor
ability, although performance changes over repeated trials that were absent when
considering only TD children were noted when considering all children. This finding
is informative when combined with the discrepancy in reports about movement times
of children with DCD presented in the background section. If the discrepancy of
slower, faster, and “same as” movement times represents existing subtypes of
children with DCD and was present within this sample, it could possibly explain why
motor ability did not come out as a significant predictor of movement time for the
simple task. Further investigation of individual subtypes of visual-motor task
performance would have to be explored in a larger sample of children with DCD to
consider this possibility.
It is interesting to note that for the intermediate task, which was well known neither
to children with DCD nor to TD children, motor ability was not a factor that affected
distance travelled or movement time over repeated trials. Here again, while
movement time over repeated trials for the intermediate task was not influenced by
motor ability, performance changes over repeated trials that were absent when
considering only TD children were noted when considering all children. As discussed
above for the simple task, this finding is informative when considered alongside the
discrepant reports about movement times of children with DCD presented in the
background section. The findings that motor ability does not influence visual-motor
task performance for the intermediate task reminds researchers to be cautious when
making general statements about the poor accuracy of children with DCD; it would
seem that it is important to consider the difficulty of the visual-motor task discussed
before making such statements.
Interestingly, children’s age was only found to influence the performance of the
complex task. Missiuna (1994) also identified an age interaction in her study of
visual-motor skill acquisition in children with DCD. Just like motor ability can be
discussed as a skill level potentially influencing performance, age could be
conceptualized similarly. In fact, as discussed earlier, developmental trends are
seen in many components affecting successful visual-motor performance, and, in
91
children, growth and maturation are considered influential factors explaining
changes in performance (Thomas & Thomas, 2008). The age range selected in this
study attempted to avoid the critical period related to the reweighting of motor-
control mechanisms known to occur around 8 years of age (Bard et al., 1990; Fayt
et al., 1993; Ferrel-Chapus et al., 2002). Yet that is not to say that this critical period,
or other critical periods related to other neural processes that could potentially affect
skill acquisition and task performance, were completely avoided in the group of
children tested. The paucity of studies having explored components known to affect
visual-motor skill acquisition and task performance in typically developing children
limits our efforts to explore this finding further.
5.3 VISUAL-MOTOR SKILL ACQUISITION: PATTERNS OF CHANGE
The third objective of this study was to further describe skill acquisition in children
with and without DCD by exploring the impact of motor ability on patterns of change
in task performance over repeated trials. The results from the two analyses
performed are discussed here.
5.3.1 PATTERNS OF CHANGE IN TD CHILDREN
For TD children, the non-linear regression analysis suggested that, for the complex
task, 63.2% of the visual-motor performance improvement that occurred over
repeated trials had occurred by 6.25 trials for movement time, and by 6.73 trials for
distance travelled. Indeed, visual graph analysis revealed that before the 8th trial,
visual-motor task performance consistently fell above the upper control limit of the
control charts. A sharp improvement in visual-motor task performance was evident
between the 8th and 15th trials, followed by a more gradual improvement throughout
the remaining trials.
It is interesting to compare the results obtained here to those reported by Sailer and
colleagues (2005), since the complex task used in this study was inspired by their
task design. Similar to the findings here for TD children, Sailer and colleagues
reported a three-stage pattern of skill acquisition for 8 of their 10 adult participants:
exploratory, skill-acquisition, and skill-refinement stages. A more direct comparison
92
between their results and those found in this study is not possible because of
differences in their task design, variables reported and analysis. Nevertheless, it is
interesting to note that the patterns of visual-motor task performance changes of the
TD children in this study were similar in shape to those of most of the adults in the
study by Sailer and colleagues.
5.3.2 THE EFFECT OF MOTOR ABILITY ON PATTERNS OF CHANGE
When all children were considered together, visual-motor skill acquisition was also
detected for the simple and intermediate tasks. Movement time decreased over
repeated trials for all three tasks and distance travelled decreased over repeated
trials for the simple and complex tasks. The parameter estimates and associated
confidence limits obtained from the non-linear regression analysis offered a general
idea of how performance changed across trials on those tasks. Visual patterns of
change in visual-motor performance were also examined to explore and
characterize the modulating effect of motor abilities on visual-motor acquisition.
With regards to the simple task, in her study of visual-motor skill acquisition,
Missiuna (1994) had detected that although the movement times of children with
DCD were initially similar to those of TD children, TD children improved more than
children with DCD and their movement times decreased after repeated trials. Yet her
statistical analysis revealed that the rate of change of movement time was similar for
both groups. Similar results were obtained in this study. As determined in previous
studies, motor ability did not affect initial visual-motor task performance.
Nevertheless, when children with DCD were added to the group, trial number was
found to be a significant predictor of visual-motor task performance, suggesting
changes across trials. An examination of the non-linear regression parameter
estimates and of related confidence limits suggests that the lower value of
movement time and distance travelled approached by TD children was smaller than
that of all children considered together. Nonetheless, a visual analysis of control
charts suggests that both for TD children and for the group consisting of all children,
performance rapidly decreased within the first three trials. Changes in movement
time were then characterized by a gradual decrease until Trial 20 and a plateau with
93
performance oscillating within a range of approximately 0.25 seconds thereafter.
After the initial rapid decrease in performance, distance travelled seemed to reach a
plateau and vary within a range of about 30 mm without demonstrating further
improvement in performance. An analysis of special causes from the control charts
did not yield additional characterizing information. A visual analysis of skill
acquisition curves thus suggests that the addition of children with DCD to the TD
group did not affect patterns of change, other than raising the lower limit approached
by the skill acquisition curve. This result is thus similar to the results obtained by
Missiuna (1994), suggesting that in this study, for the simple task, patterns of
changes over repeated trials were similar, regardless of motor ability.
For the intermediate task, when children with DCD were added to the group, trial
number modulated visual-motor task performance only for movement time; and
movement time was found to decrease over repeated trials. Parameter estimates
and confidence limits from the non-linear regression analysis suggest that changes
in movement time over repeated trials were similar for both groups, including the
lowest value approached by the acquisition curves. This time, a visual analysis of
patterns of change for both groups suggests an initial exploration period where
movement times were systematically higher until Trial 9, followed by a rapid
decrease in movement time until Trial 14, and a plateau where performance
oscillated within approximately 0.25 seconds thereafter. An analysis of special
causes from the control charts did not yield additional characterizing information.
Thus, for the intermediate task, patterns of change in movement time over repeated
trials were similar, regardless of motor ability.
