Psychometric study of the test of variables of attention: Preliminary findings on Taiwanese children with attention-deficit/hyperactivity disorder

Transcription

1 Psychiatry and Clinical Neurosciences (2007), 61, doi: /j x Regular Article Psychometric study of the test of variables of attention: Preliminary findings on Taiwanese children with attention-deficit/hyperactivity disorder YU-YU WU, md, 1 YU-SHU HUANG, md, 1 YU-YING CHEN, bs, 1 CHIH-KEN CHEN, md, phd, 2 TZYH-CHYANG CHANG, ma 3 AND CHIA-CHEN CHAO, phd 4 1 Department of Child Psychiatry, Chang Gung Children s Hospital, Linkou, 2 Department of Psychiatry, Chang Gung Memorial Hospital, Keelung, 3 Department of Occupational Therapy, Chang Gung University, Taoyuan, and 4 Graduate Institute of Clinical Behavioral Science, Chang Gung University, Taoyuan, Taiwan Abstract Attention-deficit/hyperactivity disorder (ADHD) is a common mental disorder in children. Unfortunately, reliable means of measuring attention and impulsivity to help with diagnoses are scarce.the test of variables of attention (TOVA) is a computer-administered continuous performance test measuring attention and impulsivity, designed to avoid confounding arising from language processing skills or short-term memory problems. Some evidence has indicated the TOVA can be useful in diagnosing ADHD. This study examines its validity and reliability in helping diagnose Taiwanese ADHD children. The study included 31 ADHD children (24 males, seven females) from a northern Taiwan children s hospital and 30 normal controls (18 males, 12 females) from the local community. The TOVA and the Child Behavior Checklist (CBCL) were administered to all children. TOVA scores for omissions, commissions, response time, response time variability, D and ADHD scores were analyzed. Results showed a mean internal consistency of 0.81 for all six TOVA variables across conditions, with moderate convergent and discriminant validities. Groups showed significant differences in response time variability, D and ADHD scores, with the normal group outperforming the ADHD group. Significant group differences were also found in all CBCL subscale scores except somatic complaints. The ADHD group obtained a clinically significant score on the hyperactivity subscale of the CBCL. The findings partially support the usefulness of the TOVA in assessing attention and impulsivity problems for a Taiwanese sample. Future studies should increase the sample size, use multiple measures, and collect behavior ratings from both parents and teachers. Key words attention-deficit/hyperactivity disorder, reliability, Taiwan, test of variables of attention, validity. INTRODUCTION Attention deficit/hyperactivity disorder (ADHD) is a behavior abnormality commonly seen in childhood and adolescence. It includes three core symptoms: inattention, hyperactivity, and impulsivity, and patients can be Correspondence address: Chia-Chen Chao, PhD, Graduate Institute of Clinical Behavioral Science, Chang Gung University, 259, Wen-Hwa 1st Road, Kweisan, Taoyuan, Taiwan ccchao54@yahoo.com.tw Received 18 February 2005; revised 10 January 2007; accepted 4 February classified into three subtypes: inattentive, hyperactiveimpulsive, and combined. 1 The prevalence of ADHD is estimated at 3 5% of the population in the USA, 1,2 though the rate varies in other parts of the world. For example, it is estimated at about % in Taiwan, % in China, 4 2.4% in Australia, 5 and 4% in Japan. 6 Despite abundant research on ADHD, its pathophysiology is still unclear. Indeed, it is quite difficult to make a diagnosis of ADHD based on a single procedure, observation, or behavioral characteristic. Due to its clinical heterogeneity, many specialists believe that ADHD is of multiple etiologies. Previous studies

