SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Aim  It has been suggested that one approach to identifying motor impairment in children is to use the Child Behavior Checklist (CBCL) as a screening tool. The current study examined the validity of the CBCL in identifying motor impairment.

Method  A total of 398 children, 206 females and 192 males, aged from 3 years 9 months to 14 years 10 months were assessed on the McCarron Assessment of Neuromuscular Development to determine their motor ability. Parents completed the CBCL.

Results  The ‘Clumsy’ item on the CBCL was found to predict motor ability independent of the child’s age, sex, and scores on other items of the CBCL. However, the sensitivity of the ‘Clumsy’ item in terms of identifying motor impairment was found to be a low 16.7% compared with specificity of 93.2%. The item ‘Not liked’ was also found to be a significant predictor of motor impairment.

Interpretation  Although the ‘Clumsy’ and ‘Not liked’ items were found to have a relationship with motor ability, they should not be relied upon to categorize children as motor impaired versus not impaired. It is possible that these items may be better indicators of motor impairment in children with developmental disorders such as attention-deficit-hyperactivity disorder, but clinical samples are needed to address this.

List of Abbreviations
CBCL

Child Behavior Checklist

DCD

Developmental coordination disorder

DCDQ

Developmental Coordination Disorder Questionnaire

MAND

McCarron Assessment of Neuromuscular Development

NDI

Neurodevelopmental index

Poor motor coordination accompanies many developmental disorders, including attention-deficit-hyperactivity disorder (ADHD),1,2 autism spectrum disorders,3 and reading disorder.4 It is the core deficit in developmental coordination disorder (DCD), described by the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV)5 as ‘a marked impairment in the development of motor coordination that significantly interferes with academic competence or daily living skills’. Children with DCD have low self-perception and self-worth6,7 and higher levels of anxiety8 and depression.9 These children are less likely to engage in physical activity,10 with consequences for physical health and social adjustment. It is therefore important to know whether children with a diagnosed psychiatric disorder also have comorbid DCD or motor impairment to ensure that they receive the appropriate intervention.

Diagnosis of motor impairment is difficult as it is heterogeneous in nature,11 with impairments involving locomotion, ball skills, balance, or fine motor tasks such as handwriting or using a knife and fork. Although there is no criterion standard measure of motor impairment, several tests are commonly used for school-aged children. These include the Movement Assessment Battery for Children12 or the more recent 2nd version,13 the Bruininks–Oseretsky Test of Motor Proficiency14 or the more recent 2nd version15, and the McCarron Assessment of Neuromuscular Development (MAND).16 However, these require individual assessment and take 30 to 40 minutes to administer. In addition to these performance tests, there are screening measures for motor impairment which parents or teachers can fill in, such as the Developmental Coordination Disorder Questionnaire (DCDQ),17 which have proved effective in identifying children at risk of DCD.

In a recent study examining the relationship between parent-reported motor problems, autistic symptoms, and ADHD, Reiersen et al.18 used two items from the Child Behavior Checklist (CBCL)19– Item 36, ‘Gets hurt a lot, accident-prone’ and Item 62, ‘Poorly coordinated or clumsy’– to identify parent-reported motor impairment. It is significant that participants with or without ADHD who showed no endorsement of these items rarely had clinically significant levels of autistic symptoms. However, significant autistic symptoms were seen in over 70% of children with the combination of ADHD and endorsement of both CBCL motor items. The authors noted that they could not find any published studies examining the relationship between these two CBCL items and examination-based motor impairment, so it was unclear what type or degree of motor impairment was likely in children classified using the CBCL motor items. On the basis of the above study, the CBCL motor items do appear to have some utility in identifying children with ADHD who warrant further assessment for autistic features, but the relationship of these CBCL items with examination-based motor impairment is still unclear. The CBCL is widely used for screening purposes, and if the motor items identified by Reiersen et al. do predict poor motor skills, they would provide a cost-effective way of determining which children should be more fully assessed for motor impairment.

Item 62 has been previously linked with psychiatric disorder in a Norwegian study by Novik.20 Novik identified eight CBCL items that were most strongly related to psychiatric diagnoses (determined through interview). These included the item ‘Identifying poor coordination’ (item 62) plus the following items: l, ‘Acts too young for his/her age; 11, ‘Clings to adults or too dependent’; 12, ‘Lonely’; 19, ‘Demands attention’; 35, ‘Feels worthless or inferior’; 48, ‘Not liked’; and 103, ‘Unhappy, sad or depressed’. Most of the psychiatric disorders identified by Novik were emotional disorders, and as DCD has been linked with emotional problems in the past, it is not surprising that a relationship was found.