For the complex task, since the effect of motor ability on initial visual-motor task
performance was already determined to be statistically significant, acquisition rates
were expected to also differ between groups, unless the initially measured difference
in visual-motor performance was to remain the same over repeated trials. The
examination of the non-linear regression parameter estimates and related
confidence limits suggests that the lower value of movement time and distance
travelled approached by the TD children was smaller than modelled when all
94
children were considered. In fact, the magnitude of acquisition needed to reach
those lower values was larger for the complete group (N = 24) than for TD children
alone. Thus, because the magnitude of the acquisition required for the complete
group was larger than that of the TD group, it can be concluded that the initially
measured differences in visual-motor performance did not remain constant over
repeated trials.
It took TD children almost seven trials for their visual-motor performance to reach
63.2% of the required acquisition, predicted movement time being 17.8 s and
predicted distance travelled being 682.1 mm. When all children were considered,
they reached 63.2% of the magnitude of the acquisition required by Trial 4 for
movement time and by Trial 3 for distance travelled. At Trial 4, predicted movement
time was 31.9 s, and at Trial 3, predicted distance travelled was 1252.9 mm.
Considering the known difference in initial visual-motor task performance, it is
important to interpret these results with caution. Taken at face value, the findings
would seem to suggest that children with DCD acquire visual-motor skills more
quickly than TD children, which is in fact not entirely true. The group with all children
reached the same predicted movement time of TD children (17.8 s) by Trial 8, and
their predicted distance travelled (682.1 mm) by Trial 7. Thus, the initial effect of task
difficulty was greater on the visual-motor performance of children with DCD for the
complex task. However, after the initial exploratory stage during which they acquired
the structural and parametric features of the task, their visual-motor performance
quickly improved to approach that of their typically developing peers. In fact, it took
children with DCD the same number of trials to reach the level of performance of TD
children.
Although motor ability was a significant predictor of visual-motor task performance
over repeated trials, visual graph analysis clearly suggests that the patterns of skill
acquisition were similar when children with DCD were added to the group of TD
children. For both groups, visual-motor acquisition was characterized by an initial
exploration period, during which visual-motor task performance was poor, followed
by a rapid skill-acquisition period, followed by gradual improvement in performance
95
thereafter. Such a pattern of skill acquisition resembles the one obtained by Sailer
and colleagues (Sailer et al., 2005) in most of their adult participants when using a
complex task similar to the one used in this study.
These findings suggest that it is the modulation of motor ability on the initial effect of
task difficulty that best explains visual-motor task performance of children over
repeated trials. Indeed, although the initial movement times and distances travelled
are higher during the exploratory stage when considering all children together in
comparison to the TD group, the number of trials it takes to move from the
exploratory stage to the acquisition stage seems to be similar, and so are the
gradual improvements in visual-motor performance.
5.4 CHARACTERIZING VISUAL-MOTOR ACQUISITION AND PERFORMANCE IN
CHILDREN WITH DCD
This study was conducted with the general objective of describing visual-motor skill
acquisition and task performance in children with and without DCD. Visual-motor
skill acquisition and task performance were first considered in TD children and then
in a group comprised of both children with DCD and TD children to explore how
motor ability modulated initial visual-motor task performance and changes in visual-
motor task performance over repeated trials. Given that children with DCD are
characterized by motor ability that falls below the norms of TD children, the findings
obtained in this study can be applied to characterize visual-motor skill acquisition
and task performance in children with DCD. Findings will thus be discussed in
relation to the current literature on DCD to add to the characterization of visual-
motor task performance and skill acquisition in children with DCD.
5.4.1 CHILDREN WITH DCD ARE LESS ACCURATE THAN THEIR PEERS
Findings from studies discussed in the background section had offered a partial
characterization of processes and end products resulting from the performance of
visual-motor tasks in children with DCD. Emerging from those studies was the
characterization that children with DCD have difficulty meeting the accuracy
demands of visual-motor tasks when compared to their peers, regardless of whether
96
ongoing (Chang & Yu, 2010; Di Brina, 2008; Rosenblum & Livneh-Zirinski, 2008;
Smits-Engelsman et al., 2001, 2003; Tseng et al., 2007) or end-point control (Smits-
Engelsman et al., 2006) is required. In this study, in an attempt to further explore this
characterization, distance travelled was chosen as a variable that would capture
children’s ability to meet the ongoing and end-point accuracy demands of the visual-
motor tasks as difficulty in either would yield longer distances travelled.
The general characterization that children with DCD are less accurate than their
peers when performing visual-motor tasks remains, although findings from this study
further specify this description. Indeed, task difficulty was found to have a differential
effect on children’s ability to meet a task’s accuracy demands. Furthermore, the
interaction of task difficulty with initial visual-motor task performance versus task
performance over repeated trials offered two different pictures. When considering
initial visual-motor task performance, children with DCD seem to be as able as their
typically developing peers to use motor control mechanisms to meet the accuracy
demands of a simple, well-learned visual-motor task or of a task for which the
structural features are already known and only parametric adaptation is required.
However, for a complex task for which structural and parametric features have to be
acquired, children with DCD are challenged to a greater extent than their peers, and
their performance consequently becomes less accurate.
When their task performance over repeated trials is considered, children with DCD
seem to be less accurate than their peers when performing a simple, well-learned
visual-motor task. While their peers demonstrate stable task performance with no
improvements over time, changes in distance travelled are evident in children with
DCD. This would suggest that, contrary to their peers, children with DCD continue to
show variable task performance, even for a well-learned visual-motor task. For a
task whose structural features are already known and for which parametric
adaptation is required, children with DCD perform similarly to their peers. While
performing a novel, complex task whose structural and parametric features must yet
be acquired, children with DCD were less accurate.
97
Accordingly, although the general characterization proposing that children with DCD
are less accurate than their peers when performing visual-motor tasks remains, it
should be further specified to include the impact of task difficulty on measures of
accuracy, as well as the impact of repeated trials. Furthermore, this study has
rekindled the discussion about changes in performance over time and suggests that
the accuracy of children with DCD does improve over repeated trials, even for
simple tasks that are thought to be well learned.
5.4.2 CHILDREN WITH DCD ARE SLOWER, FASTER, “SAME AS” THEIR PEERS
With regards to movement time, diverging characterizations emerged from the
studies discussed in the background section: in some studies children with DCD
moved slower than their peers (Chang & Yu, 2010; Rosenblum & Livneh-Zirinski,
2008; Smits-Engelsman et al., 2006), in others they moved faster (Chang & Yu,
2010; Di Brina et al., 2008; Smits-Engelsman et al., 2001, 2003), and in two they
moved at the same speed as their peers (Missiuna, 1994; Zwicker et al., 2011),
although after repeated practice they were slower for one of two studies (Missiuna,
1994).