2 212 Y.-Y. Wu et al. have looked into ways of measuring activity such as stabilimetric chairs, 7 actometers, 8 grid-marked room measures, 9 ultrasound and photoelectric cells, 10 and actigraphs. 11 Unfortunately, these techniques have not been widely used because of their poor reliability and high cost. 12 Instead, the continuous performance test (CPT) 13,14 and matching familiar figures test 15 are often used to assess inattention and impulsivity, as they have been considered more sensitive and reliable. In all versions of the CPT, a variety of stimuli are presented to subjects who are instructed to respond to a target stimulus. 13,14,16 19 For example, in the A-X CPT task, a response is required only if a specific stimulus (A) is followed by another specific stimulus (X). Most CPTs use English letters or numbers as stimuli. The tasks on these tests generally confound attention problems, short-term memory difficulties, language processing skills, and/or other cognitive deficits In addition, there is considerable variation in the duration of CPT tasks, both in terms of the number of trials presented and the times required for task completion. 20 Greenberg developed the test of variables of attention (TOVA), formerly known as the Minnesota computer assessment, for the assessment of attention and impulsivity. 14,20 24 It is a computer-administered, visual or audio CPT. Normative data from the TOVA has been collected from thousands of male and female children and adults. 14,23,24 Psychometric studies indicate that the TOVA has proven reliable and demonstrated adequate sensitivity and specificity. 24 Unlike other CPTs, the TOVA was carefully designed to be free of many confounding variables. 23 It uses simple geometric forms as both target and non-target stimuli to minimize the effects of cultural differences, language demands, short-term memory difficulties, and language processing skills. 23,24 Moreover, the test requires no left right discrimination. 21 Many features of the TOVA, such as its relatively brief stimuli presentations, shorter interstimulus intervals, and longer total testing time (a total of 21.6 min compared to 15 min in other CPTs), require better sustained attention and are more likely to accurately represent real-world demands in the workplace These features are particularly important because they might be the most effective parameters in differentiating ADHD and normal children. 19,21 With the above-mentioned advantages, it seems promising that the TOVA will be a useful contribution to the clinical diagnosis of ADHD. This study intends to examine the application of the TOVA to ADHD children in Taiwan and to evaluate the reliability and validity of the TOVA as a diagnostic tool for ADHD. METHOD Subjects The sample consisted of 31 ADHD children (aged 5 13 years) from the child psychiatry clinic of a medical center in northern Taiwan. Their diagnoses were made by two senior child psychiatrists based on the DSM-IV criteria 1 and cross-validated by another child psychiatrist using the standardized semi-structured interview the Chinese version of the Schedule for Affective Disorder and Schizophrenia for School-Age Children Epidemiologic Version (Chinese K-SADS-E). 25,26 Physical and neurological examinations were performed to rule out central nervous system diseases such as brain trauma, epilepsy, and others. Children with mental retardation, developmental disturbances, severe psychiatric disorders, and other mental problems were also excluded. All the ADHD subjects were free from any medication for at least 1 week before being tested. Thirty normal children (aged 5 13 years) were recruited from the local community to serve as controls. The inclusion criteria were: (i) no symptoms of ADHD; (ii) no developmental or severe psychiatric disorders; and (iii) no health problems reported by parents. Measures Test of variables of attention visual The test of variables of attention visual (TOVA-V) 23,24 is a computerized CPT that uses colored squares with a small square adjacent to either the top or bottom edge as visual stimuli. The stimulus with the inner square near the top edge is the target and the one with the inner square near the bottom edge the non-target. The subject is instructed to respond by pressing a hand-held microswitch whenever the target stimulus appears, and not to respond when the non-target stimulus is shown on the screen. Each