The aim of the current study was to determine whether the CBCL could be used effectively to screen for children with motor impairment. If it proved effective in identifying children at risk of motor problems this would provide a simple and reliable way for practitioners to identify comorbid motor impairment in children with other psychiatric disorders in order to determine suitable intervention. Given the link between motor impairment and emotional problems such as anxiety and depression, the items found by Novik20 to be linked with psychiatric disorders were also evaluated to ensure that the ‘motor’ items were more effective at picking up motor impairment than these other items. The MAND16 was used as the criterion measure.

Method

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Participants

A total of 404 children and adolescents were included in this study. However, five were removed because of incomplete data on the CBCL, and one other was removed as she was an extreme outlier. This left a total of 398 children (192 males, 206 females) aged from 3 years 9 months to 14 years 10 months (mean 9y; SD 3.05y). They had a mean MAND Neurodevelopmental Index (NDI) score of 96.33 (SD 15.54) with a range of 55 to 147. Of those found to have a motor impairment, 70 were mildly impaired (NDI score 71–85) and 20 were moderately impaired (NDI score 55–70). None were identified with a severe impairment.

Participants were recruited from 42 schools/preschools in the metropolitan region of Perth, Western Australia. Schools were chosen on the basis of their position on a state-wide index of student achievement, and therefore represented the distribution of academic achievement within the state. As this was a community sample, the presence of psychiatric symptoms was not considered as part of the selection process (see Dyck et al.21 for further details on the selection of this sample).

Materials

McCarron Assessment of Neuromuscular Development16

The MAND comprises 10 tasks, five assessing fine motor and five assessing gross motor skills. The scaled scores on each of these tasks are added and age norms, provided for children aged 3 years 6 months to 18 years, are used to determine the NDI with a mean of 100 and SD of 15. The MAND has been found to be a sensitive and valid measure of identifying motor impairment in Australian children.22 The reliability coefficients for each of the 10 tasks ranged from r=0.67 to r=0.98, with a total score reliability coefficient of r=0.99.16

Child Behavior Checklist19

The CBCL was designed for children from 4 to 18 years of age and lists symptoms that parents rate as ‘not at all true’ (0), ‘sometimes true’ (1), or ‘mostly true’ (2) of their child. Only the following nine items were used in the current study: item 1, ‘Acts too young for his/her age’; item 11, ‘Clings to adults’; item 12, ‘Lonely’; item 19, ‘Demands attention’; item 35, ‘Feels worthless or inferior’; item 36, ‘Gets hurt a lot, accident prone’; item 48, ‘Not liked’; item 62, ‘Poorly coordinated or clumsy’; and item 103, ‘Unhappy, sad or depressed’. As with the study by Reiersen et al.,18 the scores on these items were dichotimized to either 0, meaning no endorsement of the item, or 1 and 2, meaning the item was endorsed.

Procedure

Ethical approval for this study was obtained from the human research ethics committee at Curtin University. All testing was carried out with the informed consent of both the participants and their parents in accordance with the guidelines set out by the Australian National Health and Medical Research Council. The data for this study were collected as part of a larger study which assessed children on a wide range of abilities including motor and empathic ability, intelligence, executive functioning, and language in order to understand the relationship between these abilities and behavioral problems. These tests were assessed individually by trained assessors for each child during three to four separate testing sessions. The CBCL was sent home for parents to complete along with the consent form for the child’s participation in the project. Further details of the methodology can be found in Dyck et al.21

