13
JOURNAL OF LEARNING DISABILITIES VOLUME 40, NUMBER 3, MAY/JUNE 2007, PAGES 244–255 Expressive Writing Difficulties in Children Described as Exhibiting ADHD Symptoms Anna Maria Re, Martina Pedron, and Cesare Cornoldi Abstract Three groups of children of different ages who were considered by their teachers as showing symptoms of attention-deficit/hyperactivity disorder (ADHD) and matched controls were tested in a series of expressive writing tasks, derived from a standardized writing test. In the first study, 24 sixth- and seventh-grade children with ADHD symptoms wrote a description of an image. The ADHD group’s ex- pressive writing was worse than that of the control group and associated with a higher number of errors, mainly concerning accents and geminates. The second study showed the generality of the effect by testing younger groups of children with ADHD symptoms and con- trols with another description task where a verbal description was substituted for the picture stimulus. The third study extended the pre- vious observations with another type of writing task, the request of writing a narrative text. In all the three studies, children with ADHD symptoms scored lower than controls on four qualitative parameters (adequacy, structure, grammar, and lexicon), produced shorter texts, and made more errors. These studies show that children with ADHD symptoms have school difficulties also in writing—both in spelling and expression—and that these difficulties are extended to different tasks and ages. N umerous studies have shown that children described by their teachers as showing attention- deficit/hyperactivity disorder (ADHD) symptoms also present academic diffi- culties in a variety of fields. Until re- cently, the emphasis has been on the more basic skills, such as reading (Bo- nafia, Newcorn, McKay, Koda, & Hal- perin, 2000; Johnson, Altmaier, & Rich- man, 1999; Pisecco, Baker, Silva, & Brooke, 2001; Swanson, Mink, & Bo- cian, 1999; Willcutt & Pennington, 2000) and arithmetic (Benedetto-Nasho & Tannock, 1999; Shimabukuro, Prater, Jenkins, & Edelen-Smith, 1999). Given the planning and organizational diffi- culties shown by children with ADHD, one would expect them to have equal if not greater difficulties in skills re- quiring greater organization and con- trol, for example when having to com- pose a short essay. Nonetheless, and in contrast with the large literature on the writing prob- lems of children with learning disabil- ities (LD; Englert & Raphael, 1988; Gra- ham & Harris, 1989; Graham, Harris, MacArthur, & Schwartz, 1991; Mon- tague, Graves, & Leavell, 1991; New- comer & Barenbaum, 1991), much less is known about the writing skills of children with ADHD. We found only two studies that specifically examined writing skills in children with ADHD. One study focused on general writing skills, the other on speed of writing. In the first study, as part of a writing test, Resta and Eliot (1994) examined the performance of 32 students between 8 and 13 years old (Grill & Kirwin, 1989). The study included 21 boys with ADHD and 11 matched controls, who were required to write three essays about an expressive (about hands), a creative (in response to a picture of a cat), and an instructive (the danger of fire) topic. The results showed that children with ADHD performed worse and were less productive, in that they wrote fewer words than control children. In the second study, Ross, Poide- vant, and Miner (1995) assessed the writing speed of 48 children with ADHD, from first to fifth grades, and 48 controls matched for gender and schooling. The task consisted of writ- ing, as rapidly as possible, the num- bers from 0 to 9 and the student’s first name for 1 min. The results showed no difference between the two groups. The pattern of these results suggests that children with ADHD write less be- cause they have difficulties in the text production process and not because of fluency or writing speed problems. However, these data, coming from dif- ferent populations and different stud- ies, do not allow clear conclusions and generalizations to be drawn. Further- more, these studies did not systemati- cally examine different aspects of ex- pressive writing or spelling errors. Yet the study of different performance parameters in expressive writing may be important, because it could reveal specific peculiarities of children with

Expressive Writing Difficulties in Children Described as ... · PDF fileExpressive Writing Difficulties in Children Described as Exhibiting ADHD Symptoms Anna Maria Re, Martina Pedron,

  • Upload
    trandat

  • View
    215

  • Download
    0

Embed Size (px)

Citation preview

JOURNAL OF LEARNING DISABILITIESVOLUME 40, NUMBER 3, MAY/JUNE 2007, PAGES 244–255

Expressive Writing Difficulties inChildren Described as ExhibitingADHD Symptoms

Anna Maria Re, Martina Pedron, and Cesare Cornoldi

Abstract

Three groups of children of different ages who were considered by their teachers as showing symptoms of attention-deficit/hyperactivitydisorder (ADHD) and matched controls were tested in a series of expressive writing tasks, derived from a standardized writing test. Inthe first study, 24 sixth- and seventh-grade children with ADHD symptoms wrote a description of an image. The ADHD group’s ex-pressive writing was worse than that of the control group and associated with a higher number of errors, mainly concerning accents andgeminates. The second study showed the generality of the effect by testing younger groups of children with ADHD symptoms and con-trols with another description task where a verbal description was substituted for the picture stimulus. The third study extended the pre-vious observations with another type of writing task, the request of writing a narrative text. In all the three studies, children with ADHDsymptoms scored lower than controls on four qualitative parameters (adequacy, structure, grammar, and lexicon), produced shorter texts,and made more errors. These studies show that children with ADHD symptoms have school difficulties also in writing—both in spellingand expression—and that these difficulties are extended to different tasks and ages.

Numerous studies have shownthat children described by theirteachers as showing attention-

deficit/hyperactivity disorder (ADHD)symptoms also present academic diffi-culties in a variety of fields. Until re-cently, the emphasis has been on themore basic skills, such as reading (Bo-nafia, Newcorn, McKay, Koda, & Hal-perin, 2000; Johnson, Altmaier, & Rich-man, 1999; Pisecco, Baker, Silva, &Brooke, 2001; Swanson, Mink, & Bo-cian, 1999; Willcutt & Pennington, 2000)and arithmetic (Benedetto-Nasho &Tannock, 1999; Shimabukuro, Prater,Jenkins, & Edelen-Smith, 1999). Giventhe planning and organizational diffi-culties shown by children with ADHD,one would expect them to have equalif not greater difficulties in skills re-quiring greater organization and con-trol, for example when having to com-pose a short essay.

Nonetheless, and in contrast withthe large literature on the writing prob-lems of children with learning disabil-

ities (LD; Englert & Raphael, 1988; Gra-ham & Harris, 1989; Graham, Harris,MacArthur, & Schwartz, 1991; Mon-tague, Graves, & Leavell, 1991; New-comer & Barenbaum, 1991), much lessis known about the writing skills ofchildren with ADHD. We found onlytwo studies that specifically examinedwriting skills in children with ADHD.One study focused on general writingskills, the other on speed of writing. Inthe first study, as part of a writing test,Resta and Eliot (1994) examined theperformance of 32 students between 8and 13 years old (Grill & Kirwin, 1989).The study included 21 boys withADHD and 11 matched controls, whowere required to write three essaysabout an expressive (about hands), acreative (in response to a picture of a cat), and an instructive (the danger of fire) topic. The results showed thatchildren with ADHD performed worseand were less productive, in that they wrote fewer words than controlchildren.