stimulus is randomly presented for 100 msec every 2 s. The total duration of the TOVA is 21.6 min. The target stimulus is presented in 22.5% of the trials during the first half (condition 1, stimulus infrequent condition) and in 77.5% of the trials during the second half (condition 2, stimulus frequent condition) of the test. This design of varying target nontarget ratio examines the effects of differing response demands on attention and impulsivity. The first and second halves of the test are further divided into two even parts for a total of four quarters to the test, to facilitate statistical analyses. Each quarter lasts 5.4 min. The variables measured in the TOVA-V include: 1 Omission errors: this score is measured as the failure to respond to the target stimulus. Omission error

3 TOVA performance of Taiwanese children 213 scores are presented as percentages and are considered to be a measure of inattention. 2 Commission errors: this score is measured as an inappropriate response to the non-target stimulus. Commission error scores are presented as percentages and are considered to be a measure of impulsivity or disinhibition. 3 Response time (RT, in msec): this score is measured as the average of the correct response times. This score indicates response latency in information processing and motor response speed. 4 Response time variability (RTV): this score is measured as the standard deviation of the mean of correct response times. It is a measure of the subject s inconsistency in response times. 5 D score: this score is a response sensitivity score reflecting the ratio of the hit rate to false alarm rate. It refers to the accuracy of target and non-target discrimination and is interpreted as a measure of perceptual sensitivity. 6 ADHD score: this score is a composite score generated by the TOVA program. It is calculated by comparing an individual s performance on the TOVA (i.e. RT, RTV, D score) to those of an ADHD sample collected by the authors of the TOVA.The score tells how similar an individual s performance is to the ADHD profile. Child behavior checklist The child behavior checklist (CBCL) has been widely used in screening children s behavioral and emotional problems in research and clinical practice in more than 30 countries. 29 It was originally developed by Achenbach and Edelbrock to evaluate children s competencies and problems as reported by parents or parent surrogates in a standardized format. 27 Its problem scale (118 items) is designed to assess seven narrow-band syndromes (i.e. depression/anxiety, thought/obsession, somatic complaints, social withdrawal, hyperactivity, aggressive behavior, and delinquent behavior). Each item is scored as 0 (if the problem is not true of the child), 1 (if the problem is somewhat or sometimes true), or 2 (if it is very true or often true). A Chinese version of the CBCL has been used in a variety of studies. 30,31 In general, these studies have shown evidence of promising reliability and validity of the CBCL in Chinese cultures. Procedure After careful screening, all the selected subjects and their parents were provided with a clear description of the present study. Informed consent was obtained from all the parents prior to the assessment. The TOVA was administered to both the ADHD and control groups, and the CBCL was completed by their parents. Several factors were taken into consideration when evaluating the reliability of the TOVA. First, it is considered inappropriate to evaluate Cronbach alpha, split-half, and Kuder-Richardson reliability for timed tasks such as those used with the TOVA. 23,24 Second, the design of the TOVA involves manipulation of stimulus frequency; that is, in the first half of the TOVA the target stimulus occurs infrequently (condition 1, including quarters 1 and 2), while in the second half of the TOVA the target stimulus occurs frequently (condition 2, including quarters 3 and 4). Therefore, the Pearson product-moment correlation coefficients (r) for internal consistency were computed for all variables across the two conditions (condition 1 and condition 2). Two