Statistical methods

The data were initially analysed using a hierarchical regression procedure with NDI score as the criterion variable. In a hierarchical regression predictors are entered in a sequential fashion in which the order of entry is decided on theoretical grounds and not statistical grounds. The hierarchical regression reported here proceeded in three steps. On step 1 the predictors of age and sex were entered since we wished to control for these variables before evaluating the contribution of the other predictors. On step 2 the seven CBCL items predictive of psychiatric disorders20 were added to the model. On step 2, before evaluating the contribution of items 36 and 62, we wanted to examine the contribution of these seven CBCL items once age and sex were controlled for. On step 3 items 36 and 62 were entered to examine whether these predictors accounted for variance in NDI beyond the variance accounted for by the seven CBCL items entered on step 2. For any step, unique variance refers to the proportion of variance accounted for in the NDI by each predictor after the effects of the other predictors in the model are netted out. Unique variance is measured as the squared semi-partial correlation coefficient (sr2). Variance accounted for in NDI by the predictors in combination is measured by R2. Incremental variance, on the other hand, refers to the increase in the proportion of variance accounted for when a set of variables is added to a model and is assessed by R2 change. An inspection of the incremental variance on the final step allowed us to determine whether items 36 and 62 accounted for any variance in predicting motor impairment (the NDI) once the other items identified with psychiatric disorder (items 1, 11, 12, 19, 35, 48, and 103) were accounted for. Statistical significance was evaluated at an alpha level of 0.05.

Statistical power of the study was calculated for the incremental variance on the final step of a hierarchical regression. Since the effects were expected to be small, power was calculated for a change in R2 of 0.02. Given an initial model with nine predictors (age and sex as control variables, items 1, 11, 12, 19, 35, 48, and 103) and R2=0.30, in order to detect an incremental change in variance of sr2=0.02 at alpha=0.05 and power=0.80, a sample size of 340 was required.

Discrimination validity was determined by dichotomizing the NDI into children with (≤85) and without (>85) motor problems and then determining the sensitivity and specificity of the CBCL items in relation to the NDI categories. Finally, level of impairment was determined by categorizing the children’s MAND NDI score based on the criteria for mild, moderate, and severe disability and then determining the proportion of children identified by the CBCL in these categories.

On the basis of the standard residual, Mahalanobis distance and Cook’s distance, one case was found to be both unusual and influential for the combination of predictors. It was decided that this participant did not belong to the population and was removed. There were missing values for the items measured for five participants and these participants were eliminated from the analysis.

Results

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Hierarchical regression between CBCL items and NDI

Table I reports the validities (zero-order correlations) for the relationships between the predictor variables and the criterion variable (NDI score) as well as the semi-partial correlations for the final model in which all the predictors had been included. The zero-order correlation is a correlation between variables in which no other variables have been controlled for. The semi-partial correlation is the correlation between the predictor and NDI controlling for the other predictors. An inspection of the zero-order correlations shows that all items except sex and item 103 (‘Unhappy, sad, or depressed’) were significantly related to the NDI at the 0.05 level. For the CBCL items a negative correlation indicates that endorsement of the item is associated with a lower NDI score.

Table I.   Zero-order and semi-partial correlations between the predictor variables and the dependent variable (neurodevelopmental index)
Predictor variable (coding)Correlation
Zero-orderSemi-partial
  1. For all items 0 = not endorsed, 1 = endorsed. For the zero-order correlations, the Pearson product moment correlation is reported for age and the point-biserial correlation is reported for the remaining predictors. The semi-partial correlations are for the final model with all predictors included (n=398). ap<0.05; bp<0.01; cp<0.001 (two-tailed).

Sex (0=male, 1=female)0.0110.025
Age0.142b0.146b
Item 1−0.148b−0.086
Item 11−0.164c−0.086
Item 12−0.098a−0.005
Item 19−0.116a0.020
Item 35−0.118a−0.061
Item 36−0.0820.038
Item 48−0.157b−0.113a
Item 62−0.187c−0.145b
Item 103−0.0560.049

A model including age and sex was examined on step 1 of the hierarchical regression. Age and sex in combination accounted for a statistically significant proportion of the variance in NDI scores: F(2,395)=4.458, p=0.012, R2=0.022. Only age accounted for unique variance in the NDI: t(395)=2.974, p=0.003 (sr2=0.022). There were no sex differences in NDI; however, older participants were associated with higher NDI scores.

On step 2, the seven CBCL items predictive of psychiatric disorder20 were entered and added statistically significant incremental variance: inline imageChange=0.058; FChange(7,388)=3.476, p=0.001. An inspection of unique variance within the model for step 2 showed that among the CBCL items, only item 48 (‘Not liked’) accounted for statistically significant unique variance: t(388)=−2.504, p=0.013 (sr2=0.015). Participants who endorsed the item ‘Not liked’ were associated with a lower score on the NDI. The other CBCL items in the model were redundant in the sense that they did not explain a statistically significant proportion of the variance in NDI once the other predictors had been controlled for. The statistically significant zero-order correlations reported for items such as item 11 (‘Clings to adults’) and item 1 (‘Acts too young for his/her age’) were spurious in the sense that they were explained by the variance that they shared with the other predictors.

Items 36 (‘Gets hurt a lot, accident prone’) and 62 (‘Poorly coordinated or clumsy’) were added to the model on step 3 and significantly increased the variance accounted for: inline imageChange=0.021; FChange(2,386)=4.545, p=0.011. In the final model item 62 accounted for unique variance: t(386)=−3.014, p=0.003 (sr2=0.021). However, item 36 was a redundant predictor: t(386)=0.780, p=0.436 (sr2=0.001). In the final model, item 48 remained a significant predictor –t(386)=−2.44, p=0.020 (sr2=0.013) – as did age: t(386)=3.035, p=0.003 (sr2=0.021).

Discrimination accuracy

Table II shows the number of participants identified by item 62 as having motor impairment compared with those identified by the MAND. The sensitivity (a/[a+c]) of CBCL item 62 was 16.7% (95% confidence interval [CI] 10.3–25.6), whereas its specificity (d/[b+d]) for identifying the absence of motor impairment was 93.2% (CI 89.8–95.5). The positive predictive value of item 62 was 41.7% (a/[a+b]), indicating that fewer than half of the participants identified as having motor problems actually had them according to the MAND. Item 62 had a negative predictive value of 79.3% (d/[c+d]), indicating that 21% of the participants identified as not having motor problems did have them according to the MAND score. Overall, the discrimination accuracy of the CBCL item was low.

Table II.   Number of cases identified by the CBCL Item 62 (‘poorly coordinated or clumsy’) compared with the MAND
 MAND (NDI)
Motor impairmentNo motor impairmentTotal
  1. aTrue positive; bfalse positive; cfalse negative; dtrue negative. CBCL, Child Behavior Checklist; MAND, McCarron Assessment of Neuromuscular Development; NDI, neurodevelopmental index.

CBCL item 62
Motor impairment15a21b36
No motor impairment75c287d362
Total90308404

The sensitivity of CBCL item 62 combined with item 48 in identifying motor impairment was 31.1% (CI 22.4–41.2), whereas its specificity for identifying the absence of motor impairment was 85.7% (CI 81.3–89.1). The positive predictive value of combining items 48 and 62 was 38.9% (a/[a+b]), indicating that fewer than half of the total participants identified as having motor problems actually had them according to the MAND. Items 62 and 48 in combination had a negative predictive value of 81% (d/[c+d]), indicating that 19% of the participants identified as not having motor problems did have them according to the MAND score. Overall, the discrimination accuracy of the combined CBCL items was low.

Level of impairment and CBCL identification

Table III splits the participants into children identified by the MAND as having no impairment (NDI>85), mild impairment (NDI 71–85), and moderate impairment (NDI 55–70). It can be seen that CBCL item 62 identified only eight of the 20 children with a moderate impairment (40%) and only seven of the 70 children with a mild impairment (10%). When item 48 was included, this increased to 10 of 20 children with a moderate impairment (50%) and 18 of the 70 children with a mild impairment (25.7%).

Table III.   Number of cases identified by CBCL item 62 (‘poorly coordinated or clumsy’) and combined CBCL items 48 (‘Not liked’) and 62 (one or both items endorsed) compared with the NDI subcategories
 MAND (NDI)
Moderate MIMild MINo MITotal
  1. MAND, McCarron Assessment of Neuromuscular Development; NDI, neurodevelopmental index; MI, motor impairment; CBCL, Child Behavior Checklist.

CBCL item 62
 MI872136
 No MI1263287362
CBCL items 48 & 62
 MI10184472
 No MI1052264326
 Total2070308398