The general characterization that children with DCD are slower than their peers
when performing visual-motor tasks remains; when motor ability was found to
modulate visual-motor task performance, visual-motor task performance was always
slower. However, the findings from this study further specify this description. Here
again, task difficulty was found to have a modulating effect on the movement time of
children with DCD when compared to their peers, but only when considering the
complex task.
When considering initial visual-motor task performance for a well-learned, simple
task for which the structural and parametric features were known, children with DCD
moved as fast as their peers. The same was true for a task for which the structural
features were known and parametric features had to be adapted. However, for a
complex task for which both structural and parametric features had to be acquired,
children with DCD were initially slower than their peers.
98
The same characterization remains when considering the performance of children
with DCD over repeated trials. However, on a simple, well-learned task, while their
peers did not show any improvements in movement time over repeated trials,
changes were evident in children with DCD. This would suggest that, contrary to TD
children, movement time in children with DCD did not reach a stable plateau even
for a well-learned visual-motor task. Decreases in movement time over repeated
trials were also evident for a task requiring adaptation to new parametric features, as
well as for a task requiring acquisition of both structural and parametric features.
Although the findings suggest that the general characterization that children with
DCD are slower than their peers when performing visual-motor tasks remains, it
should be qualified to include the impact of task difficulty and of repeated trials on
measures of movement time. Interestingly, none of the results suggested that
children with DCD performed faster than their peers on any of the visual-motor
tasks. Finally, just like for accuracy, this study has rekindled the discussion about
changes in performance over time and suggests that the movement times of
children with DCD do improve over repeated trials, even for simple tasks that are
thought to be well learned.
5.5 LIMITATIONS
5.5.1 EXPLORATORY NATURE OF THIS STUDY
This study was the first to use this particular computer-based task in children;
therefore, no normative data was available. While the issue of having no normative
data was addressed by including a control group of TD children, this was a small
and non-random group, and thus the representativeness of the TD children can be
questioned. Administering the visual-motor task to a larger group of children would
be important to have complete confidence in the TD data. However, the findings for
TD children did perform as expected based on the background literature discussed
in chapter 2. As such, it is reasonable to assume that the TD children constituted a
representative group.
99
In addition, power issues remain an important limitation of exploratory studies using
small samples. Since effect sizes were not available from prior studies, no data
could be used to estimate the sample size needed for this study. It is thus possible
that some of the results obtained could be due in part to the impact of a small
sample size on the power of the statistical analyses conducted (Portney & Watkins,
1993). Nevertheless, consultation with a statistician, the careful consideration of
parameter estimates and standard errors and fit statistics when modelling removed
concerns regarding the sample size.
5.5.2 MEASURING MOTOR ABILITY
The Movement Assessment Battery for Children (M-ABC; Henderson & Sugden,
1992) was selected as a measure of motor ability. However, the primary purpose of
the M-ABC is not to quantify motor ability but rather to identify motor difficulties in
children. Furthermore, although the M-ABC is typically used to identify motor
difficulties in children with DCD, some have reported that, at times, children with
DCD obtain scores above the 16th percentile (Rodger et al., 2007). Accordingly, the
equivalence proposed between motor ability and the M-ABC score could be
questioned. Nevertheless, given the exploratory nature of this study, the M-ABC was
selected as the measure that would be used to estimate motor ability. The M-ABC
was also selected because it has gained international recognition amongst
researchers and clinicians working with children with DCD (Geuze et al., 2001); it is
the standardized norm-referenced motor measure most frequently used in studies of
children with DCD. In consideration of these attributes of the M-ABC, the results
obtained from the analyses discussed above would suggest that it constituted an
appropriate measure to estimate motor ability. Certainly, when considering the
findings that motor ability modulates task performance for the complex task
alongside Figure 4-3 illustrating visual-motor task performance for DCD and TD
children, the equivalence proposed between motor ability and the M-ABC score
holds.
Another issue with the M-ABC is its ceiling effect and the resultant skewed
distribution of scores when administered to the general population. Nevertheless,
100
this distribution of scores was not found in the children assessed for this study, thus
avoiding the issues that a skewed distribution would have had on the regression
analyses conducted (Appendix A).
5.5.3 TECHNICAL ISSUES AND MISSING DATA
Numerous technical issues were encountered throughout data collection, resulting in
missing data. Typically, the issue with missing data is their effect on statistical
analyses: in analyses such as ANOVA, when there is data missing in a data set, the
entire row of data associated with that trial is ignored in the analysis. In this study,
the statistical analyses selected limited the impact of missing data by fitting the
mixed-effect models by maximum likelihood. This method of analysis thus handled
missing observations and unbalanced designs efficiently, leading to reliable
conclusions. However, the impact of missing data on the end results is unknown.
Since this study is an exploratory study, the results are, however, still extremely
valuable. The findings could be used as the basis for a larger-scale study that limits
the presence of missing data.
5.6 CLINICAL IMPLICATIONS AND FUTURE DIRECTIONS
5.6.1 A CONSIDERATION FOR IDENTIFYING CHILDREN WITH DCD
When establishing the rationale for this study, it was proposed that if visual-motor
skill acquisition difficulties were found to be a defining characteristic of DCD, the
presence of such a marker could possibly be used to offer novel diagnostic avenues
to explore in the future. The exploration of the effect of task difficulty on the visual-
motor performance of children with DCD seems to be a promising avenue for such
an endeavour. The findings strongly support that the greatest differentiation of
children with DCD from their peers occurs when the initial visual-motor task
performance of a complex visual-motor task is considered. A larger-scale study
based on children spanning multiple critical developmental periods would need to be
conducted to further explore this possibility.
101
5.6.2 A CONSIDERATION FOR INTERVENING WITH CHILDREN WITH DCD
When establishing the rationale for this study, it was proposed that if we were to
successfully develop interventions that would enable children with DCD to overcome
their visual-motor performance difficulties, such as their handwriting difficulties, a
strong understanding of their acquisition of visual-motor skills would be essential.
Findings from this study have demonstrated that task difficulty has a differential
effect on the initial visual-motor task performance of children with DCD. The
requirements of a complex task for which movement structure and parametric
features must be acquired seem to be especially challenging. Efforts to help children
with DCD understand the structural and parametric requirements of complex novel
tasks could potentially bridge the gap between their performance and that of their
peers, especially since findings from this study demonstrated that once complex
novel tasks were acquired, the visual-motor performance of children with DCD
improved, approaching that of their peers.