kinds of validity were examined for the TOVA. First, the discriminant validity was examined by comparing the TOVA scores between the ADHD and the control group to see if the TOVA scores could discriminate these two groups. Second, the construct validity of the TOVA was examined by calculating the correlations between the ADHD score of the TOVA and the CBCL subscale scores. The convergent validity was estimated from the correlation between the ADHD scores of the TOVA and the hyperactivity subscale score of the CBCL, using the ADHD group. The discriminant validity was estimated from the correlations between the ADHD scores of the TOVA and all the CBCL subscale scores, except hyperactivity, using the ADHD group. Data analysis Descriptive statistics was used to organize the demographic data. A c 2 test and an independent t-test were used to compare age and gender between groups. Pearson moment-product correlations were computed to estimate the reliability and validity of the TOVA. The raw scores of the six TOVA measures were first converted to standard scores (z score) based upon normative data for age and gender. Then manova and an independent t-test were conducted to compare performance on the TOVA and the CBCL scores between groups. RESULTS Demographic characteristics Both ADHD and control groups were equivalent in terms of age and gender distribution. The mean age of

4 214 Y.-Y. Wu et al. Table 1. Variable Demographic characteristics of ADHD and control groups ADHD (n = 31) Control (n = 30) n (%) Mean SD n (%) Mean SD t/c 2 Age (years) Gender 2.16 Boys 24 (77.4) 18 (60.0) Girls 7 (22.6) 12 (40.0) ADHD subtype Inattentive 8 (25.8) Hyperactive 4 (12.9) Combined 19 (61.3) Comorbidity LD 7 (22.6) ODD 3 (9.7) ODD+CD 2 (6.5) Independent t-test was used here. c 2 test was used here. LD, learning disorder; ODD, oppositional defiant disorder; CD, conduct disorder. Table 2. The reliability coefficients for all variables within condition 1 (raw scores) Time segment Omission Commission RT RTV D Q1:Q2(n = 54) 0.807*** 0.322* 0.937*** 0.553*** 0.393** Q1:H1(n = 54) 0.930*** 0.934*** 0.984*** 0.874*** 0.733*** Q1:T(n = 61) 0.761*** 0.552*** 0.893*** 0.741*** 0.623*** Q2:H1(n = 54) 0.950*** 0.639*** 0.984*** 0.878*** 0.788*** Q2:T(n = 54) 0.802*** 0.627*** 0.913*** 0.687*** 0.586*** H1:T(n = 54) 0.848*** 0.627*** 0.918*** 0.827*** 0.696*** * P < 0.05, ** P < 0.01, *** P < Seven subjects in this study were under 6 years old.they were administered a shorter version of TOVA, consisting of only one quarter for each condition (stimulus infrequent and stimulus frequent), Q1 and Q3. Therefore, their scores for Q2 and H1 cannot be calculated here. Condition 1, Stimulus infrequent condition, including Q1 and Q2; H1, half 1, the first half of elapsed time of the TOVA; Q1, quarter 1, the first quarter of elapsed time of the TOVA; Q2, quarter 2, the second quarter of elapsed time of the TOVA; RT, response time; RTV, response time variability; T, total elapsed time of the TOVA. ADHD children was 8.1 years ( 2.11), while that of the control group was 8.3 years ( 2.31). The distribution of ADHD subtypes in the 31 ADHD children was: eight (25.8%) inattentive type, four (12.9%) hyperactive type, and 19 (61.3%) combined type. Twelve (38.7%) ADHD children also had other comorbid psychiatric problems. The demographic characteristics of the sample are presented in Table 1. Reliability Tables 2 and 3 provide the reliability coefficients for the six TOVA variables within conditions 1 and 2, respectively. All 60 correlation coefficients reached significant levels (P < 0.05), with a range of 0.31 to 0.99 and a mean of This finding demonstrates an overall satisfactory internal consistency for the TOVA. Validity The results showed that ADHD scores correlated significantly with hyperactivity scores on the CBCL (r =-0.42, P < 0.05), but not with other CBCL subscale scores (i.e. depression/anxiety, thought/obsession, somatic complaint, social withdrawal, aggressive behavior, delinquent behavior). This finding demonstrates adequate convergent and discriminant validity for the TOVA.