Discussion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

A quick and effective screening tool to identify motor impairment in children that is also reliable and valid would be helpful in clinical practice and in large research studies where individual assessment is costly and time-consuming. Two items from the CBCL, items 36 and 62, have been previously used as indicators of parent-reported motor problems in an ADHD-enriched sample of 851 child and adolescent twins.18 In that study, these CBCL items were associated with autistic-like social impairment, particularly in children with ADHD. This result was expected given previous reports of the association between motor problems and autistic symptoms, but it was unclear what type or degree of motor impairment the CBCL items might indicate. The current study partially supported the utility of item 62 as a significant predictor of motor ability. Item 36, however, was a redundant predictor and did not account for any unique variance in motor ability when considered in combination with item 62. The statistically non-significant result for item 36 reflected a weak effect size rather than low statistical power. After controlling for the other predictors, item 36 accounted for 0.001 of the variance in the NDI. This result suggested that item 36 does not provide a unique explanation of motor difficulties beyond the variance it shares with the items related to psychiatric disorders. The clinical significance of item 62, however, was also low for the current study population. The clinical outcome of using item 62 was a measure in which both the sensitivity (16.7%) and the positive predictive value (41.7%) were considerably lower than the acceptable level of 80% or more. Fewer than half of the participants identified from item 62 as having motor problems actually had these according to the MAND. Overall, the discrimination accuracy of the CBCL item was poor.

Item 62 was one of the eight items of the CBCL that Novik20 identified as being predictive of psychiatric disorders in a sample of 1170 children aged 4 to 16 years. By contrast, the current study showed item 62 to have poor sensitivity and positive predictive value for motor ability. We also found that another item, item 48 (‘Not liked’), explained unique variance in motor ability, although even when both items 48 and 62 were used to identify motor impairment, sensitivity and positive predictive value remained low at 31.1 and 38.9%, respectively.

A relationship between being liked and motor ability has been reported elsewhere. Chase and Dummer23 examined the determinants of social status in children and found that males rated athletic ability and physical appearance as the most important determinants of social status for males. Young females viewed physical appearance as the single most important determinant of social status for both males and females. It should be noted that children with poor motor ability are more often overweight,10 which may lead to low perceptions of their physical appearance. Other studies8 cite poor social support and poor peer integration in children with movement problems. Also, being described as ‘not liked’ may sometimes be a result of autistic-like social impairment, so the association between motor impairment and this CBCL item could be partly due to the high prevalence of motor impairment in autism spectrum disorders.

Novik20 also pointed out that both school psychologists and paediatricians in Norwegian schools now use the CBCL for screening behavioural and emotional problems, but warns that the predictive value depends on the prevalence of the disorder. Novik found a prevalence of 20% and suggested a that lower prevalence would reduce the predictive value. This may account for the low sensitivity and positive predictive values identified in the current study. Although the overall prevalence of motor impairment was 22.6%, 78% of these participants had only mild motor impairment. Only 5% of the total sample had moderate impairment and no individual had severe impairment. Also, caution should be used when speculating about the likely degree of motor impairment in the participants studied by Reiersen et al.18 since their sample was enriched for ADHD and the current sample was not. Future research should consider investigating the CBCL in a sample of children with more severe motor impairment, as well as in children with a variety of developmental disorders. Furthermore, some screening tools for motor impairment have been found to be less effective in children with developmental disorders such as ADHD, as symptoms may be misdiagnosed as clumsiness.24,25

In general, the use of screening instruments to identify motor impairment has had limited success. Recent instruments such as the DCDQ17 have focused on the performance of daily activities that require motor coordination to determine whether such activities have been disrupted. These behaviours are more easily recognized by parents and teachers and they generally cover several aspects of movement control. A recent investigation of the DCDQ25 found that this test had low specificity and sensitivity when the MAND was used as the criterion variable. However, the DCDQ was found to be accurate in identifying children with moderate or severe motor impairment. Since the current study did not include children with severe motor dysfunction on the basis of the MAND, it is still unclear how accurate the CBCL would be in identifying children with severe motor impairment. Given that the CBCL includes only the one item related specifically to motor ability, and given the very general nature of this item (i.e. ‘Poorly coordinated or clumsy’), it is not surprising that it lacks predictive validity in the sample studied here.

In conclusion, although the CBCL has excellent psychometric characteristics, is frequently used, and has well established norms, the current study does not support the use of the single ‘clumsy’ item to identify mild to moderate motor impairment in non-clinical populations.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This research was funded by the National Health and Medical Research Council of Australia. We wish to thank all the parents and children who participated in this study.