5.6.3 A CONSIDERATION FOR FUTURE RESEARCH TO CHARACTERIZE VISUAL-
MOTOR PERFORMANCE IN DCD
An interesting avenue for future research would be to continue with the study design
used by Sailer and colleagues (2005) and add eye-movement analysis during task
performance to develop a greater understanding of motor control mechanisms at
play. In their study, Sailer and colleagues demonstrated that gaze behaviour
followed the same three distinct stages of skill acquisition described earlier. During
the initial exploratory stage, gaze tended to pursue the cursor; during skill
acquisition, it typically began to mark predictive desired cursor positions; and during
skill refinement, gaze went directly to the targets as they appeared. Accordingly,
they proposed that during the learning stage, gaze pursued the cursor in an early
attempt to link hand movements with their visual consequences and to develop a
new visual-motor transformation. Then, once a basic knowledge of this new
transformation was developed, hand and eye movements started to be congruently
programmed, eye movement leading the cursor and predicting cursor direction more
and more accurately. During the last stage, gaze shifted directly to the target upon
102
its appearance, evidencing that the new visual-motor transformation had been
successfully implemented. Subjects now used their gaze to locate the target, and
this information could then be used for motor control to program the appropriate
hand movements (Sailer et al., 2005). As such, in addition to changes in visual-
motor task performance during skill acquisition, eye-movement analysis revealed to
be a good objective measure of the acquisition process. In particular, the patterns of
eye movements supporting manual action changed markedly depending on the
learning stage (Sailer et al., 2005).
5.7 CONCLUSIONS
This study explored visual-motor skill acquisition and task performance in children
with varying motor abilities in an attempt to specify the characterization of visual-
motor task performance in children with DCD and to move forward the discussion on
visual-motor skill acquisition. The exploration of the impact of task difficulty on
visual-motor task performance, and the special attention to the initial performance of
a task versus performance over repeated trials, is unique among studies discussing
visual-motor task performance in children with DCD. The major finding of this study
is that motor ability modulates the impact of task difficulty on visual-motor skill
acquisition and task performance. While children with DCD can still be characterized
as less accurate and slower than their peers, this characterization can now be
further specified and qualified. Children with DCD are as fast and as accurate as
their peers in their initial performance of a simple, well-learned task. However, they
are slower and less accurate when performing a complex and novel visual-motor
task. Over repeated trials, the visual-motor task performance of children with DCD
does improve, even for a simple task thought to be well learned. For a complex,
novel task, once children with DCD understand the features of the task, their
performance also improves and even approaches that of their peers. These findings
provide great hope for future interventions.
103
REFERENCES
104
REFERENCES
American Psychiatric Association [APA] (2000). Category 315.4. Developmental
coordination disorder. In Diagnostic and Statistical Manual of Mental Disorders (Text
Revision) (4th edition) (pp. 56-58). Washington, DC: APA.
Arbib M. A., Bonaiuto J. B., Jacobs S., & Frey, S. H. (2009). Tool use and the
distalization of the end-effector. Psychological Research, 73:441-62.
Astill, S. & Utley, A. (2008). Motor control, learning and development. CITY: Taylor &
Francis.
Bard, C., Hay, L., & Fleury, M. (1990). Timing and accuracy of visually directed
movements in children: control of direction and amplitude components. Journal of
Experimental Child Psychology, 50, 102-118.
Beery, K. & Beery, N. (2006). The Beery-Buktenica Developmental Test of Visual
Motor Integration Administration, Scoring, and Teaching Manual. Bloomington, MN:
Pearson.
Bo, J., Contreras-Vidal, J. L., Kagerer, F. A., & Clark, J. E. (2006). Effects of
increased complexity of visuo-motor transformations on children's arm movements.
Human Movement Science, 25, 553-567.
Breckenridge, M. B., Tallia, A. F., & Like, R. C. (1988). Display of small-area
variation in health-related data: a methodology using resistant statistics. Social
Science Medicine, 26, 141–151.
Brown, G. T. & Gaboury, I. (2006). The measurement properties and factor structure
of the test of visual-perceptual skills-revised: Implications for occupational therapy
assessment and practice. American Journal of Occupational Therapy, 60, 182-193.
Cairney, J., Veldhuizen, S., & Szatmari, P. (2010). Motor coordination and
emotional-behavioral problems in children. Current Opinion Psychiatry, 23, 324-329.
105
Callahan, C. D. & Barisa, M. T. (2005). Introduction to the special issue of
rehabilitation psychology: Issues in outcome measurement. Rehabilitation
Psychology, 50, 5.
Cantin, N., Polatajko, H. J., Thach, W. T., & Jaglal, S. (2004). Developmental
coordination disorder: exploration of a cerebellar hypothesis. Human Movement
Science, 26, 491-509.
Chang, S. & Yu, N. (2010). Characterization of motor control in handwriting
difficulties in children with or without developmental coordination disorder.
Developmental Medicine & Child Neurology, 52, 244-250.
Crawford, J. D., Medendorp, W. P., & Marotta, J. J. (2004). Spatial transformations
for hand-eye coordination. Journal of Neurophysiology, 92, 10-19.
Davidson, P. R. & Wolpert, D. M. (2005). Widespread access to predictive models in
the motor system: A short review. Journal of Neural Engineering, 2, 8313-8319.
Dewey, D. & Wilson, B. N. (2001). Developmental Coordination Disorder: What is it?
Physical & Occupational Therapy in Pediatrics, 20, 5-28.
Di Brina, C., Niels, R., Overvelde, A., Levi, G., & Hulstijn, W. (2008). Dynamic time
warping: a new method in the study of poor handwriting. Human Movement Science,
27, 242-55.
Donker, A. & Reitsma, P. (2007). Drap-and-drop errors in young children’s use of
the mouse. Interacting with Computers, 19, 257-266.
Dunford, C., Missiuna, C., Street, E., & Sibert, J. (2005). Children's perceptions of
the impact of developmental coordination disorder on activities of daily living. British
Journal of Occupational Therapy, 68, 207-214.
Draper, N. R. & Smith, H. (1998). Applied Regression Analysis, 3rd Edition,
Hoboken, NJ: Wiley.
106
Elders, V., Sheehan, S., Wilson, A. D., Levesley, M., Bhakta, B., & Mon-Williams, M.
(2010). Head-torso-hand coordination in children with and without developmental
coordination disorder. Developmental Medecine & Child Neurology, 52, 238-43.
European Consensus Conference (2010). Brussel
Estill, L., Ingvaldsen, R., & Whiting, H. (2002). Spatial and temporal constraints on
performance in children with movement co-ordination problems. Experimental Brain
Research, 147, 153-161.
Fayt, C., Minet, M., & Schepens, N. (1993). Children’s and adult’s learning of a
visuomanual coordination: Role of ongoing visual feedback and of spatial errors as a
function of age. Perceptual and Motor Skills, 77, 659-669.
