Upload
amar-ram
View
212
Download
0
Embed Size (px)
Citation preview
8/7/2019 Copy of v Research Design
1/45
V Research Design
8/7/2019 Copy of v Research Design
2/45
What is research design?
Blueprint for collection, measurement and
analysis of data.
It is the conceptual structurew
ithinw
hichresearch is conducted ; it comprises the planfor collection, measurement & analysis ofdata.
It is logical and systematic plan prepared fordirecting research study.
8/7/2019 Copy of v Research Design
3/45
Definition of research design
Research design is an aid for researcher in the
allocation of limited resources by posing
crucial choices in methodology.
It is procedural plan that is adopted by the
researcher to answer questions validly,
objectively, accurately and economically.
8/7/2019 Copy of v Research Design
4/45
What is Research Design
What is essential
Activity & time based plan
Based on research problem / objectives
Guide for selecting sources and type of information
Procedural outlines
Framework for specifying relationship amongst study
variable
8/7/2019 Copy of v Research Design
5/45
Components of research design
Sampling design
Deals with the method of selecting items to be observed for the givenstudy.
Observational Design
Relates to the condition under which observations are to be made.
Statistical Design Concerns with the questions of how many items are to be observed
(sample size) and
how information and data gathered are to be analysed
Operational Design Deals with implementation part.
How the techniques and procedures specified will carried out.
8/7/2019 Copy of v Research Design
6/45
Features of good design
Minimise bias and maximises the reliability of the
data collected and analysed.
Yields maximal information and provides an
opportunity for considering many different aspect of
the problem.
Design is related with purpose and objective of the
study.
Flexible, appropriate, efficient, economical.
8/7/2019 Copy of v Research Design
7/45
Types of research Design
Research Design for exploratory research
Research Design for Conclusion oriented research Research Design for Descriptive research
Cross sectional
Longitudinal
Research Design for Causal research Pre experiment
True Experiment
Quasi Experiment
Complex designs Simple factorial
Complex factorial
8/7/2019 Copy of v Research Design
8/45
Exploratory Vs Descriptive research Designs
Exploratory /
Formulative
Descriptive /
Diagnostic
Overall
Design
Flexible (opportunity to
explore the different aspectsof problem)
Rigid (maximize reliability of
the outcome and reducesbias)
Sampling
Design
Non probability Probability
Statistical
Design
No preplanned Design for
analysis
Pre planned design for
analysis
Observation
al Design
Unstructured Instruments
used for data collection
Structured and well thought
and designed instruments for
data collection
Operational
Design
No fixed decision regarding
operational procedures
Advanced decision regarding
operational procedures
8/7/2019 Copy of v Research Design
9/45
Experimental Method
8/7/2019 Copy of v Research Design
10/45
What is experimentation?
Studies involving
intervention (to manipulate some variable)by researcher
8/7/2019 Copy of v Research Design
11/45
Advantages
Manipulation
Control
Convenience
Cost
Replication
8/7/2019 Copy of v Research Design
12/45
Disadvantage
Setting up the laboratory
Chances of cost overruns
Past things can not be studied
Ethicality (as social research is related with
people extent to which one can
manipulate??)
8/7/2019 Copy of v Research Design
13/45
Experimental Designs
Examines differences betweenexperimentally manipulated groups orvariables (e.g., one group gets a certain drugand the other gets a placebo).
At minimum, experimental (independent)variable has two levels (e.g., drug vs.placebo).
Advantage is that you can determine causality. Disadvantage is cost and many variables cannot
be experimentally manipulated (e.g., smokeexposure over time).
8/7/2019 Copy of v Research Design
14/45
Experimental Designs
Four Canons for Identifying Causality
Method of Agreement
Observe the element common to several
instances of the event Problem is you may inadvertently overlook a
significant variable.
Method of Difference
Identify the different effects produced by twosituations that are alike in all ways but one.
Fairly robust and strong method.