In the second study, Ross, Poide-vant, and Miner (1995) assessed thewriting speed of 48 children withADHD, from first to fifth grades, and48 controls matched for gender andschooling. The task consisted of writ-ing, as rapidly as possible, the num-bers from 0 to 9 and the student’s firstname for 1 min. The results showed nodifference between the two groups.The pattern of these results suggeststhat children with ADHD write less be-cause they have difficulties in the textproduction process and not because offluency or writing speed problems.However, these data, coming from dif-ferent populations and different stud-ies, do not allow clear conclusions andgeneralizations to be drawn. Further-more, these studies did not systemati-cally examine different aspects of ex-pressive writing or spelling errors. Yetthe study of different performanceparameters in expressive writing may be important, because it could revealspecific peculiarities of children with

VOLUME 40, NUMBER 3, MAY/JUNE 2007 245

ADHD, like those that have emergedin other learning areas (e.g., problemsolving; Marzocchi, Lucangeli, DeMeo, Fini, & Cornoldi, 2002). Writing isnot a simple transcription of thoughtsand concepts; it requires the involve-ment of a high degree of complexcognitive procedures. For example,classical views of expressive writing(Burnett & Kastman, 1997; Hayes &Flower, 1980) showed the importanceof several processes, and in particularthe planning phase of an essay, the pro-duction of ideas, their organization,the transcription, and the final revi-sion. The planning problems of chil-dren with ADHD have been docu-mented in the literature (Barkley, 1995;Cornoldi, Barbieri, Gaiani, & Zocchi,1999; Seidman, Biederman, Monu-teaux, Doyle, & Faraone, 2001); it isthus highly probable that these chil-dren also have difficulties in expres-sive writing. For example, one couldpredict that children with ADHD canproduce many ideas, but—given theirorganizational and planning difficul-ties—they do poorly at writing a textthat is adequate and well organized(i.e., adequately matching the writingrequests both in terms of the quantityof expressed ideas and their organiza-tion). Thus, despite the fact that theirexpression difficulty should mainlyconcern the organization of ideas, thepoorer text organization and planningof children with ADHD could also havethe consequence of limiting the num-ber of ideas actually being expressed.

Another problem in text produc-tion concerns spelling: Children withADHD could be less capable of simul-taneously paying attention to theirideas and to spelling and, therefore,make more mistakes. In fact, the pres-ence of a larger number of spelling er-rors in the written text of children withADHD can also be predicted on thebasis of their frequently observed read-ing and phonological difficulties (e.g.,Kroese, Hynd, Knight, Hiemenz, &Hall, 2000). Nevertheless, it is not clearwhether a hypothesized spelling diffi-culty in children with ADHD equallyextends to any type of error or mainly

concerns a particular error type. A suc-cessful classification of writing errorsis based on a basic model for learningto read and write (Frith, 1985) that pro-poses a series of learning stages, someof which are associated with specifictypes of spelling errors. In the firststage, called logographic, a child canonly associate a particular graphic con-figuration with a certain concept. Inthe second stage, called alphabetic, achild discovers the concept of pho-nemes and learns to associate everyphoneme with its peculiar graphic signpattern. In this stage, errors are of thephonological type, due to the incorrectassociation between a grapheme andthe corresponding phoneme. In thethird stage, called orthographic, a childlearns that writing is governed by syn-tactic and orthographic rules and nolonger works with phonemes but withsyllables or other sublexical units, sothat the writing process becomes morerapid and correct. In the fourth stage,children learn specific lexical entriesand are able to read and write wordsthat do not follow the phonologicalrules (e.g., typical writing errors in Ital-ian are represented by illegal fusionsand separations). Finally, Tressoldi andCornoldi (1991) described a fifth stage,requiring a lexicon-based refinementof lexical and phonological analysis, inwhich the main difficulties are repre-sented by the correct use of last sylla-ble accents and geminates. Evidencefor a specific writing disorder and aconsequent autonomous representa-tion of geminates in Italian has alsobeen reported by Miceli, Benvegnù,Capasso, and Caramazza (1995). Atthis stage, the selection of phonemesand graphemes may be appropriate,but may miss the details regardinglonger phoneme duration (which inItalian requires the writing of doubleletters: e.g., correre, to run) or of an ac-cent (which in the Italian writing sys-tem is required only when the accent ison the last syllable).

The present research examinedthe performance of children who weredescribed by their teachers as showingADHD symptoms (see Note 1) on writ-

ing tasks derived from the Batteria perla Valutazione della Scrittura e della Com-petenza Ortografica nella Scuola dell’Ob-bligo (BVSCO; Battery for the Assessmentof Writing Skills in 7- to 13-Year-OldChildren), devised by Tressoldi andCornoldi (1991), which is the only stan-dardized writing test available in Italy.Because the BVSCO offers an overallview of expressive writing competen-cies, the first goal of the present re-search was to obtain a general descrip-tion of the writing skills of childrenwith ADHD symptoms. Concerningthe children’s writing abilities we fo-cused on the four main aspects consid-ered in the BVSCO (adequacy, struc-ture, lexicon, and grammar) and onsome other indices as recommendedby the BVSCO’s authors. Furthermore,the BVSCO also offered the opportu-nity of measuring writing speed andaccuracy. Concerning writing speed,we predicted that, as Ross et al. (1995)already found, children with ADHDsymptoms would be as rapid as con-trols. Concerning accuracy, on the con-trary, we predicted that children withADHD symptoms would show a poorerperformance due to their spelling diffi-culties (Kroese et al., 2000).

The second goal of the study con-cerned spelling accuracy in expressivewriting. In particular, and followingthe indication of the BVSCO manual,we considered three types of errors:phonological, nonphonological, and athird type (geminates and accents; seeNote 2). Because these types of errorsreflect different stages in learning towrite correctly, we expected that thelower competence of older childrenwould show itself in the form of third-type errors, whereas for younger chil-dren who were still involved in thepreceding learning stages, the differ-ences would be found at earlier levels(Tressoldi & Cornoldi, 1991).

In the present research, three dif-ferent studies examined these issueswith their specific associated goals. Inthe first study, sixth- and seventh-grade children with ADHD symptomsand matched controls were engaged inan expressive writing task, in which

JOURNAL OF LEARNING DISABILITIES246

they had to describe an image, and ina speed writing task, in which they hadto write as many numbers in letters aspossible in one minute. The secondstudy investigated the generality of theeffect and its developmental pattern bytesting younger groups of childrenwith ADHD symptoms and controls,using the same task as in Study 1 andanother task in which the picture wasreplaced by a verbal description. Infact, it has been suggested that not onlyimages, but also verbal prompts canimprove the quality of expressive writ-ing (Marchisan & Alber, 2001). Finally,the third study examined the generalwriting skills with respect to anothertype of writing request, by asking chil-dren to write a narrative text.

STUDY 1

Study 1 examined the expressive writ-ing, spelling, and writing speed abili-ties of a group of children described bytheir teachers as exhibiting ADHDsymptoms and of a matched controlgroup.

Method

Participants

Two groups of sixth- and seventh-grade children were selected: 24 chil-dren (17 boys and 7 girls) described bytheir teachers as exhibiting ADHDsymptoms, and a comparison group of24 children (17 boys and 7 girls; controlchildren) of the same age and school-ing. Participants were selected frompublic schools in the Catania area ofSicily, Italy, on the basis of the teacher’srating scale, Scala per i Disturbi diAttenzione/Iperattività per Insegnanti(SDAI; ADHD Rating Scale for Teachers;Marzocchi & Cornoldi, 2000, whichwas used to assess 817 children. Theschools were located in the environs ofthe city of Catania, which are predom-inantly inhabited by White familiesworking in the public sector or en-gaged in agricultural activities. All

children came from families with Ital-ian as their first language.

The SDAI includes 18 items, basedon the 18 ADHD symptoms listed inthe fourth edition of the Diagnostic andStatistical Manual of Mental Disorders(DSM-IV; American Psychiatric Asso-ciation, 1994), and has been validatedand standardized for the Italian popu-lation, showing very high validity, reli-ability (r = .81), and interrater agree-ment (r = .78; Marzocchi & Cornoldi,2000). The scale includes two sub-scales, one for Inattention and theother for Hyperactivity/Impulsivity.Teachers have to observe the children’sbehavior and report the frequency ofsymptomatic behaviors described foreach item. Scores range from 0 (whenthe problematic behavior is never pres-ent) to 3 (for very frequent behavior).Children whose mean score was above1.5 per item on one of the two subscaleswere considered for inclusion in thegroup of children with ADHD symp-toms (selected children were mainly ofthe inattentive or combined DSM-IVsubtypes), and those who did not meetthis criterion were considered for in-clusion in the control group.

Teachers were also asked to ratesocioeconomic status (SES) and oppo-sitional and aggressive behavior. In thegroup with children showing ADHDsymptoms, we included only childrenwho did not manifest oppositional andaggressive behavior. Control childrenwere selected from the same classes asthe children with ADHD symptoms, soas to be matched for gender, age,schooling, estimated IQ, rated ability,and SES. The IQ estimation was ob-tained by administering part of the Pri-mary Mental Abilities Battery 11–17(PMA) test. The PMA Reasoning sub-scale is part of the classical battery de-vised by Thurstone and Thurstone(1963/1981) on the basic componentsof intelligence, which is still in use inItaly, and requires finding, within 6minutes, the logical criterion in a se-quence of letters (30 items). For exam-ple, the first item presents the letter se-ries a a b c c d e e f g g, and requires thatthe child select between the following

possible choices to continue the series:a b c f g h. The Italian norms of the testinclude a transformation table, whichpermits the translation of the rawscores into IQ scores. All children in-cluded in the study scored above the40th percentile on the PMA Reasoningtask, roughly corresponding to an IQhigher than the 95th percentile. Finally,children with ADHD symptoms scoredbelow the 30th percentile in an atten-tional visual search task, the Contin-uous Performance (CP) task (Cornoldi,Gardinale, Masi, & Pettenò, 1996), whichrequires finding a sequence of threeletters (F Z B) presented 54 timeswithin three random arrays of letters.

Procedure

Children were administered the De-scription test from the BVSCO (Tres-soldi & Cornoldi, 1991) in their class-room. This task is based on thepresentation of a colored figure show-ing people at the zoo, with a variety ofanimals in their surroundings. Chil-dren receive a copy of the figure and aresponse sheet corresponding to thetype of paper on which they are usedto writing. They have 10 min to write atext on the basis of the following in-struction: “Imagine that you have beenat the zoo and you have to describe thescene you see in the picture to somefriends who were not there.” As a con-trol, children were administered awriting speed subtest taken from thesame BVSCO. This test requires writ-ing, in letters, as many numbers as pos-sible in 1 min, starting from uno (one)and proceeding with the subsequentnumbers of the number series. Therewas no significant difference betweenthe two groups’ performance on thistask.

Results

Children were able to understand theinstructions and meet task requests.The available time of 10 min was suffi-cient for all children, and many chil-dren in both groups finished with a

VOLUME 40, NUMBER 3, MAY/JUNE 2007 247

few minutes to spare. Because childrenwere instructed to keep the responsesheet until the end of the availabletime, so they could check over their an-swers, it was not possible to record theactual time used by each child to com-plete the task. The protocols were ana-lyzed for the properties of expressivewriting, following the procedure de-scribed in the test manual. First, twoindependent raters, blind with respectto the goals of the study, rated the pro-tocols for the following four qualitativeparameters:

1. adequacy, defined as the adequacyof the written text with respect tothe task request;

2. structure, based on the organizationof the text;

3. grammar, concerning the correctuse of punctuation, subdivision inparagraphs, correct use of verbtenses, and correct concordancebetween gender and number ofnouns, verbs, and adjectives (veryimportant in Italian); and

4. lexicon, defined as the quantity ofdifferent words used.

The agreement of the two raterswas high for each of the four parame-ters (i.e., adequacy, r = .89; structure, r = .84; lexicon, r = .83; and grammar, r = .78). Therefore, in the subsequentanalyses, we only considered the rat-ings given by the first rater.

Children with ADHD symptomsobtained significantly lower scoresthan controls on adequacy (ADHD, M = 2.62, SD = 0.78; controls, M = 3.92,SD = 0.87, t(46) = 5.41, p < .01), struc-ture (ADHD, M = 2.25, SD = 0.99; con-trols, M = 3.04, SD = 0.99, t(46) = 2.76,p < .01), lexicon (ADHD, M = 1.92, SD = 0.93; controls, M = 3.29, SD = 1.30,t(46) = 4.21, p < .01), and grammar(ADHD, M = 1.83, SD = 0.82; controls,M = 3.12, SD = 1.11, t(46) = 4.58, p < .01).In other words, children with ADHDsymptoms obtained lower scores thancontrols on all qualitative parameters.

We also computed the quantity ofwords written by every child, and wefound that children with ADHD symp-

toms wrote significantly less, on aver-age, than controls. In the group withchildren showing ADHD symptoms,the mean length of the text was 51.79words (SD = 21; 95% CI = 42.92–60.66),with a range between 20 and 90. In thecontrol group, the mean length of thetext was 68.21 words (SD = 32.96; 95%CI = 54.29–82.12), with a range between32 and 143. A t-test comparison be-tween group means showed a signifi-cant difference, t(46) = 2.06, p < .05.

Furthermore, we computed thepercentage of errors distinguished ac-cording to the manual classification:phonological errors (PhE), where thereading of the written word wouldhave a different phonology than thereal word; nonphonological errors(NphE), with the same sound for thewritten word and the real word; andthird-type errors. In the last category,the test manual includes errors wherethe sequence of written letters corre-sponds to the correct sequence, butsome subtle phonological markers ofthe Italian writing system are lost (i.e.,geminates and accents; e.g., girafa forgiraffa, citta for città).

Because the length of the essaywas different for each participant, wecomputed the percentages of the threetypes of errors with respect to the totalnumber of written words. We foundthat large groups of children had avery low percentage of errors, suggest-ing that a control for the normality ofthe distribution and the use of non-parametric tests would be appropriate.In fact, the Kolmogorov-Smirnov test,a statistical test for the evaluation ofdistribution normality, showed thatthe distributions deviated significantlyfrom the normal distribution for alltypes of error (p < .05), so we used non-parametric tests. First, we comparedthe two groups on the overall percent-age of errors with the nonparametricMann-Whitney U test and found asignificant difference between groups, U = 97.50, p < .001. In fact, the groupwith children showing ADHD symp-toms made more than four times moreerrors than the control group, with amean percentage of errors higher than

7% (ADHD, M = 7.88, SD = 6.48, range =0–24.24; control, M = 1.66, SD = 2.45,range = 0–9.38).

When we considered the threedifferent types of errors separately, wecould see that many children were per-fectly accurate in one or more cate-gories. For this reason, we decided todivide children into two categories ac-cording to whether they had made oneor more errors of a particular type orno such errors. We made this divisionfor the three types of error. The resultsshowed that the two groups were sig-nificantly different only for the thirdtype of error (see Figure 1), whereasthe differences between groups wereslight both for PhE (62.5% of ADHDand 41.7% of controls made errors),χ2(1, N = 48) = 2.09, p > .05, and forNPhE (29.2% ADHD vs. 12.5% con-trol), χ2(1, N = 48) = 2.02, p > .05. In con-trast, there was a highly significant dif-ference in the third error type, whichwas present for 87.5% of children withADHD and only 29.2% of controls,χ2(1, N = 48) = 16.8, p < .001.

Discussion

Study 1 confirmed that children withADHD symptoms, despite showing anadequate standard of general abilities,presented expressive writing difficul-ties. Their problems were rather gen-eral, as they concerned all the mea-sured aspects. Children with ADHDsymptoms were rated as less proficientwriters on all four basic parameters(adequacy, structure, lexicon, and gram-mar). They produced shorter texts andmade a higher percentage of errors.These difficulties were not related towriting speed, because the two groupsdid not differ on this measure. How-ever, in some aspects of expressivewriting, the differences appearedstronger. In particular, the difference inthe adequacy parameter was very con-spicuous. Furthermore, the pattern oferrors mainly involved errors that re-quire a sophisticated use of both pho-nological and nonphonological indices.

One problem in this study was re-lated to the fact that we did not have a

JOURNAL OF LEARNING DISABILITIES248

measure of the linguistic (includingreading) ability of the participants, anda particularly low linguistic ability ofchildren with ADHD symptoms couldhave affected their writing perfor-mance. Another problem concerned thegeneralizability of the results obtained.In particular, evidence concerning chil-dren in primary school was also neededto examine the presence of difficultiesat earlier phases of the writing skill ac-quisition process. Furthermore, the dif-ficulties of children with ADHD symp-toms could be strictly related to theprocedure adopted in the task, whichwas based on the presentation of a pic-ture (see the procedure also adoptedby Grill & Kirwin, 1989). Effects couldbe different if a verbal scaffold ratherthan a picture were presented and chil-dren were given verbal cues to starttheir description. In fact, it has beensuggested that expressive writing canbe improved on the basis of a promptrepresented by a sufficiently largequantity of verbal material (Marchisan& Alber, 2001). An advantage of theverbal scaffold could be especiallypresent in children with ADHD symp-toms who have difficulties in autono-mously initiating and controlling cog-nitive activities. In fact, the availability

of a short text could facilitate the pro-duction process. For these reasons, in asecond study, we tested children of dif-ferent grades and compared the stan-dard procedure used in Study 1 with adifferent procedure in which the pic-ture was substituted with a verbal scaf-folding description. Children werepresented with the same descriptionrequest used in Study 1, but one of thetasks was based on the standard pre-sentation of the picture, whereas theother task was based on a verbaldescription.

STUDY 2

Study 2 examined whether childrenwith ADHD symptoms who attendeddifferent grades of primary schoolscored lower in expressive writing andspelling, and whether this was the casealso when their productions wereprompted by a detailed verbal prompt.

Method

Participants

Two groups consisting of a total of 163children participated in the study. One

group comprised children describedby their teachers as exhibiting ADHDsymptoms, and a control group wasmatched for gender, age, and school-ing. In total, groups consisted of 46 sec-ond graders (33 boys and 13 girls), 40third graders (32 boys and 8 girls), 44fourth graders (36 boys and 8 girls) and33 fifth graders (29 boys and 4 girls).Participants were selected from publicschools located in northeastern Italyand mainly came from White familiesworking in agriculture, public employ-ment, and industry. The teacher ratingscale (SDAI), already described in thepreceding study, was used for assess-ing 1,468 children. Also in this case,children whose mean score was above1.5 per item on one of the two subscaleswere considered for inclusion in thegroup with children showing ADHDsymptoms, and those who did notmeet this criterion were considered forinclusion in the control group. Teach-ers were interviewed to have con-firmation concerning the presence ofADHD symptoms and were also askedto rate on a 0 to 3 rating scale the pres-ence of general cognitive difficulties,linguistic and mathematics learningdifficulties, problems in the social skillsdomain, oppositional and aggressivebehavior, anxiety problems, and de-pressive behavior (children with thesekinds of problems were excluded fromboth control and experimental groups).Control children were also matched forestimated IQ, rated ability, and SES.The IQ estimation was obtained by ad-ministering part of the PMA 2–4 and4–6 Spatial Reasoning subscale (Thur-stone & Thurstone, 1963/1981), whichrequires finding, in 6 minutes, amongfour different alternatives, the figurethat—combined with the given model—can produce a square. The control groupand the group with children showingADHD symptoms were not signifi-cantly different in their estimatedcognitive abilities (M = .30 for bothgroups), language abilities (M = .53 forthe ADHD group and M = .40 for thecontrol group), and IQ, which, due tothe overestimation in the Italian PMAnorms, was particularly high in both

FIGURE 1. Study 1: Percentages of children with attention-deficit/hyperactivity dis-order (ADHD) and control children who made errors on the Description test. *p <.001. PhE = phonological errors; NPhE = nonphonological errors; 3rd Type = errorson geminates and last syllable accents.

VOLUME 40, NUMBER 3, MAY/JUNE 2007 249

groups: 118.35 (SD = 16.35) in thegroup of children with ADHD symp-toms and 119.30 (SD = 13.50) in thecontrol group. On the other hand, thetwo groups were significantly different(p < .01) on estimated mathematicalabilities (.80 vs. .45)

Tasks

All children were administered two ex-pressive writing tasks (i.e., two De-scription tasks) in their classroom. Allthe tasks were derived from the stan-dardized writing battery devised byTressoldi and Cornoldi (1991). Thewriting tasks were based on the pre-sentation of one of two colored figuresappropriate for the child’s grade andone of the corresponding verbal illus-trations. Figures and verbal illustra-tions concerned the zoo scenario al-ready proposed in the preceding study,and the picture task was also the sameas in Study 1, but all children were alsopresented with another descriptiontask using another stimulus modality,namely, a verbal scaffold rather than apicture. The verbal scaffold includedthe main elements present in the pic-ture; for example, for one of the pic-tures, it included the following: “Try toimagine that you and another childhave been to visit the zoo, where therewere a lot of people and animals. Atone point you stopped in front of acage in which there were many parrotsof different colors.”

Stimulus modality (verbal scaf-fold vs. picture) and order of presenta-tion of the situation (zoo with monkeysvs. zoo with parrots) were counterbal-anced across participants. In all otherrespects, the procedure was the sameas in Study 1.

ResultsA preliminary analysis examinedwhether grade level affected the twogroups in different ways. However, nosignificant interaction was found be-tween grades and groups. Therefore,we will present the overall pattern ofresults, distinguishing only betweengroups (children with ADHD symp-toms and controls) and types of tasks.

For the analysis of the children’sproduction, we followed the proce-dure described in the BVSCO manual.The qualitative judgment was made bytwo blind and independent raters. Theinterrater agreement was high for eachparameter in both tasks (i.e., for theDescription task with the picture: ade-quacy, r = .96; structure, r = .92; lexicon,r = .88, and grammar, r = .91; and forthe Description task with verbal scaf-fold: adequacy, r = .96; structure, r =.90; lexicon, r = .90; and grammar, r =.94). Therefore, in successive analyses,we only considered the ratings givenby the first rater.

Four 2 × 2 (Group × Task [verbalvs. picture]) ANOVAs were run for thedifferent parameters. We did not find a

significant difference between the twotasks, but we found differences betweenthe two groups on all aspects: ade-quacy: F(1, 324 = 261.50, MSE = 307.10,p < .001, partial η2 = .447; structure:F(1324) = 213.18, MSE = 290.47, p < .001, partial η2 = .397; lexicon: F(1324) =239.52, MSE = 220.48, p < .001, partialη2 = .425; and grammar: F(1324) =249.17, MSE = 209.35, p < .001, partialη2 = .435. Means and standard devia-tions of Description task scores with animage and with verbal scaffolding forboth ADHD and groups are detailed inTable 1.

We also carried out a 2 × 2 (Group ×Task) ANOVA to compare the length ofthe descriptions and found a signifi-cant difference between groups (ADHD:M = 52.13, SD = 1.55, 95% CI = 49.09–55.18; control: M = 67.97, SD = 1.55,95% CI = 64.92–71.01; F(1, 324) = 52.33,MSE = 780.88, p < .001, partial η2 = .139,but no significant difference betweenthe two conditions. As can be seenfrom Table 2, there was no difference inthe number of words used to write thedescription of an image alone or withthe aid of a verbal scaffold. That is,children with ADHD symptoms andcontrols wrote on average a similarnumber of words in the two tasks, butin general, as in Study 1, the childrenwith ADHD symptoms wrote fewerwords than the control group.

Following the suggestion of Tres-soldi and Cornoldi (1991), we carriedout further analyses on the protocols,

TABLE 1Study 2: Means and Standard Deviations on Four Basic Parameters of Expressive Writing in

Two Conditions for ADHD and Control Groups

ADHD groupa Control groupa

Image condition Scaffold condition Image condition Scaffold condition

Parameter M SD 95% CI M SD 95% CI M SD 95% CI M SD 95% CI

Adequacy 2.87 0.84 2.73–3.00 2.98 0.91 2.84–3.12 4.15 0.65 4.05–4.25 4.16 0.66 4.06–4.26

Structure 2.77 0.70 2.67–2.88 2.82 0.68 2.72–2.93 3.88 0.81 3.75–4.00 3.88 0.79 3.76–4.01

Lexicon 2.39 0.66 2.29–2.49 2.39 0.58 2.30–2.48 3.40 0.66 3.30–3.50 3.39 0.75 3.27–3.50

Grammar 2.34 0.64 2.25–2.43 2.31 0.60 2.22–2.40 3.29 0.71 3.18–3.40 3.34 0.67 3.24–3.45

Note. ADHD = attention-deficit/hyperactivity disorder.an = 163.

JOURNAL OF LEARNING DISABILITIES250

concerning not only the text length, butalso the percentage of qualitative ad-jectives, the percentage of repetition(both calculated on the first 50 words),and the percentage of subordinateclauses present. These indexes havebeen suggested (Tressoldi and Cor-noldi, 1991) to be associated with goodproduction, mainly reflecting the rich-ness of the ideational process (in thecase of text length), their organizationin the text (in the case of a high per-centage of subordinate clauses), a sat-isfactory use of the lexicon (in the caseof a low percentage of word repeti-tions), and an articulated and sophisti-cated transcription (in the case of ahigh percentage of qualifying adjec-tives).

In this case (see Table 3), the twodescription tasks produced differentoutcomes. Wilcoxon tests—a nonpara-metric test that enables the comparisonof within-subject measures—showedthat the children with ADHD symp-toms showed a significantly lower per-formance on all parameters. Further-more, they wrote significantly morequalitative adjectives in the conditionwith verbal scaffolding than in the con-dition with the picture, Wilcoxon z =3.54, p < .001; they made more repeti-tions in the condition with the imagethan in the other condition, z = 2.4, p <.05; but the difference in the number ofsubordinate clauses only approachedsignificance, z = 1.9, p = .05. In contrast,for the control group, there was no dif-

ference between the two conditions inthe percentages of qualitative adjec-tives, z = 0.99, p > .05, and repetitions,z = 0.17, p > .05, but there was a signif-icantly higher percentage of subordi-nate clauses in the picture task than inthe verbal scaffolding task, z = 2.11, p <.05. These results are shown in Table 3.

In summary, the performance ofchildren with ADHD symptoms waspoorer (used less adjectives and mademore repetitions) on the task with thepicture than on the task with the verbalscaffolding. In contrast, control chil-dren scored higher (used more subor-dinate clauses) in the condition withimages than in the verbal scaffoldingcondition. Furthermore, comparingthe two groups on these parameters,we found that children with ADHDsymptoms used less qualitative adjec-tives, made more repetitions, andwrote less subordinate clauses thancontrol children, both in the task withan image and in the task with verbalscaffolding (see Table 3).

The lower performance of thegroup with children showing ADHDsymptoms was also evident from thetotal number of spelling errors; herethere was no significant difference be-tween the two tasks. Also in this study,when we considered the three different

TABLE 2Study 2: Means and Standard Deviations of Description Text Lengths in

Two Conditions for ADHD and Control Groups

ADHD groupa Control groupa

Condition M range SD M Range SD

Description with image 51.89 18–104 19.63 68.82 33–132 21.66

Description with verbal scaffolding 52.38 18–121 18.90 67.12 28–131 22.28

Note. ADHD = attention-deficit/hyperactivity disorder.an = 163.

TABLE 3Study 2: Means and Standard Deviations of Percentages of Adjectives, Repetitions, and Subordinate

Clauses for ADHD and Control Groups

Adjectives Repetitions Subordinates

Group M Range SD M Range SD M Range SD

Image condition

ADHDa 2.9 0–9 2.3 7.4 0–27 4.0 29.44 0–133 27.86

Controla 6.6 1–18 3.5 2.7 0–12 2.5 46.70 0 –125 25.21

Mann-Whitney U 5248.5* 3754.5* 8692.5*

Verbal scaffolding condition

ADHDa 3.4 0–13 2.7 6.6 0–29 3.3 27.95 0–133 28.43

Controla 6.7 0–16 3.7 2.8 0–11 2.4 44.26 0–125 22.65

Mann-Whitney U 6529.5* 4139.0* 8568.0*

Note. ADHD = attention-deficit/hyperactivity disorder.an = 163.*p < .001.

VOLUME 40, NUMBER 3, MAY/JUNE 2007 251

types of errors separately, we could seethat many children’s performanceswere accurate in one or more cate-gories. For this reason, we again di-vided children into two categories ac-cording to whether they had made oneor more errors or none.

As can be seen in Figure 2, chil-dren with ADHD symptoms mademore errors than control children bothin the condition with images (ADHD:M = 3.6, SD = 2.6, range = 0–13; con-trols: M = 1.6, SD = 1.5, range = 0–7,Mann-Whitney U = 6,800, p < .001) andin the condition with verbal scaffold-ing (ADHD: M = 3.7, SD = 2.5, range =0–14; controls: M = 1.7, SD = 1.6, range =0–9, Mann-Whitney U = 6,178, p < .001).In particular, in the condition with im-ages, 22.7% of children with ADHDsymptoms versus 13.5% of control chil-dren made PhE, χ2(1, N = 326) = 11.95,p = .001; 36.2% of children with ADHDsymptoms versus 23.9% of control chil-dren made NPhE, χ2(1, N = 326) =20.47, p < .001; and 39.9% of childrenwith ADHD symptoms versus 30.1%of controls made third-type errors,χ2(1, N = 326) = 14.94, p < .001. A simi-lar result could be observed in the ver-bal scaffolding condition: 22.2% of chil-dren with ADHD symptoms versus11.4% of controls made PhE, χ2 (1, N =326) = 17.24, p < .001; 35% of childrenwith ADHD symptoms versus 24.8% ofcontrols made NPhE, χ2(1, N = 326) =13.90, p < .001; and 43.4% of childrenwith ADHD symptoms versus 32.3%of controls made third-type errors,χ2(1, N = 326) = 20.77, p < .001.

Discussion

Study 2 confirms the results of the pre-ceding one: Children with ADHD symp-toms present writing difficulties. Inparticular, children with ADHD symp-toms are rated as less proficient writ-ers, make more errors, and writeshorter texts than controls. These er-rors cannot be due to a general lin-guistic disability in the children withADHD symptoms because the twogroups were matched on this aspect.

However, their impaired writing profi-ciency could be affected by a more spe-cific linguistic deficit. The deficit couldalso be related to a more general scho-lastic difficulty, as suggested by theirlack of high performance in math (asreported by their teachers). The find-ing that—in contrast to the precedingstudy—control children also made aconsiderable number of errors of thethird type can be explained by consid-ering the young age of the children inthis study, at a stage in which they arestill involved in learning to managesubtle spelling skills.

This study examined the writingdifficulties encountered by childrenwith ADHD symptoms and testedwhether producing a written descrip-tion of a scene aided by a verbal scaf-fold could be simpler than with the aidof an image alone, by offering a verbalstructure to be used for planning andproducing a verbal text. In general,children showed a similar pattern ofdifficulties on both tasks. Resultsshowed that the availability of a verbalscaffold during a writing task does notrepresent a strong help for a child en-gaged in a writing task, and this result

is valid both for children with ADHDsymptoms and for controls. Childrenwith ADHD symptoms performedlower than controls on all parameters,they made a higher percentage of er-rors, and they produced shorter texts.However, the presence of a very simpleverbal scaffold slightly reduced theirperformance difficulties, improving thequality of their texts by increasing the number of adjectives and reducingthe number of repetitions. It is also im-portant to note that the only differencebetween the two conditions (with pic-ture or with verbal instructions) forcontrols was a very slight improve-ment in the percentage of subordinateclauses in the picture condition.

STUDY 3

The difficulties encountered by chil-dren with ADHD symptoms couldhave been emphasized by the particu-lar type of expressive task set. In fact,it has been shown (e.g., Ellis, Taylor, &Drury, 2005; Graham, Harris, & Mason,2005; Segev-Miller, 2004) that expres-

FIGURE 2. Study 2: Percentages of children with attention-deficit/hyperactivitydisorder (ADHD) and control children who made errors on the Description task con-dition with image and on the condition with verbal scaffolding. PhE = phonologicalerrors; NPhE = nonphonological errors; 3rd = third-type errors on geminates andlast syllable accents.

JOURNAL OF LEARNING DISABILITIES252

sive writing is largely dependent onthe type of writing task required. Asthe BVSCO standardized test batteryfor Italy includes another writing task(i.e., a narrative one), we further testedchildren with ADHD symptoms ontheir expressive writing skills by pre-senting this task. Thus, the same chil-dren examined in the preceding studywere invited to write a narrative textbased on the presentation of a series ofimages describing a short story.

Method

Participants were the same as in Study 2,and the procedure was the same as inStudy 1, with the presentation of picto-rial material and the request to pro-duce a written text. However, ratherthan being presented with a singleimage, children were presented with aseries of pictures forming a short story.The pictorial material for the writingtest (Tressoldi & Cornoldi, 1991) forsecond and third graders consisted ofthree cartoons depicting the story of achild and his dog. For fourth and fifthgraders, it included five cartoons de-picting the story of a child falling off atree. To give a more complete exam-ple, the cartoons for younger childrenshowed the story of a child who firstplays with his dog (Cartoon 1), thengoes to school with the dog (Cartoon 2)and leaves the dog waiting for himoutside while he stays in school (Car-

toon 3). All children had 10 min towrite their text.

Results

As for the other studies, we followedthe BVSCO manual’s procedure. First,two independent and blind raters eval-uated the protocols. The correlationsbetween the two raters’ assessments ofthe qualitative parameters for the nar-rative task were very high for ade-quacy (r = .95, p < .001), structure (r =.91, p < .01), lexicon (r = .85, p < .001),and grammar (r = .81, p < .001). For thisreason, we decided also in this case toconsider the evaluations of the firstrater only.

The performance of childrenshowing ADHD symptoms was signif-icantly lower on all qualitative para-meters compared to the control group,as shown in Table 4. We compared thetwo groups on the number of wordswritten, and we found, again, that thechildren with ADHD symptoms wrotefewer words than the control group. Infact, the children with ADHD symp-toms wrote on average 50.77 words(SD = 14.75, range = 21–99), whereasthe control group wrote 64.32 words(SD = 17.78, range = 27–118), a differ-ence found to be significant, t(324) =7.49, p < .01.

When we considered the first 50text words for all children, we foundthat children with ADHD symptoms

used less qualitative adjectives (ADHD:M = 3.2, SD = 2.8, range = 0–11; con-trols: M = 6.3, SD = 4, range = 0–15,Mann-Whitney U = 7,300.5, p < .001),made more repetitions (ADHD: M =9.8, SD = 4.6, range = 0–22; controls: M = 3.9, SD = 3.6, range = 0–16, Mann-Whitney U = 4,193.5, p < .001), and wrotefewer subordinate clauses (ADHD: M =26.28, SD = 25.11, range = 0–133; con-trols: M = 44.7, SD = 23.84, range =0–150, Mann-Whitney U = 7,510.5, p <.001) than control children (see Fig-ure 3). Furthermore, children withADHD made significantly more errorsthan controls (ADHD: M = 3.4, SD =2.01, range = 0–11; controls: M = 1.7, SD = 1.3, range = 0–7, Mann-WhitneyU = 6,177, p < .001). Comparing the twogroups on the different types of errors,we divided children into two groupsaccording to whether they had madeone or more errors or none. The resultsshowed that 19.3% of children withADHD symptoms made phonologicalerrors versus 12.6% of controls, χ2(1, N = 326) = 6.83, p = .009; that 36.4% ofchildren with ADHD symptoms madenonphonological errors versus 24.4%of controls, χ2(1, N = 326) = 18.49, p <.001; and that 40.6% of children withADHD symptoms made third-type er-rors versus 36.3% of controls, χ2(1, N =326) = 3.03, p > .05.

As can be seen in Figure 4, chil-dren with ADHD symptoms alwaysmade more errors than controls, butthere was no relevant difference be-tween the two groups on third-typeerrors. A final analysis comparing thedifferent types of errors showed that inboth groups, phonological errors wereless frequent than nonphonological er-rors (ADHD: PhE = 38.7% vs. NPhE =72.4%, McNemar’s p < .001; control:PhE = 24.8% vs. NPhE = 49.1%, McNe-mar’s p < .001) and nonphonologicalerrors were less frequent than third-type errors for the control group(NPhE = 49.1% vs. 3rd type = 73.3%,McNemar’s p < .001) but not for chil-dren with ADHD symptoms; in fact,for ADHD, NPhE were not signifi-cantly less frequent than third-type er-rors (NPhE = 72.4% vs. 3rd type = 81%,

TABLE 4Study 3: Means and Standard Deviations on Four Basic Parameters of Narrative

Writing for ADHD and Control Groups

ADHD groupa Control groupa

Parameter M SD 95% CI M SD 95% CI t (324)

Adequacy 3.23 1.01 3.08–3.39 4.36 0.70 4.25–4.47 11.67*

Structure 2.34 0.60 2.24–2.43 3.24 0.64 3.14–3.34 13.05*

Grammar 2.42 0.62 2.32–2.51 3.38 0.73 3.27–3.49 12.86*

Lexicon 2.92 0.73 2.80–3.02 4.01 0.74 3.89–4.12 13.59*

Note. ADHD = attention-deficit/hyperactivity disorder.an = 163.*p < .001.

VOLUME 40, NUMBER 3, MAY/JUNE 2007 253

McNemar’s p > .05). Finally, children inboth groups made numerous errors ofthe third type.

Discussion

The same patterns of performance ob-served in the preceding studies withthe Description tasks were also found in the Narrative task: Children withADHD symptoms always made moreerrors than control children, wroteshorter texts, and were judged to bepoorer writers than controls. This re-sult shows that the writing difficultiesof children with ADHD symptoms donot depend on the type of writing taskor on the type of instructions given. Inall the present studies, and in all con-ditions, children with ADHD symp-toms always wrote less proficientlyand produced shorter texts than thecontrol children.

In this third study, we found thesame pattern of specific problems forthe children with ADHD symptomsthat were found in the previous stud-ies. The only difference was that on thistask, children with ADHD symptomsand controls did not differ in the num-ber of third-type errors made. In fact, arelevant proportion of control childrenalso made this type of errors. It is thuspossible that the narrative task wasmore difficult for both groups, sub-tracting critical resources from payingattention to spelling, so that all chil-dren made third-type errors.

GENERAL CONCLUSION

The three studies presented here shedlight on the expressive writing skills ofchildren with ADHD symptoms—anaspect of this disorder that has beenrather understudied so far. The resultsshowed that the expressive writingperformance of children described bytheir teachers as exhibiting ADHDsymptoms was always lower than thecontrol group’s performance under allthe considered aspects (number of er-rors made, qualitative and objective

parameters of the production). Thus,our results show that the difficultiesthat children with ADHD symptomsface at school are not limited to the tra-ditionally considered areas of reading(e.g., Pisecco et al., 2001) and mathe-matics (e.g., Marzocchi et al., 2002), butalso concern different aspects of writ-ing. If an increase in spelling errors inchildren with ADHD symptoms may

be associated with a reading difficulty(see Frith, 1985), the relevant difficul-ties faced by children with ADHDsymptoms in producing an adequate,well-organized text could be due totheir well-known planning and organi-zational difficulties (e.g., Barkley,1995).

Concerning spelling errors, de-spite following the same developmen-

FIGURE 3. Study 3: Mean writing performance of children with attention-deficit/hyperactivity disorder (ADHD) and control children on adjectives, repetitions, andsubordinate clauses on the Narrative test. Error bars represent 1 SD. *p < .001.

FIGURE 4. Study 3: Percentages of children with attention-deficit/hyperactivity dis-order (ADHD) and control children who made errors on the Narrative test. *p <.001; PhE = phonological errors; NPhE = nonphonological errors; 3rd = third-typeerrors on geminates and last syllable accents.

JOURNAL OF LEARNING DISABILITIES254

tal path described by the stage model(Frith, 1985; Tressoldi & Cornoldi,1991), children with ADHD symptomspresented more of the three types of er-rors analyzed here (i.e., phonological,nonphonological, and third type, thelatter involving last-syllable accents,which in Italian must be reported, andgeminates). This type of error later dis-appears in all children, but is still pres-ent in children with ADHD symptomswho are attending secondary school.This evidence is consistent with amodel that outlines a number of stagesin learning to write, starting from theacquisition of the ability to transcribesimple words (phonological errors),then more complex ones (nonphono-logical errors), and finally acquiringthe rules regulating accents and gemi-nates (third-type errors). The presentarticle focused on writing skills and,unfortunately, no information was col-lected on the children’s reading abili-ties—an issue that should be consid-ered in future research in the area. Forthe children tested in Studies 2 and 3,we only had the teachers’ general rat-ing on language abilities, which didnot produce significant differencesbetween the group of children withADHD symptoms and the controlgroup. However, due to the typical dif-ficulties that children with ADHDsymptoms encounter in reading (e.g.,Bonafia et al., 2000) and the interactionbetween reading and spelling (Frith,1985), one could predict that the chil-dren examined in the present studiesshould also present with reading diffi-culties, and the latter should have aninfluence on the observed spelling dif-ficulties to produce in the child somehesitations also in the expressive writ-ing context. In a similar vein, one couldhave predicted that the reading diffi-culties observed in preceding studieswith children with ADHD symptomscould be at least partially affected bytheir spelling difficulties. The fact thatboth spelling and reading difficultiescan be found leaves open the questionof the nature of their frequent comor-bidity with ADHD and of the potential

causal links—an issue that cannot beresolved on the basis of the presentdata.

However, in the case of expres-sive writing, a negative influence ofADHD on academic performance seemsclear. In fact, for children with ADHDsymptoms, producing ideas, organiz-ing concepts, and writing them downis hard work. For this reason, probably,the children with ADHD symptoms inthe present studies wrote less, orga-nized the text poorly, used a limitedvocabulary, and made more errors.Usually, their text was very simple andnot well articulated—more similar to alist of elements than to an organizedtext. The adequacy and structure of thewriting of children with ADHD symp-toms was particularly poor, but also forlexicon and grammar, the differencesbetween children with ADHD symp-toms and controls were dramaticallyevident (Studies 1, 2, and 3). The poorerlexicon of the children with ADHDsymptoms was also confirmed by thegreater number of repetitions and thelower number of adjectives used (Stud-ies 2 and 3). In this respect, the presentresearch confirms that the inclusion ofadjectives in a text is associated withhigher writing proficiency. The poorgrammatical and syntactic qualities ofthe texts produced by children withADHD symptoms are also confirmedby the less complex syntactic structuredemonstrated by the lower number ofsubordinate clauses (Studies 2 and 3).

There are a number of issues thatremain to be investigated and clarified.In particular, our observations shouldbe replicated in contexts using morestandardized procedures for the di-agnosis of ADHD and its comorbidsymptoms. However, we believe that apriority in future research should begiven to the study of how to facilitatethe writing process of children withADHD. Thus, we think that research-ers in the field should investigate waysto train and help children with ADHDto write more proficiently and in ac-cordance with their age. Developmentsin this field could also take advantage

of some of the results of these studies,for example the fact that ADHD chil-dren prefer verbal to pictorial prompts.

ABOUT THE AUTHORS

Anna Maria Re, BA, is a PhD student in psy-chology at the University of Padova, Italy. Hermain interests include learning disabilities,attention-deficit/hyperactivity disorder (ADHD),and children with ADHD, especially in pre-school years. Martina Pedron, BA, is a psy-chologist with expertise in learning disabilities.Currently she conducts assessment and treat-ment at the University of Padova. Cesare Cor-noldi, PhD, is a full professor of generalpsychology at the University of Padova. His re-search interests include human learning andmemory, cognitive processes in learning dis-abilities, working memory, metacognition andlearning, and mental imagery. Address: AnnaM. Re, Dipartimento di Psicologia Generale,Universita degli studi di Padova, via Venezia,8—35131, Padova, Italy; e-mail: [email protected]

NOTES

1. In these studies, we used only the teachers’ratings. It must be noted that in Italy, nostandard procedures are used for ADHD as-sessment and diagnosis, and medication wasnot permitted at the time of these studies.However, teachers had a long-term familiar-ity with the children and knowledge abouttheir behavior in different contexts. More-over, the identification of children withADHD symptoms was checked through suc-cessive informal interviews.

2. The manual presents validation studies.Concerning the test’s psychometric proper-ties, the manual only reports a mean test–retest reliability of .57 for the classificationsystem of the three types of spelling errors.

REFERENCES

American Psychiatric Association. (1994).Diagnostic and statistical manual of mentaldisorders (4th ed.). Washington, DC: Au-thor.

Barkley, R. A. (1995). Taking charge of ADHD:The complete, authoritative guide for parents.New York: Guilford Press.

Benedetto-Nasho, B., & Tannock, R. (1999).Math computation, error patterns andstimulant effect in children with attention

VOLUME 40, NUMBER 3, MAY/JUNE 2007 255

deficit hyperactivity disorder. Journal ofAttention Disorders, 3, 121–134.

Bonafia, M. A., Newcorn, J. H., McKay, K. E., Koda, V. H., & Halperin, J. M.(2000). ADHD and reading disabilities: Acluster analytic approach for distinguish-ing subgroups. Journal of Learning Disabil-ities, 33, 297–307.

Burnett, R. E., & Kastman, L. M. (1997).Teaching composition: Current theoriesand practices. In G. D. Phye (Ed.), Hand-book of academic learning (pp. 268–305).San Diego: Academic Press.

Cornoldi, C., Barbieri, A., Gaiani, C., & Zoc-chi, S. (1999). Strategic memory deficits inattention deficit disorder with hyperac-tivity participants: The role of executiveprocesses. Developmental Neuropsychology,15, 53–71.

Cornoldi, C., Gardinale, M., Masi, A., & Pet-tenò, L. (1996). Impulsività e autocontrollo[Impulsivity and self-control]. Trento,Italy: Erickson.

Ellis, R., Taylor, C. E., & Drury, H. (2005).Evaluating writing instruction throughan investigation of students’ experiencesof learning through writing. InstructionalScience, 33, 49–71.

Englert, C., & Raphael, T. (1988). Construct-ing well-formed prose: Process, structure,and metacognitive knowledge. Excep-tional Children, 54, 513–520.

Frith, U. (1985). Beneath the surface of sur-face dyslexia. In J. C. Marshall, M. Colt-heart, & K. Patterson (Eds.), Surfacedyslexia and surface dysgraphia (pp. 310–330). London: Routledge & Kegan Paul.

Graham, S., & Harris, K. (1989). Compo-nents analysis of cognitive strategy in-struction: Effect on learning disabled stu-dents’ compositions and self-efficacy.Journal of Educational Psychology, 81, 353–361.

Graham, S., Harris, K., MacArthur, C., &Schwartz, S. (1991). Writing and writinginstruction with students with learningdisabilities: A review of a program of re-search. Learning Disability Quarterly, 14,89–114.

Graham, S., Harris, K., & Mason, L. (2005).Improving the writing performance,knowledge, and self-efficacy of strug-gling young writers: The effects of self-regulated strategy development. Contem-porary Educational Psychology, 30, 207–241.

Grill, J. J., & Kirwin, M. M. (1989). Writtenlanguage assessment manual. Novato, CA:Academic Therapy.

Hayes, J. R., & Flower, L. S. (1980). Identi-fying the organization of the writingprocess. In L. W. Gregg & E. R. Stein-berg (Eds.), Cognitive processes in writing(pp. 3–30). Hillsdale, NJ: Erlbaum.

Johnson, B. D., Altmaier, E. M., & Richman,L. C. (1999). Attention deficits and read-ing disabilities: Are immediate memorydefects additive? Developmental Neuropsy-chology, 15, 213–226.

Kroese, J. M., Hynd, G. W., Knight, D. F.,Hiemenz, J. R., & Hall, J. (2000). Clinicalappraisal of spelling ability and its rela-tionship to phonemic awareness (blend-ing, segmenting, elision and reversal),phonological memory and reading inreading disabled, ADHD and normalchildren. Reading and Writing, 13, 105–131.

Marchisan, M. L., & Alber, S. R. (2001). Thewrite way: Tips for teaching the writingprocess to resistant writers. Intervention inSchool and Clinic, 36, 154–162.

Marzocchi, G. M., & Cornoldi, C. (2000).Una scala di facile uso per la rilevazionedei comportamenti problematici dei bam-bini con deficit di attenzione e iperattivita[An easy-to-use scale for the descriptionof ADHD symptoms]. Psicologia Clinicadello Sviluppo, 4, 43–63.

Marzocchi, G. M., Lucangeli, D., De Meo,T., Fini, F., & Cornoldi, C. (2002). The dis-turbing effect of irrelevant informationon arithmetic problem solving in inatten-tive children. Developmental Neuropsychol-ogy, 21, 73–92.

Miceli, G., Benvegnù, B., Capasso, R., &Caramazza, A. (1995). Selective deficit inprocessing double letters. Cortex, 31, 161–171.

Montague, M., Graves, A., & Leavell, A.(1991). Planning, procedural facilitation,and narrative composition of junior highstudents with learning disabilities. Learn-ing Disabilities Research & Practice, 6, 219–224.

Newcomer, P. L., & Barenbaum, E. M.(1991). The written composing ability ofchildren with learning disabilities: A re-view of the literature from 1980 to 1990.Journal of Learning Disabilities, 24, 578–593.

Pisecco, S., Baker, D. B., Silva, P. A., &Brooke, M. (2001). Boys with reading dis-

abilities and/or ADHD: Distinctions inearly childhood. Journal of Learning Dis-abilities, 34, 98–106.

Resta, P. S., & Eliot, J. (1994). Written ex-pression in boys with attention deficitdisorder. Perceptual and Motor Skills, 79,1131–1138.

Ross, P. A., Poidevant, J. M., & Miner, C. U.(1995). Curriculum-based assessment ofwriting fluency in children with attentiondeficit hyperactivity disorder and normalchildren. Reading and Writing Quarterly:Overcoming Learning Difficulties, 11, 201–208.

Segev-Miller, R. (2004). Writing fromsource: The effect of explicit instructionon college students’ process and prod-ucts. L1–Educational Studies in Languageand Literature, 4, 5–33.

Seidman, L. J., Biederman, J., Monuteaux,M. C., Doyle, A. E., & Faraone, S. V. (2001).Learning disabilities and executive dys-function in boys with attention deficit hy-peractivity disorder. Neuropsychology, 15,544–556.

Shimabukuro, S. M., Prater, M. A., Jenkins,A., & Edelen-Smith, P. (1999). The effectof self-monitoring of academic perfor-mance on students with learning disabil-ities and ADD/ADHD. Education andTreatment of Children, 22, 397–424.

Swanson, H. L., Mink, J., & Bocian, K. M.(1999). Cognitive processing deficits inpoor readers with symptoms of readingdisabilities and ADHD: More alike thandifferent? Journal of Educational Psychol-ogy, 91, 321–333.

Thurstone, T. G., & Thurstone, L. L. (1963).Primary mental abilities. Chicago: ScienceResearch Associates. (Italian translation,1981)

Tressoldi, P. E., & Cornoldi, C. (1991). Batte-ria per la valutazione della scrittura e dellacompetenza ortografica nella scuola dell’ob-bligo [Battery for the assessment of writ-ing skills of children from 7 to 13 yearsold]. Florence, Italy: Organizzazioni Spe-ciali.

Willcutt, E. G., & Pennington, B. F. (2000).Comorbidity of reading disability and at-tention deficit hyperactivity disorder:Differences by gender and subtype. Jour-nal of Learning Disabilities, 33, 179–191.