Ferrel-Chapus, C., Hay, L., Oliver, I., Bard, C., & Fleury, M. (2002). Visuomanual
coordination in childhood: adaptation to visual distortion. Experimental Brain
Research, 144, 506-517.
Fitts, P. M. (1954). The information capacity of the human motor system in
controlling the amplitude of movement. Journal of Experimental Psychology, 47,
381–391. (reprinted in Journal of Experimental Psychology: General, 121, 1992:
262–269).
Flanagan, J. R. & Johansson, R. S. (2002). Hand movements. Encyclopedia of the
Human Brain, 2, 399-414.
Flanagan, J. R., Vetter, P., Johansson, R. S., & Wolpert, D. M. (2003). Predictions
precedes control in motor learning. Current Biology, 13, 146-150.
Fox, A. M. & Lent, B. (1996). Clumsy children. Primer on developmental
coordination disorder. Canadian Family Physician, 42, 1965-1971.
Frey, S. H. (2007). What puts the how in where? Tool use and the divided visual
streams hypothesis. Cortex. 2007 43, 368-75.
107
Gentile, A. M. (1998). Movement science. Implicit and explicit processes during
acquisition of functional skills. Scandinavian Journal of Occupational Therapy, 5, 7-
16.
Geuze, R. H., Jongmans, M. J., Schoemaker, M. M., & Smits-Engelsman, B. C. M.
(2001). Clinical and research diagnostic criteria for developmental coordination
disorder: A review and discussion. Human Movement Science, 20, 7-47.
Green, D., Chambers, M. E., & Sugden, D. A. (2008). Does subtype of
developmental coordination disorder count: is there a differential effect on outcome
following intervention? Human Movement Science, 27, 363-82.
Guadagnoli, M. A. & Lee, T. D. (2004). Challenge point. A framework for
conceptualizing the effects of various practice conditions in motor learning. Journal
of Motor Behavior, 36, 212-224.
Hammill, D. D., Pearson, N. A., & Voress, J. K. (1993). Developmental Test of Visual
Perception (2nd edition). Austin, TX: Pro Ed.
Hawkins, D. M. (1980). Identification of Outliers. New-York: Chapman and Hall.
Henderson, S. E. & Sugden, D. A (1992). Movement Assessment Battery for
Children. London: Psychological Corporation.
Jansen-Osmann, P., Ritcher, S., Konczak, J., & Kalveram , K.T. (2002). Force
adaptation transfers to untrained workspace regions in children: Evidence for
developing inverse dynamic motor models. Experimental Brain Research, 143, 212-
220.
Jensen, J. L. & Korff, T. (2004). Adapting to changing task demands: Variability in
children’s response to manipulations of resistance and cadence during pedaling.
Research quarterly for exercise and sport, 24, 361-369.
Johnston, L. M., Burns, Y. R., Brauer, S. G., & Richardson, C. A. (2002). Differences
in postural control and movement performance during goal directed reaching in
108
children with developmental coordination disorder. Human Movement Science, 21,
583-601.
Kagerer, F. A., Bo, J., Contreras-Vidal, J. L., & Clark, J. E. (2004). Visuomotor
adaptation in children with developmental coordination disorder. Motor Control, 8,
450-460.
Kagerer, F. A., Contreras-Vidal, J. L., Bo, J., & Clark, J. E. (2006). Abrupt, but not
gradual visuomotor distortion facilitates adaptation in children with developmental
coordination disorder. Human Movement Science, 25, 622-633.
Kagerer, F. A., Contreras-Vidal, J. L., & Stelmach, G. E. (1997). Adaptation to
gradual compared with sudden visuo-motor distortions. Experimental Brain
Research, 115, 557-561.
Kaufman, A. S., & Kaufman, N. L. (1990). Kaufman Brief Intelligence Test manual.
Circle Pines, MN: American Guidance Service.
Kennedy-Behr, A., Rodger, S., & Mickan, S. (2011). Physical and social play of
preschool children with and without coordination difficulties: preliminary findings.
British Journal of Occupational Therapy, 74, 348-354.
King, B. R., Harring, J. R., Oliveira, M. A., Clark, J. E. (2011). Statistically
characterizing intra- and inter-individual variability in children with Developmental
Coordination Disorder. Research in Developmental Disability, 32, 1388-98.
Kirby A. (2011). Dyspraxia series: part one. At sixes and sevens. Journal of Family
Health Care, 21, 29-31.
Magalhães, L.C., Cardoso, A. A., & Missiuna, C. (2011). Activities and participation
in children with developmental coordination disorder: a systematic review. Research
Developmental Disability, 32, 1309-16.
Mandich, A., Polatajko, H. J., Missiuna, C., & Miller, L. (2001). Cognitive strategies
and motor performance in children with developmental coordination disorder.
Physical and Occupational Therapy in Pediatrics, 20, 125-144.
109
Mandich, A. D., Polatajko, H. D., & Rodger, S. (2003). Rites of passage:
Understanding participation of children with developmental coordination disorder.
Human Movement Science, 22, 583-95.
Martin, N. A. (2006). Test of Visual-Motor Skills (3rd edition). Academic Therapy
Publications.
Martin, T. A., Keating, J. G., Goodkin, H. P., Bastian, A. J., & Thach, W. T. (1996).
Throwing while looking through prisms: Focal olivocerebellar lesions impair
adaptation. Brain: A Journal of Neurology, 119, 1183-1198.
Maruff, P., Wilson, P., Trebilcock, M., Currie, J. (1999). Abnormalities of imaged
motor sequences in children with developmental coordination disorder.
Neuropsychologia, 37, 1317-24.
McHale, K., & Cermak, S. A. (1992). Fine motor activities in elementary school:
Preliminary findings and provisional implications for children with fine motor
problems. American Journal of Occupational Therapy, 46, 898-903.
Mehta, B. & Schaal, S. (2002). Forward models in visuomotor control. Journal of
Neurophysiology, 88, 942-953.
Miller, L. T., Missiuna, C. A., Macnab, J. J., Malloy-Miller, T., Polatajko, H. J. (2001).
Clinical description of children with developmental coordination disorder. Canadian
Journal Occupational Therapy, 68, 5-15.
Milner, A. D. & Goodale, M. A. (2006). The Visual Brain in Action (2nd Edition).
Oxford: Oxford University Press.
Missiuna, C. (1994). Motor skill acquisition in children with developmental
coordination disorder. Adapted Physical Activity Quarterly, 11, 214-235.