5 TOVA performance of Taiwanese children 215 Table 3. The reliability coefficients for all variables within condition 2 (raw scores) Time segment Omission Commission RT RTV D Q3:Q4(n = 54) 0.921*** 0.743*** 0.919*** 0.805*** 0.320* Q3:H2(n = 54) 0.978*** 0.923*** 0.979*** 0.934*** 0.439*** Q3:T(n = 61) 0.980*** 0.853*** 0.977*** 0.919*** 0.314* Q4:H2(n = 54) 0.982*** 0.943*** 0.980*** 0.960*** 0.905*** Q4:T(n = 54) 0.949*** 0.802*** 0.974*** 0.950*** 0.752*** H2:T(n = 54) 0.981*** 0.851*** 0.994*** 0.987*** 0.816*** * P < 0.05, ** P < 0.01, *** P < Seven subjects in this study were under 6 years old.they were administered a shorter version of TOVA, consisting of only one quarter for each condition (stimulus infrequent and stimulus frequent), Q1 and Q3. Therefore, their scores for Q4 and H2 cannot be calculated here. Condition 2, Stimulus frequent condition, including Q3 and Q4; H2, half 2, the second half of elapsed time of the TOVA; Q3, quarter 3, the third quarter of elapsed time of the TOVA; Q4, quarter 4, the fourth quarter of elapsed time of the TOVA; RT, response time; RTV, response time variability; T, total elapsed time of the TOVA. Table 4. Variable Comparison of TOVA scores between ADHD and control groups ADHD Control n Mean SD n Mean SD t Omission Commission RT RTV ** D * ADHD score ** * P < 0.05, ** P < 0.01, *** P < RT, response time; RTV, response time variability. Comparison of TOVA performance between ADHD and control groups As expected, the ADHD group performed worse than the control group on all the TOVA variables: omissions, commissions, response time, response time variability, D score, and ADHD score (Table 4). However, only on the last three variables (i.e. response time variability, D score, and ADHD score) did differences reach statistically significant levels. Further analysis of TOVA performance over the four quarters revealed that, compared to the control group, there was significantly more variability in the performance of the ADHD group over time. Comparison of CBCL scores between ADHD and control groups Parents of ADHD children, compared to their counterparts, reported significantly more problems in their children. Table 5 shows ADHD children received significantly higher ratings on all CBCL subscales (i.e. depression/anxiety, thought/obsession, social withdrawal, hyperactivity, aggressive behavior, delinquent behavior, internalizing problems, and externalizing problems), except somatic complaints. Within the ADHD group, it was found that only hyperactivity scores reached the clinical cut-off (i.e. T = 70) among all the CBCL subscales, thus supporting the diagnosis for the ADHD group. DISCUSSION The 8 12% prevalence rate of ADHD in Taiwan 3 is relatively high compared with the rate in other areas. Because clinical diagnosis of ADHD often relies on parental and/or teacher report, cultural differences in the adult perception of the child s behavior might have affected the objectivity of such a diagnosis. In addition, research evidence reveals only a modest

6 216 Y.-Y. Wu et al. Table 5. Subscale Comparison of CBCL subscale scores between ADHD and control groups ADHD Control n Mean SD n Mean SD t Depression/anxiety *** Thought/obsession ** Somatic complaints Social withdrawal ** Hyperactivity *** Aggressive behavior *** Delinquent behavior *** Internalizing problems *** Externalizing problems *** * P < 0.05, ** P < 0.01, *** P < correspondence between different informants. Therefore, it is important to apply objective measures in evaluating ADHD. When comparing our findings with data from other cultures, some consistency emerges. For example, using American samples, Forbes reported that omission errors, response time, and response time variability were higher in ADHD group than in disease controls (i.e. children with oppositional defiant disorder, conduct disorder, learning disabilities, adjustment disorders, or depression). 21 Using Japanese samples of ADHD children and age-matched normal controls, Wada et al. reported significant differences between ADHD and control males for all TOVA variables (i.e. omission errors, commission errors, response time, response time variability, anticipatory responses, and multiple responses). 