References

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • 1
    Kadesjo B, Gillberg C. Attention deficits and clumsiness in Swedish 7-year-old children. Dev Med Child Neurol 1998; 40: 796804.
  • 2
    Pitcher TM, Piek JP, Hay DA. Fine and gross motor ability in males with ADHD. Dev Med Child Neurol 2003; 45: 52535.
  • 3
    Dziuk MA, Gidley Larson JC, Apostu A, Mahone EM, Denckla MB, Mostofsky SH. Dyspraxia in autism: association with motor, social, and communicative deficits. Dev Med Child Neurol 2007; 49: 7349.
  • 4
    Crawford SG, Dewey D. Co-occurring disorders: a possible key to visual perceptual deficits in children with developmental coordination disorder? Hum Mov Sci 2008; 27: 15469.
  • 5
    American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Text Revision (4th edn). Washington: American Psychiatric Association, 2000.
  • 6
    Piek JP, Dworcan M, Barrett NC, Coleman R. Determinants of self-worth in children with and without developmental coordination disorder. Int J Disabil Dev Educ 2000; 47: 25972.
  • 7
    Rose B, Larkin D, Berger BG. Coordination and gender influences on the perceived competence of children. Adapt Phys Activ Q 1997; 14: 21021.
  • 8
    Skinner RA, Piek JP. Psychological implications of poor motor coordination in children and adolescents. Hum Mov Sci 2001; 20: 7394.
  • 9
    Piek JP, Rigoli D, Pearsall-Jones JG, et al. Depressive symptomatology in child and adolescent twins with attention deficit hyperactivity disorder and/or developmental coordination disorder. Twin Res Hum Genet 2007; 10: 58796.
  • 10
    Cantell MS, Crawford G, Doyle-Baker PK. Physical fitness and health indices in children, adolescents and adults with high or low motor competence. Hum Mov Sci 2008; 27: 34462.
  • 11
    Cermak SA, Gubbay SS, Larkin D. What is developmental coordination disorder? In: CermakSA, LarkinD, editors. Developmental Coordination Disorder. Canada: Delmar, 2002: 222.
  • 12
    Henderson SE, Sugden DA. Movement Assessment Battery for Children manual. New York: The Psychological Corporation/Harcourt, 1992.
  • 13
    Henderson SE, Sugden DA, Barnett AL. Movement Assessment Battery for Children, 2nd edn. London: Harcourt Assessment, 2007.
  • 14
    Bruininks RH. Bruininks–Oseretsky Test of Motor Proficiency Examiner’s Manual. Circle Pine, MN: American Guidance Service, 1978.
  • 15
    Bruininks RH, Bruininks BD. Bruininks–Oseretsky Test of Motor Proficiency, 2nd edn. Windsor, UK: NFER-Nelson, 2005.
  • 16
    McCarron LT. MAND McCarron Assessment of Neuromuscular Development: Fine and Gross Motor Abilities. Revised. Dallas, TX: Common Market Press, 1997.
  • 17
    Wilson BN, Kaplan BJ, Crawford SG, Campbell A, Dewey D. Reliability and validity of a parent questionnaire on childhood motor skills. Am J Occup Ther 2000; 54: 48493.
  • 18
    Reiersen AM, Constantino JN, Todd RD. Co-occurrence of motor problems and autistic symptoms in attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry 2008; 47: 66272.
  • 19
    Achenbach T. Manual for the Child Behaviour Checklist: 4–18 and 1991 Profile. Burlington, VT: University of Vermont, Department of Psychiatry, 1991.
  • 20
    Novik TS. Child Behavior Checklist item scores in Norwegian children. Eur Child Adolesc Psychiatry 2000; 9: 5460.
  • 21
    Dyck MJ, Hay D, Anderson M, Smith LM, Piek J, Hallmayer J. Is the discrepancy criterion for defining developmental disorders valid? J Child Psychol Psychiatry 2004; 45: 97995.
  • 22
    Tan SK, Parker HE, Larkin D. Concurrent validity of motor tests used to identify children with motor impairment. Adapt Phys Activ Q 2001; 18: 16882.
  • 23
    Chase MA, Dummer GM. The role of sports as a social status determinant for children. Res Q Exerc Sport 1992; 63: 41824.
  • 24
    Green D, Bishop T, Wilson BN, et al. Is questionnaire-based screening part of the solution to waiting lists for children with Developmental Coordination Disorder? Br J Occup Ther 2005; 68: 210.
  • 25
    Loh PR, Piek JP, Barrett NC. The use of the Developmental Coordination Disorder Questionnaire in Australian children. Adapt Phys Activ Q 2009; 26: 3853.