Missiuna, C., Moll, S., King, G., Stewart, D., MacDonald, K. (2008). Life experiences
of young adults who have coordination difficulties. Canadian Journal of Occupational
Therapy, 75, 157-166.
110
Moser, E. B. (2004). Repeated measures modelling with PROC MIXED. Paper 188–
29, In Proceedings of the 29th Annual SAS Users Group International Conference,
SAS Institute Inc., Cary: NC.
Motulsky, H. J. & Brown, R. E. (2006). Detecting outliers when fitting data with
nonlinear regression – a new method based on robust nonlinear regression and the
false discovery rate. BMC Bioinformatics,7, 123
Newell & Vaillancourt, K. M. & Vaillancourt, D. E. (2001). Dimensional change in
motor learning. Human Movement Science, 20, 695-715.
Pearson, K. (2000). Motor systems. Current Opinion in Neurobiology, 10, 649-654
Polatajko, H. J. (1999). Developmental coordination disorder (DCD): Alias the
clumsy child syndrome. In K. Whitmore, H. Hart, & G. Willems (Eds.), A
Neurodevelopmental Approach to Specific Learning Disorders. Clinics in
Developmental Medicine (pp. 119-133). London: Mac Keith Press.
Polatajko, H. J., Davis, J., Stewart, D., et al. (2007). Specifying the domain of
concern: occupation as core. In: E. A.Townsend, H. J. Polatjako, (Eds). Enabling
Occupation II: Advancing an Occupational Therapy Vision for Health, Well-being, &
Justice through Occupation (pp. 13-36). Ottawa: CAOT Publications ACE.
Polatajko, H. J. & Fox, A. M. (1995). An international consensus on children with
developmental coordination disorder. Canadian Journal of Occupational Therapy,
62, 3-6.
Portney, L. G. & Watkins, M. P. (1993). Foundations of clinical research:
Applications to practice. Norwalk, Connecticut: Appleton & Lange.
Raudenbush, S. W. (1993). Chapter 13. Hierarchical linear models as
generalizations of certain common experimental design models. In L. Edwards (Ed.).
Applied Analysis of Variance in Behavioral Science (pp. 459‐496). New York:
Marcell Decker.
111
Rodger, S., Watter, P., Marinac, J., Woodyatt, G., Ziviani, J., & Ozanne, A. (2007).
Assessment of children with Developmental Coordination Disorder (DCD): motor,
functional, self-efficacy and communication abilities. New Zealand Journal of
Physiotherapy, 35, 99-109.
Rodger, S., Ziviani, J., Watter, P., Ozanne, A., Woodyatt, G., & Springfield, E.
(2003). Motor and functional skills of children with Developmental Coordination
Disorder (DCD): A pilot investigation of measurement issues. Human Movement
Science 22, 461-478.
Rosenblum, S., & Livneh-Zirinski, M. (2008). Handwriting process and product
characteristics of children diagnosed with developmental coordination disorder.
Human Movement Science, 27, 200-214.
Sailer, U., Flanagan, J. R., & Johansson, R. S. (2005). Eye-hand coordination during
learning of a novel visuomotor task. The Journal of Neuroscience, 25, 8833-8842.
Schmidt, R. A. & Lee, T. D. (2011). Motor control and learning: A behavioral
emphasis (5th edition). Champaign, IL: Human Kinetics.
Schmidt, R. A. & Wrisberg, C. A. (2008). Motor learning and performance: A
situation-based learning approach (4th edition). Champaign, IL: Human Kinetics.
Siegler, R. S. (2005). Children's Learning. American Psychologist, 60, 769-778
Singer, J. D. (1998). Using SAS PROC MIXED to fit multilevel models, hierarchical
models, and individual growth models. Journal of Educational and Behavioural
Statistics, 24, 323-355.
Smits-Engelsman, B., Bloem-van der Wel, H. E., & Duysens, J. (2006). Children with
developmental coordination disorder respond similarly to age-matched controls in
both speed and accuracy if goal-directed movements are made across the midline.
Child: Care, Health & Development, 32, 703-710.
112
Smits-Engelsman, B. C. M., Niemeijer, A. S., & van Galen, G. P. (2001). Fine motor
deficiencies in children diagnosed as DCD based on poor grapho-motor ability.
Human Movement Science, 20, 161-182.
Smits-Engelsman, B. C. M., Wilson, P. H., Westenberg, Y., & Duysens, J. (2003).
Fine motor deficiencies in children with developmental coordination disorder and
learning disabilities: An underlying open-loop control deficit. Human Movement
Science, 22, 495-513.
Stephenson, E. A., & Chesson, R. A. (2008). ‘Always the guiding hand’: Parents’
accounts of the long-term implications of developmental co-ordination disorder for
their children and families. Child: Care, Health and Development, 34, 335-343.
Sugden, D. A. (2006). Leeds consensus statement: Developmental coordination
disorder as a specific learning difficulty. Leeds: DCD-UK/Dyscovery Centre.
Summers, J., Larkin, D., & Dewey, D. (2008). Activities of daily living in children with
developmental coordination disorder: Dressing, personal hygiene, and eating skills,
Human Movement Science, 27, 215-229.
Szklut, S. E. & Breath, D. M. (2001). Learning Disabilities. In D. A. Umphred (Ed.),
Neurological Rehabilitation (4th edition) (pp. 308-350). St-Louis, Missouri: Mosby Inc.
Thomas & Thomas, 2008.
Tseng, M., Howe, T., Chuang, I., & Hsieh, C. (2007). Co-occurrence of problems in
activity level, attention, psychosocial adjustment, reading and writing in children with
developmental coordination disorder. International Journal of Rehabilitation
Research, 30, 327-332.
Wang, T. N., Tseng, M. H., Wilson, B. N., Hu, F. C. (2009). Functional performance
of children with developmental coordination disorder at home and at school.
Developmental Medicine & Child Neurology, 51, 817-25.
Wechsler, D. (1991). WISC-III: Wechsler intelligence scale for children. San Antonio,
TX: Psychological Corporation, Harcourt Brace Jovanovich.
113
Wickstrom, R. L. (1983). Fundamental Motor Patterns (3rd edition). Philadelphia: Lea
& Febiger.
Wilson, B. N., Dewey, D., & Campbell, A. (1998). DCD Questionnaire.
Wilson, B., Kaplan, B., Crawford, S., Campbell, A., & Dewey, D. (2000). Reliability
and validity of a parent questionnaire on childhood motor skills. American Journal of
Occupational Therapy, 54, 484-493.
Wilson, P. H., & McKenzie, B. E. (1998). Information processing deficits associated
with developmental coordination disorder: A meta-analysis of research findings. The
Journal of Child Psychology and Psychiatry, 39, 829-840.