20 Similar results have been found in our Taiwanese samples; the normal group performed better than the ADHD group on all TOVA variables, although only on three of them (response time variability, D score, and ADHD scores) did differences reach statistically significant levels. It is worth noting that compared to their normal counterparts, the ADHD group received higher ratings from their parents on most of the problem scales of the CBCL. This might suggest a higher risk of other psychiatric problems in ADHD children. Consistent with our hypothesis, among all the problem scales, only the hyperactivity score reached the clinical cut-off (T = 70), thus supporting the diagnosis for the ADHD group. Furthermore, findings that the TOVA ADHD scores correlate significantly with only hyperactivity scores but not with other problem scale scores in the CBCL supports the TOVA s validity as a tool specifically for assessing attention and impulsivity problems. This study has a number of limitations. First, the number of subjects in each group was relatively small, which may have been the reason that some of the TOVA variables were not significantly different between ADHD group and normal controls. Second, this study employed a broad-band measure (the CBCL) to evaluate the subjects general psychopathology in order to examine the convergent and discriminant validity of the TOVA. However, the inclusion of a narrow-band scale specifically designed to assess ADHD symptoms in a future study might help distinguish the ADHD and controls more clearly. Third, this study has only analyzed behavior ratings from parents because teacher ratings were not available for most subjects. Only 70% of parents in control group completed the CBCL, thus further reducing the efficacy of a group comparison. It is recommended that future studies include a larger sample of ADHD and normal subjects of both genders and a wider range of age levels in order to study the diagnostic and medication response sensitivity of TOVA in Taiwanese ADHD children and adults. Multiple measures including both broad-band and narrow-band rating scales should be implemented and behavior ratings from both parents and teachers collected in order to examine the relationships between the TOVA and parent or teacher ratings of overall problems as well as specific symptoms of inattention, hyperactivity, and impulsivity. Finally, it is essential to remember that the TOVA alone, like any other CPT, lacks sufficient diagnostic power to constitute the exclusive determinant of a diagnosis. TOVA scores are indicators of inattention, hyperactivity, and impulsivity, but such symptoms are common to pervasive developmental disorder, mental retardation, learning disabilities, reactive attachment disorder, and a variety of medical and mental problems along with ADHD. Nonetheless, the results of this study suggest that the TOVA can provide objective, reliable, and valid information as well as make a meaningful contribution to the clinical assessment of inattention

7 TOVA performance of Taiwanese children 217 and impulsivity problems for a Taiwanese sample. With regard to ADHD children in Taiwan, the TOVA has proven useful and reliable as one of the multiple sources and types of information on which any clinical diagnosis should be based. ACKNOWLEDGMENTS This research was supported by Chang Gung Memorial Hospital, Taiwan, Republic of China (CMRP 993, CMRP 1237, and CMRPG 32015). Our appreciation to Drs Shur-Fen Gau and Wei-Tsuen Soong for granting us permission to use the Chinese K-SADS-E in this study. Thanks to Hanvey Hsiung and Frank Sharp for their assistance in editing this manuscript. Deepest gratitude goes to all the children and their parents who made this project possible by sharing their experiences with us. REFERENCES 1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association, Washington DC, Barkley RA. Attention Deficits Hyperactivity Disorder: A Handbook for Diagnosis and Treatment, 3rd edn. Guilford Press, New York, Tzang RF, Wu KH, Liou CP. Prevalence of attentiondeficit/hyperactivity disorder in a Taiwanese elementary school. Taiwanese J. Psychiatry 2002; 16: (in Chinese). 4. Tao KT. Hyperactivity and attention deficit disorder syndromes in China. J. Am. Acad. Child Adolesc. Psychiatry 1992; 31: Gomez R, Harvey J, Quick C, Scharer I, Harris G. DSM-IV AD/HD: Confirmatory factor models, prevalence and gender and age difference based on parent and teacher ratings of Australian primary school children. J. Child Psychol. Psychiatry 1999; 40: Robison LM, Sclar DA, Skaer TL, Galin RS. National trends in the prevalence of Attention-Deficit/ Hyperactivity Disorder and the prescribing of methylphenidate among children: Clin. Pediatr. 