8/7/2019 Copy of v Research Design
15/45
Experimental Designs
Four Canons for Identifying Causality
Joint methods of agreement and difference
Observe the element common to several instances of theevent
Form hypothesis based on observations
Test hypothesis using method of difference
Method ofConcomitant Variation
Identify the different effects produced by more than twosituations that are alike in all ways but one.
E.g., Compare two active drugs to a placebo
8/7/2019 Copy of v Research Design
16/45
Experimental Design
Because it is so difficult with human behavior
to demonstrate causation unequivocally, some
argue that a theory or prediction can onlyachieve the status of not yet disconfirmed
(Popper, 1968).
Our scientific efforts are directed at finding thecausal factors rather than the cause per se.
8/7/2019 Copy of v Research Design
17/45
Psychological Experiment:
Is Objective
Researcher strives for freedom from bias.
Recognize that:
Mistakes can occur
Carefully scrutinize all steps of the experiment to
identifywhere mistakes are likely.
Take the steps necessary to minimize error.
8/7/2019 Copy of v Research Design
18/45
Psychological Experiment:
Focuses on a Phenomenon
This is a publicly observable behaviour.
Actions
Appearances Verbal statements
Responses to questionnaires
Physiological responses.
8/7/2019 Copy of v Research Design
19/45
Psychological Experiment:
Is Done Under Strictly Controlled
Conditions
Eliminate all factors that could influence
the outcome other than the factor being
manipulated.
Control is needed to infer causation.
All conditions are kept constant except one;
the manipulated variable.
The variable of interest is varied in order to
test its effect.
8/7/2019 Copy of v Research Design
20/45
Experimental Method
Advantages
Strength with which causal relationships can be
inferred.
Ability to manipulate one or more variables.
Proven to be a very useful and robust scientific
method (i.e., withstood the test of time).
8/7/2019 Copy of v Research Design
21/45
Experimental Method
Disadvantages
Tight controls often produce artificial
conditions that could limit the generalizabilityof the findings (i.e., internal vs. external validity
trade-off).
Time consuming.
Expensive.
Human behavior is very complex and cannot be
fully studied using experimental methods.
8/7/2019 Copy of v Research Design
22/45
Experimental Method:
Threats to Internal Validity Learning or practice effects
Scores on a measure change on repeat testing becauseparticipant has more familiarity with the measure andso answers more truthfully.
Natural history effects Something happens in the social background (e.g.,
society because more affluent generally) and thisinfluences the participants responses.
Maturation Natural developments in the participant account for
the changes (e.g., getting older).
8/7/2019 Copy of v Research Design
23/45
Experimental Method:
Threats to Internal Validity
Regression to the mean
High scores generally move down toward the
mean and low scores move up.
Instrumentation
If pre and post tests are not equivalent in all
ways (e.g., difficulty, readability) then
differences observed may be due toinstrumentation differences rather than due
to your experimental manipulation.
8/7/2019 Copy of v Research Design
24/45
Experimental Method:
Threats to Internal Validity
Subject problems
Selection bias (e.g., participation rate).
Attrition (e.g., only motivated subjects stay in
the experiment).
8/7/2019 Copy of v Research Design
25/45
Experimental Method:
Threats to External Validity Subject variables
Selection bias.
Attrition bias
Artificial conditions E.g., In order to measure a subjects blood pressure in
response to a well-fined stressor you bring him/her
into the laboratory but his/her response in the
laboratory may not reflect how his/her blood pressure
would really respond under stress in his natural
environment.
8/7/2019 Copy of v Research Design
26/45
8/7/2019 Copy of v Research Design
27/45
The independent variable is the variable that is
purposely changed. It is the manipulatedvariable.
The dependent variable changes in response to
the independent variable. It is the responding
variable.
Variables
Variables are things that change.
8/7/2019 Copy of v Research Design
28/45
Constants or Controlled Variables
in an ExperimentWhat are constants in an experiment?
Factors that are kept the same and not
allowed to change.
It is important to control all but onevariable at a time to be able to
interpret data
8/7/2019 Copy of v Research Design
29/45
Sample experiment with data
Studentswere asked to build a helicopter and
test something.
8/7/2019 Copy of v Research Design
30/45
Writing A Statement of the
Problem for the Experiment
It should state: The Effect of the Independent
Variable on the Dependent Variable.
A good statement of the problem for The
Helicopter would be
The Effect of wing length on time of decent
What should it state?
8/7/2019 Copy of v Research Design
31/45
More about our variables:
The independent variable is the variable that is
purposely changed. It is the manipulated variable.
The dependent variable changes in response to the
independent variable. It is the responding variable.
What is the Helicopter independent variable?
length of one wing as measured in centimeters
What is the dependent variable?
time of descent as measured in seconds
8/7/2019 Copy of v Research Design
32/45
Title for the Helicopter Experiment
The Effect of Wing Length on the Length
of Time the Helicopter Takes to Reach
the Ground
8/7/2019 Copy of v Research Design
33/45
Another Look
at Our Hypothesis
INDEPENDENT VARIABLE
If the length of the wings are increased,
DEPENDENT VARIABLE
then time of descent will increase.
8/7/2019 Copy of v Research Design
34/45
Constants in an ExperimentFactors that are kept the same and not
allowed to change
What should be kept constant in theHelicopter?
CONSTANTS
Same structure of helicopters
Same dropping technique
Same dropping height
8/7/2019 Copy of v Research Design
35/45
Values of the Independent Variable
How many different values of the
independent variable should we test?
Enough to clearly illustrate a trend in
the data - 5cm, 6cm, 7cm, 8cm..
8/7/2019 Copy of v Research Design
36/45
Repeated Trials
How many repeated trials are needed in the
Helicopter?
At least 3, then calculate an average value
for each level of the independent variable
8/7/2019 Copy of v Research Design
37/45
Drawing of Experiment
Hold helicopter
in the middle
Drop from
same spot
each time
8/7/2019 Copy of v Research Design
38/45
QualitativeObservations and Results
As we dropped the helicopter, we
observed that sometimes they
seemed to be affected by aircurrents.
It appeared that helicopters with
longer wingspans took a greater
time to reach the ground
8/7/2019 Copy of v Research Design
39/45
Quantitative Observations
Data Table, Graph and
Statistics
8/7/2019 Copy of v Research Design
40/45
Sample Data Table
Title: The Effect ofwing length on the time of descent
Wing Length
(cm)
Average Descent
Time (s)
Descent Time (s)
Trials
1 2 3
56
7
89
3 5 4 44 5 6 5
6 7 5 6
7 8 6 79 7 8 8
8/7/2019 Copy of v Research Design
41/45
The Effect f Wi Length n Descent Time
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10
Average Wing Length (cm)
AverageDescen
tTime(s)
8/7/2019 Copy of v Research Design
42/45
Analysis and Interpretation of
ResultsWhile individual trial results showed some
variation; overall, as average wing length
increased from 5 through 9 centimeters,
average time of descent also increasedfrom 4 through 8 seconds.
This shows a direct correlation.
8/7/2019 Copy of v Research Design
43/45
Possible Experimental Errors
Although the average results supported the
hypothesis, the individual trial results showedsome variation. Some possible reasons for this
could be due to air currents in the room or
reaction time differences between the timer
and the helicopter dropper.
8/7/2019 Copy of v Research Design
44/45
Conclusion
The overall results supported the
hypothesis.
A possible reason for this could be
due to increased surface area of the
wing in contact with the air - allowing
for greater air support.
8/7/2019 Copy of v Research Design
45/45
Recommendations for Further Experimentation/
Practical Applications
Increased time of descent due to greater wing length
suggests that increased wing length could lead to
greater fuel efficiency of an actual helicopter.
Other experiments investigating factors that would
contribute to greater descent time might include:
making modifications in the shape of the wings,
changing the number of wing blades, and making foldsin the wing blades