Wolpert, D. M. & Flanagan, J. R. (2001). Motor prediction. Current Biology, 18,
R729-R732.
Wolpert, D. M., & Flanagan, J. R. (2010). Primer: Motor learning. Current Biology,
20, R467-R472.
Wolpert, D. M., Ghahramani, Z., & Flanagan, J. R. (2001). Perspectives and
problems in motor learning. Trends in Cognitive Science, 5, 487-494.
Zipp, G. P. & Gentile, A. M. (2010). Practice Schedule and the Learning of Motor
Skills in Children and Adults: Teaching Implications. Journal of College Teaching
and Learning, 7, 35-42.
Zwicker, J. G., Missiuna, C., Harris, S. R., & Boyd, L. A. (2011). Brain activation
associated with motor skill practice in children with Developmental Coordination
Disorder: an fMRI study. International Journal of Developmental Neuroscience, 29,
145-152.
114
APPENDICES
115
APPENDIX A DISTRIBUTION OF M-ABC SCORES FOR ALL CHILDREN (N = 24)
Figure A-1. Distribution of M-ABC Scores across Participants; The top figure shows the distribution of
scores in the sample and the second figure clearly shows the range of scores obtained by participants.
0
5
10
15
20
25
M-A
BC
Sco
res
Participants
Distribution of M-ABC Scores
116
Appendix B Missing Data and Outliers per Trial per Controller
Table B-1. This table presents the percentage of missing data and outliers that occurred during the
simple task for movement time and distance travelled for each trial number.
SIMPLE TASK Trial Number
Movement Time Distance Travelled
% Missing Values % Outliers % Missing Values % Outliers
1 0 0 0 0 2 0 0 0 0 3 0 0 0 0 4 0 0 0 0 5 0 0 0 0 6 0 0 0 0 7 0 0 0 0 8 0 0 0 0 9 0 4 0 0
10 4 0 4 0 11 0 0 4 0 12 0 4 0 0 13 0 0 0 0 14 0 0 4 4 15 0 0 4 17 16 0 0 4 8 17 0 0 0 8 18 0 4 0 8 19 0 4 8 13 20 4 8 0 17 21 38 8 38 8 22 38 4 42 8 23 38 0 38 4 24 38 0 38 0 25 38 0 38 0 26 38 8 38 0 27 38 4 38 0 28 38 4 42 0 29 38 13 42 0 30 38 8 38 4 31 38 13 38 21 32 38 4 38 8 33 38 0 38 13 34 38 0 38 4 35 42 0 38 0 36 38 0 38 4 37 38 4 38 4 38 42 0 42 0 39 38 4 38 8 40 38 0 38 0 41 38 4 38 0 42 38 0 38 4 43 42 0 42 0 44 38 4 38 8 45 38 0 38 0 46 38 0 38 0 47 42 4 42 4 48 38 0 38 4 49 38 4 38 4 50 38 17 38 8
117
Table B-2. This table presents the percentage of missing data and outliers that occurred during the
intermediate task for movement time and distance travelled for each trial number.
INTERMEDIATE TASK Trial Number
Movement Time Distance Travelled % Missing Values % Outliers % Missing Values % Outliers
1 4 0 4 0 2 4 0 4 0 3 4 0 4 0 4 4 0 4 0 5 4 0 4 0 6 4 0 4 0 7 4 0 4 0 8 4 0 4 0 9 4 0 4 0
10 4 0 4 0 11 4 0 4 0 12 4 0 8 0 13 4 0 8 0 14 4 0 4 0 15 4 0 4 0 16 4 0 4 4 17 4 0 4 0 18 4 4 4 13 19 4 4 4 0 20 4 0 8 4 21 4 0 4 4 22 4 0 4 0 23 4 4 4 0 24 4 0 4 0 25 4 0 4 4 26 4 4 4 17 27 4 0 4 0 28 4 0 4 4 29 4 0 4 4 30 8 0 8 4 31 29 0 29 0 32 29 0 29 4 33 29 0 29 0 34 29 0 33 4 35 29 0 33 4 36 29 0 29 0 37 29 0 29 0 38 29 13 29 8 39 29 0 29 4 40 29 4 29 4 41 29 0 42 4 42 29 4 38 0 43 29 8 38 0 44 29 8 42 0 45 29 8 42 4 46 29 4 42 0 47 29 4 38 4 48 29 4 38 13 49 29 4 42 17 50 29 4 42 13
118
Table B-3. This table presents the percentage of missing data and outliers that occurred during the
complex task for movement time and distance travelled for each trial number
COMPLEX TASK Trial Number
Movement Time Distance Travelled % Missing Values % Outliers % Missing Values % Outliers
1 4 0 4 4 2 4 4 8 4 3 8 0 8 0 4 8 0 8 0 5 8 0 8 0 6 4 4 4 0 7 4 0 4 0 8 4 4 4 4 9 4 0 4 0
10 4 0 4 0 11 8 0 8 0 12 4 4 4 0 13 4 0 4 0 14 4 4 4 4 15 4 0 4 4 16 4 21 4 21 17 4 8 4 4 18 4 4 4 8 19 4 4 4 8 20 4 17 4 17 21 8 4 13 4 22 4 13 8 8 23 4 4 4 4 24 4 4 4 0 25 4 4 4 4 26 4 8 4 4 27 4 4 4 0 28 4 0 4 25 29 8 8 8 8 30 13 0 13 0 31 17 0 17 0 32 17 0 17 8 33 21 13 21 13 34 17 0 17 0 35 17 4 17 0 36 17 4 17 4 37 17 4 17 0 38 17 0 17 4 39 21 0 21 0 40 21 4 21 4 41 33 4 33 8 42 38 17 33 13 43 38 0 38 4 44 38 4 38 4 45 33 4 38 13 46 29 4 29 4 47 29 8 29 8 48 29 0 33 4 49 29 0 29 4 50 29 0 33 0
119
APPENDIX C DESCRIPTION OF PARTICIPANTS: DISTRIBUTIONS AND GROUP
DIFFERENCES
Figure C-1 Distribution Graphs and Box Plots of VMI, DCDQ, and M-ABC for children with DCD (group =
1) and TD Children (group = 0). A statistically significant difference between groups was detected on the
three measures.
120
APPENDIX D PARAMETERS ESTIMATES AND STANDARD ERRORS
Table D-1. Effect of Increased Difficulty on Visual-Motor Task Performance over Repeated Trials in TD Children (n = 12)
Source
TD Ln (MT) TD Ln (DIST)
Test value β (SE) p value
Test value β (SE) p value
TRIAL*DIFFs t(1020) = 0.26 < .001 (.002) .793 t(1005) = -.19 < -.001 (.002) .852
TRIAL*DIFFi t(1020) = -1.70 -.004 (.002) .090 t(1005) = -1.04 -.002 (.002) .300
TRIAL*DIFFc t(1020) = -9.87 -.020 (.002) .001* t(1005) = -9.12 -.016 (.002) .001*
Abbreviations: DIFF (s, i, c) = simple, intermediate, complex tasks; MT: movement time; DIST: distance travelled; M-ABC: motor ability; Ln: natural log transformation applied to data;*statistically significant at α = .05. Table D-2. Effect of Motor Ability on Initial Visual-Motor Performance (N = 24) when task difficulty is also considered
Source
Ln (MTi) Ln (Disti)
Test value β (SE) p value
Test value β (SE) p value
M-ABC*DIFFs t(42) = 0.31 .004 (.013) .758 t(42) = -.19 < -.001 (.002) p = .683
M-ABC*DIFFi t(42) = .60 .009 (.015) .551 t(42) = -1.04 -.002 (.002) p = .958
M-ABC*DIFFc t(42) = 3.46 .051 (.015) .001* t(42) = -9.12 -.016 (.002) p < .001*
Abbreviations: DIFF (s, i, c) = simple, intermediate, complex tasks; MT: movement time; DIST: distance travelled; M-ABC: motor ability; Ln: natural log transformation applied to data;*statistically significant at α = .05.
Table D-3. Effect of Motor Ability on Visual-Motor Task Performance over Repeated Trials (N = 24)
Source Ln (MTi) Ln (Disti)
Test value β (SE) p value Test value β (SE) p value
Simple Task TRIAL
t(977) = -2.98
-.006 (.002)
.003*
t(972) = -3.23
-.002 (<.001)
.001*
AGE t(21) = -.48 -.002 (.004) .638 t(21) = -.35 <-.001 (.001). .730
M-ABC t(21) = 1.92 .004 (.013) .069 t(21) = 2.82 007 (.002) .010*
Intermediate Task
TRIAL t(1155) = -2.72 -.004 (.002) .007* t(1145) = -1.79 -.003 (.001) .074
AGE t(21) = -1.46 -.005 (.003) .156 t(21) = -1.11 <-.001 (.001) .280
M-ABC t(21) = 1.41 .010 (.007) .174 t(21) = 1.53 .007 (.004) .142
Complex Task
TRIAL t(1182) = -8.33 -.021 (.002) <.001* t(1187) = -6.70 -.015 (.002) <.001*
AGE t(21) = -3.82 -.014 (.004) .001* t(21) = -3.48 -.012 (.004) .002*
M-ABC t(21) = 1.70 .013 (.008) .028* t(21) = 2.18 .016 (.007) .041*
Abbreviations: MT: movement time; DIST: distance travelled; DIFF: Task difficulty; TRIAL: Trial number; M-ABC:
motor ability; Ln: natural log transformation applied to the data; *: statistically significant at α = .05.
121
APPENDIX E COMPARING CHANGES IN VISUAL-MOTOR PERFORMANCE OF
CHILDREN WITH DCD (N = 12) TO TD CHILDREN DURING COMPLEX TASK
Figure E-1 Comparing Patterns of Change in Visual-Motor Performance of Children with DCD and TD
-
200
400
600
800
1,000
1,200
1,400
1,600
0 10 20 30 40 50
Dis
t (m
m)
Trial Number
Comparing TD-DCD Patterns of Change in Dist. during Complex Task
TD Dist
DCD Dist
0.00
10.00
20.00
30.00
40.00
50.00
60.00
0 10 20 30 40 50
MT
(s)
Trial Number
Comparing TD-DCD Visual Analysis of Patterns of Change in MT during Complex Task
MT
TD MT
122
APPENDIX F METHODS FOR CALCULATING SPECIAL CAUSES
(adapted from Callahan & Barisa, 2005, p.33)
To calculate upper (UCL) and lower control limits (LCL), compute:
1. Moving range: difference between sequential observations on dependent
variable
2. Xbar: arithmetic mean
3. Rbar: moving range mean, which is an average of the moving range scores
Then compute UCL and LCL
UCL = Xbar + (3 * Rbar)
LCL = Xbar – (3 * Rbar)
Finally, compute + 1 and + 2 sigma
+ 1 sigma = Xbar + (1 * Rbar)
+ 2 sigma = Xbar + (2 * Rbar)
123
0
1
2
3
4
5
0 5 10 15 20 25 30 35 40 45 50
MT
(s)
Trial Number
Changes in MT (TD) over Repeated Trials for Simple Task
Predicted MT(TD)
Actual MT(TD)
0
50
100
150
200
250
300
350
400
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Changes in DIST (TD) over Repeated Trials for Simple Task
Predicted DIST(TD)
Actual DIST(TD)
0
1
2
3
4
5
0 5 10 15 20 25 30 35 40 45 50
MT
(s)
Trial Number
Changes in MT (N=24) over Repeated Trials for Simple Task
Predicted MT(N=24)
Actual MT(N=24)
0
50
100
150
200
250
300
350
400
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Changes in DIST (N=24) over Repeated Trials for Simple Task
Predicted DIST(N=24)
Actual DIST(N=24)
APPENDIX G REGRESSION CURVES OF TD CHILDREN ON TASKS FOR WHICH
TRIAL WAS NOT A MODULATING VARIABLE
Figure G-2. Illustrating Changes in Visual-Motor Performance over Repeated Trials during the Simple
Task in All children (N = 24) and TD Children (n = 12)
0
1
2
3
4
5
0 5 10 15 20 25 30 35 40 45 50
MT
(s)
Trial Number
Changes in MT (N=24) over Repeated Trials for Intermediate Task
Predicted MT(N=24)
Actual MT (N=24)
0
50
100
150
200
250
300
350
400
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Changes in DIST (N=24) over Repeated Trials for Intermediate Task
Predicted DIST(N=24)
Actual DIST(N=24)
0
1
2
3
4
5
0 5 10 15 20 25 30 35 40 45 50
MT
(s)
Trial Number
Changes in MT (TD) over Repeated Trials for Intermediate Task
Predicted MT(TD)
Actual MT(TD)
0
50
100
150
200
250
300
350
400
0 5 10 15 20 25 30 35 40 45 50
DIS
T (m
m)
Trial Number
Changes in DIST (TD) over Repeated Trials for Intermediate Task
Predicted DIST(TD)
Actual DIST(TD)
Figure G-1 Illustrating Changes in Visual-Motor Performance over Repeated Trials during the Simple
Task in All children (N = 24) and TD Children (n = 12).