1999; 38: Tyron WW. Principles and methods of mechanically measuring motor activity. Behav. Assess. 1984; 6: Conners CK, Kronsberg S. Measuring activity level in children. Psychopharmacol. Bull. 1985; 21: Barkley RA, Ullman DG. A comparison of objective measures of activity and distractibility in hyperactive and nonhyperactive children. J. Abnorm. Child Psychol. 1975; 3: Johnson CF. Limits on the measurement of activity level in children using ultrasound and photoelectric cells. Am. J. Ment. Defic. 1972; 77: Colburn TR, Smith BM, Guarini JJ, Simmons NN. An ambulatory activity monitor with solid state memory. ISA Trans. 1976; 15: Inoue K, Nadaoka T, Oiji A et al. Clinical evaluation of Attention-Deficit Hyperactivity Disorder by objective quantitative measures. Child Psychiatry Hum. Dev. 1998; 28: Beck LH, Bransome E Jr, Mirsky AF, Rosvold HE, Sarason I. A continuous performance test of brain damage. J. Consult. Psychol. 1956; 20: Greenberg LM, Waldman ID. Developmental normative data on the test of variables of attention (T.O.V.A.). J. Child Psychol. Psychiatry 1993; 34: Kagan J. Reflection-impulsivity: The generality and dynamics of conceptual tempo. J. Abnorm. Psychol. 1966; 71: Epstein JN, Erkanli A, Conners CK, Klaric J, Costello JE, Angold A. Relations between continuous performance test performance measures and ADHD behaviors. J. Abnorm. Child Psychol. 2003; 31: Halperin JM, Matier K, Bedi G, Sharma V, Newcon JH. Specificity of inattention, impulsivity, and hyperactivity to the diagnosis of attention-deficit hyperactivity disorder. J. Am. Acad. Child Adolesc. Psychiatry 1992; 31: DuPaul GJ, Anastopoulos AD, Shelton TL, Guevrement DC, Metevia L. Multimethod assessment of attentiondeficit hyperactivity disorder: The diagnostic utility of clinic-based test. J. Clin. Child Psychol. 1992; 21: Corkum PV, Siegel LS. Is the continuous performance task a valuable research tool for use with children with attention-deficit-hyperactivity disorder? J. Child Psychol. Psychiatry 1993; 34: Wada N, Yamashita Y, Matsuishi T, Ohtani Y, Kato H. The test of variables of attention (TOVA) is useful in the diagnosis of Japanese male children with attention deficit hyperactivity disorder. Brain Dev. 2000; 22: Forbes GB. Clinical utility of the test of variables of attention (TOVA) in the diagnosis of attention-deficit/ hyperactivity disorder. J. Clin. Psychol. 1998; 54: Greenberg LM. An objective measure of methylphenidate response: Clinical use of the MCA. Psychopharmacol. Bull. 1987; 23: Leark RA, Dupuy TR, Greenberg LM, Corman CL, Kindschi CL. TOVA R : Professional Manual (Version 7.0). Universal Attention Disorders, Los Alamitos, CA, Greenberg LM, Kindschi CL. TOVA R : Clinical Guide. Universal Attention Disorders, Los Alamitos, CA, Puig-Antich J, Chambers WJ. The Schedule for Affective Disorders and Schizophrenia for School-Age Children- Epidemiologic Version (K-SADS-E). New York State Psychiatric Institute, Biometrics Research, New York, Gau SF, Soong WT. Psychiatric comorbidity of adolescents with sleep terrors or sleepwalking: A case-control study. Aust. NZ. J. Psychiatry 1999; 33:

8 218 Y.-Y. Wu et al. 27. Achenbach TM, Edelbrock C. Manual for the Child Behavior Checklist and Revised Child Behavior Profile. University of Vermont, Research Center for Children, Youth & Families, Burlington, VT, Achenbach TM. Manual for the Revised Child Behavior Checklist. University of Vermont, Department of Psychiatry, Burlington, VT, Achenbach TM, Ruffle TM. The Child Behavior Checklist and related forms for assessing behavioral/emotional problems and competencies. Pediatr. Rev. 2000; 21: Huang HL, Chuang SF, Wang YC. Developing the multiaxial behavioral assessment of children in Taiwan. In: Chinese Assessment Association (ed.). Psychological Assessment in Chinese-Speaking Society. Psychological Press, Taipei, Taiwan, 1994; (in Chinese). 31. Yang HJ, Soong WT, Chiang CN, Chen WJ. Competence and behavioral/emotional problems among Taiwanese adolescents as reported by parents and teachers. J. AM. Acad. Child Adolesc. Psychiatry 2000; 39: