I? “‘91"? ‘A $1.1 5: “A?" 1.11% IAIN" IA} “Aw; W'IW 'A' :IN.’ (I'Atz‘ ‘iI‘AEi . Aug “(2" .. A ‘ffiuhp‘lj :1“ ”3);: W191}?! zuffl . AA..:AAAAA1AA flak“ W :31 51:1 1‘ . {AAA-1W. AA'A' .2 AAAAfi;A1'-A..:Aat.r I‘A ‘9‘“ II A" “"9“ 1.13“???“ I‘A' 1I .A-AAALLAMAgA 1 .... AA A . A A; A A: AAAAA‘AA A'I'flr-J“ ( M .3‘ l A‘ ”I: It“ ‘EI-fiuf’ffjd '5. u I}. [021‘ 1“}.wa in“ N A“ \L‘ 3.1», . A 4’“, 3:41:45:— ;5 Afl p.“ 4.153;: ‘w,» a ‘ vul- ._— _' _‘ dfl~— p—P J —o— .4 AA" A AA ISIAEA v. M .ffif—r" ,st‘twé-v ~ - 1—:— “4,-.- ‘2- : .. . ‘l/ - m ' W_ ., u 1: . ‘- ‘Tn. o.<'_ ,I ' .I.’ .' . ~ | i€Z¢IR ' .. . 4. r .4 ”:3 '2. .- :- ‘ .4- — OF.‘ , Lav - 3.31—3“ ,uyir-gc‘; .L.- fi".‘. *- ;.-—'- . .-~ 112.4. I ...- Wharf!"- . ...u'~ ...__ ”r, “at“: vii 3A31 “.1: I”. .qu .A ‘U‘. 13%? 3““ II?“ \A. ‘3'? . I‘D .- 1. AA: AA 1 \ .. 2'0“} ‘ l I . A; AWN”? f’v‘: .4 '1: ' - ”#1353?- _ fl) .d-:‘ .4 -n v __., «~HJ- ¢ ,1..-» ' W .i—o. . ~ .4 ._ .4. .. .. "E‘E‘E‘ 1"“, ”rm-17"» ”r &1L':.-‘~1 ,0 - ‘1 ‘ I...- c 1 a k .. H a! ‘ ‘ 4 . ”n 4 , £"‘B"' . ' — -4 " _' fix" . f“, _' may. ’ n . ~ . a Ih 4 ‘~ . ‘ €“" :3! " 1' ’ c‘ a l I). ‘7 ; ‘ I‘ RI 5:: IAA“‘ " '- 1.1.3... I . .: ‘ ‘ ,A' a A‘ {II " n “.3: '9 ‘ A . " . ‘7 4 0" ' ..‘ Z'. 4 ‘. l‘ g. ‘ 1 I: ' .0 $3: — e“. s ‘ l.‘ J . ,4”- 4~—v—- rw ' ~_.' ol“ ,2.“ .-"':J:é. -.. _ N”. 71:4, 4 -— ‘5’.— 4—, — 4'4"???" fig. * . 3.x 1 - 4-_ 13 AU’ :‘Stfijgi‘ .. 4 “.4 ‘b‘n’:S-.‘. .‘:i';‘)i 1“ . A: ' i ‘ “ ‘A AA ’='=.- :21'211‘4 "‘5‘- ‘ I A): . “.l'. -'-.'j':|l“‘ ‘AA-YA' A . A: :.'A‘. AAA, ' «A: 1 .; «AA. A‘ ,Ji 7“: 11A it AAA.n..1“~AAA~:'A~A 55'- 4 l Ti..A1.A “IA.“ 13‘“; A:1}:'“AAWWI I ':~:“““ 1".“ ‘ 4‘ ”"1 I ~:A"'3‘ 1p ,AfiIAt. 1PS5 , 4:35.11 . 1A“: A: Ad, . Wynn“... 35."... ,- (A. -,,. r . “I11 .111 . . . A‘Efi-..:~i$mj{A .L';*h‘"5ri 1‘5‘ .. ""“ HA ““‘ AA Afi.~3."5’~33 ?'.‘:le: ’Etfifé' ' IA‘. .1 " A ‘AI‘ E‘J MA‘AAA'FF'AW: [DA «1:: 1L1". A As“ .._ mA. A11“ ‘1! AI “-1..." A EFAAI ll'IIHTNA'.I\['O\I ”L'- AAI . . . '-'A“:‘-A:'».:2:::A ;_ m: ”A: 113?“; . .5 A m‘AA. 1I‘..“AA?Ai’:' Ewan... J} ‘1‘ ‘ ”(qu? ‘ H" m :51” 131'... I“ ‘.~‘ “4“ ‘1‘ Am" ““'I‘ ‘A' . "A.“ ‘I IAIclII' 1"A'I'll1I1 IL'ILAAAO' (L: ‘ ‘... .1: .HA I 1" A)" ,1" 1 AA A = A .AA A A A A AA AAA .., .. A AA. :A-At'AAAAAAA-r: .. ' ..s I “‘1' “"1““ '“" AHIA‘ .13]! Al I A... A 1‘. A A 1““"'A "WA '11" , IIIA’ "' I. i' I ' A 1.1“." I.:" 'I‘ " A“: .- A.‘ "A‘A‘."3"'|IA ' .AA A ‘.‘ 'ml M" II" .31'. '.I' AH A.“ AA. HI I ' AAA I A)"; ‘3' HI“ -.3. ‘.-..-. .I'AI‘AI‘: 1 [A‘AAA'I'AI‘II‘A "‘I‘ "‘I‘AIAAAA‘I‘ m11‘1IANL‘3‘IA1MII'I‘V‘AIIX 1:" m “0“‘A5qlu‘h‘qul .h‘IJ'HI; ‘A. ‘ ”A" “A | ‘J‘All‘ '“A A “I‘m“ “" ‘ ‘ I I“ 'I‘l‘lfllm I‘M" IN Ag‘.‘lul"""AA'M‘A‘lv‘é :IQIAU‘ "'LAAAAAA‘AAIAAAAA'AAAAAAA‘ AAIIIAAA‘AAI‘AA AAA “AAA .‘AA’AAAA AHAAAAAA‘AA‘IAAAAAAA'AAAA (II: A‘AAAAAH‘” A‘AAA‘A‘AA A‘é‘ufi’flé'flb‘fi-‘zi'MAA A. Ill i \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\i 1293 10429 3893 33,3. , 33.. w73w¢ 3 “rm "' ' THESIS III _._,._. f... _..-——-==—" ”—— g _,_,—-._-=2 ,— This is to certify that the dissertation entitled Construct Vaiidity and Diagnostic Potential of the Person- ality Inventory for Chiidren (PIC) with Emotionaily Dis- turbed, Learning Disabied, and Educabie Mentaiiy Retarded Chiidren presented by E1aine Ciark has been accepted towards fulfillment of the requirements for Ph.D. degreein Educational Psychoioqy @«EW Major professor Date ,2. //~ 5,; MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to 3 remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. // [g1 CONSTRUCT VALIDITY AND DIAGNOSTIC POTENTIAL OF THE PERSONALITY INVENTORY FOR CHILDREN (PIC) WITH EMOTIONALLY DISTURBED, LEARNING DISABLED, AND EDUCABLE MENTALLY RETARDED CHILDREN By Elaine Clark A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling and Educational Psychology 1982 ABSTRACT Construct Validity and Diagnostic Potential of the Personality Inventory for Children (PIC) with Emotionally Disturbed, Learning Disabled and Educable Mentally Retarded Children by Elaine Clark This study investigated the ability of the Personality Inventory for Children (PIC) to discriminate cognitive and personality patterns among l4l learning disabled, emotionally disturbed, and educable men- tally retarded children, ages 6 to l6. In addition, the construct validity of the PIC was studied by examining the relationships among the PIC profile scales and independently derived measures of intellec- tual functioning and behavioral adjustment. Profile analysis resulted in the significant discrimination among the three groups' profiles. Subsequent analyses of variance showed that relative to a combined group of learning disabled and emotionally disturbed children, the educable mentally retarded children had sig- nificantly higher scores (indicating greater impairment) on the Lie, Intellectual Screening, Achievement, Development, Psychosis, and With- drawal scales, and significantly lower scores on the Hyperactivity scale. The emotionally disturbed children, compared to the learning disabled, had significantly higher scores on the Adjustment, Delin- quency, Anxiety, and Social Skills scales, and significantly lower scores on the Lie and Defensiveness scales. These findings suggest Elaine Clark that these educationally handicapped children can be differentiated along salient intellectual and personality dimensions, as measured by the PIC. A series of disCriminant function analyses provided further support for this contention by showing that two functions, cognitive and emotional/behavioral, separated the groups, and that the PIC which assesses both is a relatively effective instrument in making these discriminations. The correlation among the PIC profile scales and the Nechsler Intelligence Scale for Children--Revised and the Teacher Rating and School Information checklist showed that the profile scales are re- lated to variations in the intellectual abilities and classroom be- havior of children diagnosed as learning disabled, emotionally dis- turbed, and educable mentally retarded. The majority of significant correlations clustered around scales which discriminated among the three groups and identified cognitive deficits (Intellectual Screening, Development) and disruptive, acting-out behaviors (Adjustment, Hyper- activity). Examination of the scale correlates provides substantial support for the convergent and discriminant validation of the PIC profile scales. Copyright by ELAINE CLARK 1982 ACKNOWLEDGMENTS I would like to express my sincere gratitude to Dr. Donna Nanous, the director of this dissertation, who spent countless hours critiquing the research and sharing her expert knowledge of statistics and research design. Without her guidance, as well as her patient support and con- fidence in my ability as a researcher, this dissertation would still be incomplete. Dr. Don Hamachek, my committee chairperson, has assisted me through- out my doctoral program and has encouraged me to demonstrate high pro- fessional standards. Dr. William Mehrens contributed to this disserta- tion through his expertise in measurement theory, while Dr. Gary Stollak provided insights about parental perceptions. These four scholars have added to both my educational and professional growth, and, for this, I am most grateful. I have been fortunate to have the assistance of Dr. Charles Gdowski of Lafayette Clinic. He has provided valuable suggestions, as well as an enthusiasm for research with the Personality Inventory for Children. A special note of thanks is due a friend and colleague, Patricia Carrico, who volunteered numerous hours assisting in the data collection process. Janiece Pompa also deserves special mention for her painstak- ing review of this manuscript. I am also thankful for the research assistance received from Jacqueline Smith, Vikas Kapil, Sondra Clark, and Carol Blumberg. 3 To the parents and teachers who participated in this study, I am extremely appreciative. I would also like to express my sincere thanks to Dr. Katherine Elliott and Kristine Hanstrom who helped make subjects available. 3 Although it is impossible to name each one, I would like to thank my classmates and friends who have encouraged me and have been able to empathize. Their presence has been particularly important to me and has helped me to enjoy these otherwise difficult years. Finally, I would like to express my heartfelt appreciation for my parents, Frank and Jessie Clark. Not only have they provided me with constant support and encouragement throughout this, and other endeavors, but have instilled within me a sincere concern for persons who have special needs. It is to these special people, my parents, that I dedi- cate this dissertation. ii TABLE OF CONTENTS List of Tables ......................... List of Figures ........................ CHAPTER I: INTRODUCTION .................... Need for the Study .................... Purpose of the Study ................... Definitions ........................ Hypotheses ........................ CHAPTER II: REVIEW OF THE LITERATURE ............. Cognitive and Personality Characteristics ......... PIC Development and Description .............. Reliability Studies of the PIC .............. Studies Using and Examining PIC Validity ......... Summary of the Research .................. CHAPTER III: METHOD ...................... Subjects ......................... Instruments ........................ Procedures ........................ Hypotheses and Data Analyses ............... Limitations and Assumptions of the Study ......... CHAPTER IV: RESULTS ...................... Descriptive Statistics .................. Hypothesis Testing .................... Supplementary Analysis .................. CHAPTER V: DISCUSSION ..................... Differential Profiles ................... Cognitive Triad of the PIC ................ Construct Validation of the PIC .............. Diagnostic Validation ................... 116 117 124 128 137 CHAPTER VI: SUMMARY AND CONCLUSIONS .............. Findings ......................... Practical Implications .................. Recommendations for Future Research ............ Appendices A: List of PIC Scale Abbreviations ............ B: Summary of Studies Using and Examining the Validity of the PIC ............... C: Personality Inventory for Children (PIC) ....... D: Teacher Rating and School Information ......... E: Consent to Participate ................ F: Letter to Parents ................... References ........................... iv 142 143 150 152 155 156 159 172 176 177 178 3.1 3.2 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 LIST OF TABLES Age and Sex Distribution ................ WISC-R IQ Scores .................... T Score Means and Standard Deviations Across PIC Profile Scales ............. Effects of Diagnostic Group on Profile Scale Elevations .............. Effects of Diagnosis, Sex, and Age on Profile Scale Elevations ............ Contribution of Variables in the Discriminant Function Analyses ........... Relative Ability of the Discriminant Function to Differentiate Groups .......... Summary Table of Correct Classifications by Discriminant Function Analyses ......... Item Composition of Walker Problem Behavior Factors . . . . Correlations Among PIC Profile Scales and WISC-R . . . . Obtained Profile Scale Intercorrelations and Expected Intercorrelations Due to Item Overlap . . . . 4.lO Number of PIC Items which Overlap the Profile Scale ................. 4.ll Significant Correlations Between PIC Profile Scales and TRSI .............. 55 56 68 76 81 86 87 91 96 99 LIST OF FIGURES 4.l PIC Profiles for L0, ED, EMR Subjects ........... 74 4.2 Profiles for Planned Comparisons: LD and ED ....... 77 4.3 Profiles for Planned Comparisons: EMR and LD/ED ..... 80 vi CHAPTER I INTRODUCTION Need for the Study The identification and classification of children with special needs has been a persistent and major concern of our educational sys- tem, particularly since the passage of Public Law 94-l42, the Educa- tion for All Handicapped Children Act of 1975. Through this legisla- tion, the federal government has appropriated the funds for educating the handicapped child who has been formally identified and classified according to what the law defines as handicapped. Public Law 94-l42 states that handicapped children be defined as . mentally retarded, hard of hearing, deaf, speech im- paired, visually handicapped, seriously emotionally dis- turbed, orthopedically impaired or otherwise health im- paired, or children with specific learning disabilities, who by reason thereof require special education and re- lated services (Section 602, paragraph 1). An accurate and early assessment of a child who is experiencing academically-related difficulties is important for the implementation of appropriate educational programs. As Bergan and Tombari (l976) have pointed out, the psychologist's success in proposing solutions to chil- dren's problems and having these implemented is largely dependent upon the initial stage of problem identification. Failure to correctly identify and apply appropriate intervention strategies soon after the child's difficulties become apparent may compound their problems. Re- searchers have shown that difficulties in learning can be causally related to both cognitive and emotional factors (Heinicke, 1972) and that delaying intervention may serve to aggravate any interaction be- tween these factors (Connolly, 1971). Although there are a number of psychometrically-sound measures for evaluating cognitive abilities, personality assessment generally in- volves the use of less reliable and valid projective instruments (Ana- stasi, 1976). Researchers have indicated that the Personality Inven- tory for Children (PIC) provides a means by which cognitive and academically-related abilities, as well as personality characteristics, may be assessed (Wirt, Lachar, Klinedinst, & Seat, 1977). The major purpose of this study is to determine whether the PIC is a valid and diagnostically useful instrument for a select group of special educa- tion children and adolescents (emotionally disturbed, learning dis- abled, and educable mentally retarded). At present, decisions regarding the diagnosis and treatment of handicapped children are based primarily on the judgments of school and clinical psychologists. However, results of studies conducted over the past several years suggest that psychologists find it extremely difficult to reliably and meaningfully classify children into cate- gories of emotionally disturbed, learning disabled, and educable men- tally retarded (Huelsman, 1970; Bryan & Bryan, 1976; Hallahan & Kauff- man, 1977: Downey, 1979). Goldberg and Werts (1966) pointed out that little support has been provided for the validity of clinicians' judgments on an absolute basis and relatively no validity for these judgments over actuarial methods. Contrary to commonly held beliefs, studies have shown that diagnostic validity does not necessarily increase as a function of either the number of tests administered (Sines, 1959) or the psychologist's years of professional experience and level of training (McDermott, 1980). Most of the studies showed that in the diagnostic process, inferences were crystalized early and were resistent to change despite the ac- quisition of further, and perhaps contradictory, information. Studies that examine the reliability of judgments made by clinical psychologists (Oskamp, 1965: Schinka & Sines, 1974) and school psy- chologists (Peterson & Hart, 1978; Elders, 1977; McDermott, 1980) have shown littleinterclinician agreement of diagnostic labels. While some studies suggest that diagnostic reliability or accuracy is case- specific; therefore, psychologists' judgments may be highly reliable on certain types of cases and quite unreliable on others (DeWitte, 1976), most researchers agree that psychologists tend to disagree when assigning diagnostic labels (Frame, 1979). The unreliability of judg- ments is often attributed to the inconsistent application of criteria in rendering diagnoses; however, others have pointed to the lack of standardized, operational definitions that will reliably and meaning- fully differentiate children in special education categories (Hammill, 1976; Bernard, 1978). Definitions are considered vital to insure validity of continuing research aimed at identifying and understanding educational handicaps; however, as Cruickshank (1972) indicated, there is an overabundance of terminology with no specific behavioralcmiteria. The legal definitions of emotionally disturbed, learning disabled, and educable mentally retarded (Education for All Handicapped Children Act, Public Law 94-142, 1975) are imprecise and unclear. The diffi- culties of classifications are further compounded by the fact that the law specifies the number of handicapped children who can receive special education services. P.L. 94-142 states that no more than 12% of the school-aged population, children ages 5 to 17, will be served. Lloyd, Sabatino, Miller, and Miller (1977) found that definitions which seek to conform to federal legislation, thereby controlling the inci- dence of handicapping conditions, have often resulted in the failure to provide appropriate educational programs for children with special needs. Studies of educationally handicapped children designated learning disabled, emotionally disturbed, and educable mentally retarded have concentrated on assessing their intellectual functioning, academic achievement, and personality, in an attempt to discover specific pat- terns of intellectual ability, achievement, or behavior that would dif- ferentiate these children and aid in the establishment of uniform . operational definitions for each group. It should be noted that some researchers (Quay & Werry, 1979) have argued that inadequate relia- bility and validity of instruments used to assess childhood psycho- pathology, as well as the inability of professionals with different theoretical perspectives and professional training to agree on criteria for childhood disorders, mitigate against the establishment of such definitions for each special education category. However, a review of the research shows that perhaps the major difficulty in differentiating children in the three groups lies in their similarity in behavior and personality characteristics. Hallahan and Kauffman (1976) found that children classified emo- tionally disturbed, learning disabled, and educable mentally retarded shared more similarities than differences. According to the researchers, children in all three groups show developmental delays, social-emotional maladjustment and academic difficulties: however, the degree to which these problem behaviors are demonstrated varies with the group. No patterns of behavior characterized any of the groups, or differentiated them from each other. In addition, although there has been little re- search on personality factors of educable mentally retarded and learn- ing disabled children, it has been found that certain cognitive de- ficits of the educable mentally retarded child (Zigler, 1975) and learning problems of the learning disabled child (Torgensen, 1977) can be attributed to poor psychological adjustment. In the area of intellectual functioning, a number of researchers have shown that learning disabled (Kirk & Elkins, 1975) and emotionally disturbed (Kauffman, 1977) children have measured IQs below the popula- tion mean; although, by definition, children classified as educable mentally retarded have lower 105 than learning disabled and emotionally , disturbed children. Underachievement has been typically used to iden- tify learning disabled children; however, educable mentally retarded (Dunn, 1973) and emotionally disturbed (Kauffman, 1977) have also been shown to achieve below their expected levels. The issue of diagnostic diffusion and the homogeneity of charac- teristics has been addressed by a number of researchers and considered by some to be a significant enough problem to preclude the use of such a classification system (Bryan & Bryan, 1976). Special educators, how- ever, continue to conceive of each category as distinct and separate (Forness, 1976), having labels which imply certain behavioral regu- larities which are important for educational interventions. Ideally, the special education diagnosis should specify an appropriate treatment and, theoretically, suggest preventive measures; however, this has not been shown. Clarizio and McCoy (1976) pointed out that despite the intended advantage of a diagnostic system for assisting in decisions regarding interventions, studies have failed to provide evidence to show that present classification schemes enable psychologists to link specific treatments with specific diagnoses. Furthermore, Draguns and Phillips (1971) have suggested that the diagnostician's treatment options are neither numerous enough nor are they specific enough to even warrant fine discrimination among diagnostic categories. Nonetheless, in order to be in compliance with the federal legislation and receive monetary support for special education programs, psychologists working for public school systems must classify a child according to the pre- scribed categories such as emotionally disturbed, learning disabled, and educable mentally retarded. A number of attempts have been made for behavioral description of exceptional children for the purpose of matching educational treatments with specific deficits; however, most of these have been unsuccessful (Iscoe, 1962; Quay, 1968). As Myers and Hammil (1976) have shown, school systems operate according to those categorical definitions which measure easily ascertained and labeled difficulties and ignore other pertinent components of the disorder. Despite the absence of a coherent conceptual framework for de- scribing and differentiating children who share certain intellectual and behavioral problems, it is important to determine whether these children differ according to certain patterns of cognitive, intellec- tual, and emotional/behavioral factors. An assessment of these variables would seem pertinent to planning educational experiences for children who have been referred for academically-related problems. Purpose of the Study Professionals who are concerned with the psychological assessment of children have expressed a persistent need for a comprehensive, well researched instrument which has relevant applications to treatment. The major purpose of this study is to determine whether the Person- ality Inventory for Children (PIC) is a valid and relevant instrument for the psychoeducational evaluation of children. One aspect of this study is to ascertain whether the PIC profile scales can discriminate among emotionally disturbed, learning disabled, and educable mentally retarded children according to certain intel- lectual, cognitive, and personality factors. Although a number of studies have investigated cognitive and emotional variables in attempt- ing to differentially diagnose these mildly handicapped children (Gajar, 1977; Downey, 1979), few have objectively and empirically ex- amined these variables. Studies which to date have examined the PIC have shown substantial support for the diagnostic potential and basic interpretive intent of the scales (Culbert & Gdowski, 1982; Lachar, Butkus, & Hryhorczuk, 1978). Culbert and Gdowski (1982) studied the clinical usefulness of the PIC to differentiate reading disabled (a subgroup of learning disabled) from a matched control group of children. The researchers found that the profile scales which measured cognitive and academically-related abilities significantly discriminated among the groups and were sensi- tive to variations in cognitive functioning. Since the study involved a small and restricted population of children, the present study will seek to determine whether the PIC scales are sensitive to cognitive variations, as measured by the Wechsler Intelligence Scale for Children --Revised (WISC-R), of a larger and more diverse sample of children. Recent studies have provided support for construct validity of the PIC profile scales (Lachar, Butkus, & Hryhorczuk, 1978: Lachar & Gdowski, 1979a); however, these studies were conducted on a psychia- tric population. According to Public Law 94-142, tests must be vali- dated for the specific purpose for which they are used. Therefore, the present study will attempt to validate the PIC on a special educa- tion population. This will be accomplished by assessing the extent to which the PIC profile scales relate to a standardized intelligence test and a teacher-completed problem checklist. Researchers have found that teacher ratings correlate poorly with those of psychologists (Achenbach, 1978) and parents (Becker, 1960) and are not sufficiently valid as primary indices in assessing maladjust- ment (Rutter, 1967). However, teachers' observations are based on a somewhat different behavioral domain than that of parents or psycholo- gists: therefore, it is reasonable to expect low correlations between the two groups of informants. Teacher perceptions have been considered valuable in the differential diagnosis of learning disabled, emotional- ly disturbed, and normal groups of children (McCarthy & Paraskevopou- 105, 1970). As teachers are generally the only informants regarding a child's behavior in the classroom, their assessments of the children have been selected as a criterion for examining the validity of the PIC profile scales. In addition, the WISC-R was selected as a criterion to judge the construct validity of the cognitive scales of the PIC since it is a widely accepted and commonly used measure of intellectual abilities. The PIC manual (Wirt, Lachar, Klinedinst, 8: Seat, 1977) has docu- mented the reliability of the PIC in normal and psychiatric popula- tions. The present study will also determine reliability estimates of the PIC for special education children. The PIC, unlike most other personality inventories, is completed by a secondary respondent, generally the child's mother. Since chil- dren are generally not articulate or psychologically aware enough to furnish reliable self-reports of their own behavior, most instruments used to assess child personality generally rely on observations by adults. Studies using parents as informants regarding the behavior of their children have repeatedly shown that interparent agreement and parent agreement with teachers and clinicians exceeds both interclini- cian agreement and agreement between clinicians and teachers. Further- more, the information provided by parents has been shown to be more comprehensive than that from teachers and objective home and school observers (Achenbach, 1978). Substantial research has also supported the contention that parent descriptions of emotional (Maes, 1966; Novick, Rosenfeld, Bloch, & Dawson, 1966) and academically-related difficulties (Strag, 1972: Iano, 1970) are quite reliable and useful in both the identification and treatment processes. In fact, McCoy (1976) found that clinical deci- sions regarding the psychological diagnosis and disposition of a child in clinic settings depend almost entirely on parental report. 10 Recent research with the Personality Inventory for Children sug- gests that "parents can relay at a molecular level, via individual inventory responses, accurate and useful diagnostic information about school performance and cognitive abilities" (Lachar & Gdowski, 1979b, p. 51). Although researchers have contended that information supplied by parents is subject to bias and distortion (Robbins, 1963; Yarrow, 1963), studies using the PIC have shown little support for this conten- tion. For instance, Lachar and Sharp (1979) found that potential bias and distortion in reports that could result from informant psychopa- thology did not have a consistent effect on the PIC results. According to their study, systematic bias affected only a limited portion of the PIC items, such as those which identify somatic concerns. Definitions LearningDisabled "Learning disabled" is defined as: l. Disorder in one or more of the basic psychological pro- cesses involved in understanding or in using spoken or written language, which disorder may manifest itself in imperfect ability to listen, think, speak, read, write, spell, or do mathematical calculation. 2. Manifestation of symptoms characterized by diagnostic labels such as perceptual handicap, brain injury, mini- mal brain dysfunction, dyslexia, or aphasia. 3. Development at less than the expected rate of age group in the cognitive, affective, or psychomotor domains. 4. Inability to function in regular education without sup- portive special education services. 5. Unsatisfactory performance not found to be based on social, economic, or cultural background. 11 Emotionally Disturbed "Emotionally disturbed" means a person having one or more of the following behavioral characteristics: 1. Disruptive to the learning process of other students or him/herself in the regular classroom over an extended period of time. Extreme withdrawal from social interaction in the school environment over an extended period of time. Manifestation of symptoms characterized by diagnostic labels such as psychosis, schizophrenia, and autism. Disruptive behavior which has resulted in placement in a juvenile detention facility. Educable Mentally Retarded "Educable mentally retarded" is defined as: 1. Development at a rate approximately two to three stan- dard deviations below the mean as determined by intel- lectual assessment. Scores approximately within the lowest six percentiles on a standardized test in reading and arithmetic. Lack of development primarily in the cognitive domain. Unsatisfactory academic performance not found to be based on social, economic, and cultural background (Michigan Special Education Code as amended January 14, 1977). Educationally Handicapped "Educationally handicapped" is defined as all sabjects labeled learning disabled, emotionally disturbed, and educable mentally re- tarded. Special Education "Special education" refers to all programs and services under the classification of special education according to Public Law 94-142 (Education for All Handicapped Children Act, 1975). 12 Hypotheses Given that the major purpose of this study is to determine the diagnostic potential of the PIC in a special education population, the following hypotheses were tested. Hypothesis 1 The Personality Inventory for Children (PIC) will discriminate among the three groups of special education children, learning dis-~ abled, emotionally disturbed, and educable mentally retarded, accord- ing to their profile scale patterns. Hypothesis 2 A significant discriminant function using multiple variables (PIC profile scales, WISC-R 10, Walker factors, sex, age) will predict mem- bership in the learning disabled, emotionally disturbed, and educable mentally retarded groups. Hypothesis 3 There will be significant negative correlations among the cogni- tive scales of the PIC (Achievement, Intellectual Screening, Develop- ment) and the Wechsler Intelligence Scale for Children--Revised (WISC-R) subtests, and 105. CHAPTER II REVIEW OF THE LITERATURE Numerous studies have appeared in the literature designed to iden- tify cognitive and personality characteristics of children diagnosed as learning disabled, emotionally disturbed, and educable mentally retarded. However, the interpretation of the findings is often ob- scured by the lack of consistency in the way these categories are de- fined and the methodological shortcomings which seem to plague many of the studies. Consequently, few studies were selected for review. The studies which were chosen were selected because (a) personality and/or cognitive characteristics of emotionally disturbed, learning disabled, and educable mentally retarded children and adolescents were studied; (b) demographic variables were described and, in general, controlled for; and (c) the methodology was relatively sound. The literature review is divided into five sections. Studies that have attempted to identify cognitive and personality characteristics of learning disabled, emotionally disturbed, and educable mentally retarded children are presented first. The section is subdivided into studies examining (a) the intellectual characteristics of children with learn- ing and behavioral problems, and (b) the differential diagnosis of educationally handicapped children. Second, the Personality Inventory for Children (PIC) scale development and description are presented. Then, studies which have investigated the reliability of the PIC are described. Finally, in light of the questions posed in this study, 13 14 studies that have used the PIC and have provided support for its va- lidity are reviewed. Cognitive and Personality Characteristics Intellectual Character- istics Studies Stevenson (1980) studied WISC-R subtest patterns of children re- ferred for school-related difficulties to the Child Service Demonstra- tion Center at Northeastern Illinois University. The 55 subjects were between the ages of 5 years, 10 months and 17 years, 3 months. The subjects were predominantly white middle class children with 105 rang- ing from 78 to 116 (Full Scale IQ mean = 96.89, Verbal IQ mean = 97.41, and Performance 10 mean a 97.38). The study sought to identify unique profiles which would charac- terize the children. WISC-R scores were factor analyzed in order to examine those subtests which have been hypothesized to reflect func- tional variations in learning. The subjects scored lowest on Digit Span, Coding, Arithmetic, and Information. No significant differences according to grade, age, and 10 were found. The author concluded that attentional difficulty is the key factor in characterizing learning problems of children. It is difficult to generalize from this study to other popula- tions, as the sample was not clearly defined as learning disabled and was being seen at an agency outside the school. However, according to the author, these findings provide further support for the attentional and concentrational deficits of children experiencing school-related problems. These findings support those of earlier studies which 15 examined the intellectual characteristics of learning disabled chil- dren (Vance & Gaynor, 1976; Smith, Coleman, Dobecki, & Davis, 1977). Vance and Gaynor (1976) studied the WISC-R subtest patterns of children identified as learning disabled. The sample included 58 chil- dren (42 males, 16 females) with an age range of 6 years to 15 years, 10 months. Data analysis showed that the subjects scored lowest on WISC-R subtests which tested attentional and concentrational processes. The learning disabled group was characterized by low scores on Arithmetic 7'= 7.5) and Coding (7'= 7.7). Relatively high subtest scores were ob- tained on Comprehension (7'= 9.7), Picture Completion (7'= 9.8), and Object Assembly (7'= 10.4). No significant differences across age and sex were feund. This study showed general limitations including failure to re- port both the Digit Span score and control for the effect of 10 on the WISC-R subtest scores, the use of a small and restricted (rural) sam- ple, and the absence of a control group. The study's findings, there- fore, 1ack generalizability and diagnostic utility. Similar results were obtained by Smith, Coleman, Dobecki, and Davis (1977) who used a larger and more representative sample and controlled for demographic effects. Intellectual characteristics of 208 children (6-12 years) enrolled in classrooms for the learning disabled were ex- amined. The children were divided into high and low IQ groups, with the mean IQ of the high group at 93.3 and the mean IQ of the low group at 76.3. Smith et a1. (1977) found that both groups of learning disabled children scored significantly lower (p<<.Ol) on the Arithmetic, 16 Information, and Coding subtests. The total sample means were 7.2, 6.6, and 7.3, respectively. The highest scaled scores were the same for both learning disabled groups, with total sample means of 8.2 on Comprehension, 9.7 on Picture Completion, and 10.2 on Object Assembly. Across both groups, the Performance IQ was significantly higher than the Verbal IQ. The results of the study suggest that attention- concentration problems are characteristic of children classified as learning disabled. Furthermore, the authors suggest that profile analysis may be more useful than the Verbal-Performance IQ discrepancy in describing learning disabled children. The problem in interpreting these results is that 37% of the sample demonstrated significantly im- paired intellectual functioning and may be more representative of an educable mentally retarded or borderline impaired population rather than learning disabled. Dean (1977) studied the WISC-R patterns of special education chil- dren classified as emotionally disturbed. The subjects included 41 Caucasian males who were between 13 and 15 years of age. The subjects were from lower-middle and middle class backgrounds and 105 between 80 and 105. All the children had been referred for a variety of behavior problems including aggressiveness, anxiety, truancy, and temper tan- trums. The sample's WISC-R scores were compared to those of the normative sample (Wechsler, 1974). The results showed that the emotionally dis- turbed group had more subtest scatter; however, the author cautioned that scatter analysis cannot be considered conclusive without further verification. The emotionally disturbed group's performance subtest means exceeded all verbal tests, with the exception of Similarities. 17 This finding was consistent with previous studies which showed de- pressed verbal functioning of emotionally disturbed children (Saccuzzo & Lewandowski, 1976). Dawdy (1979) explored the relationship between WISC-R Verbal- Performance IQ discrepancies and behavior disorders in children. Since a number of studies had inadequately controlled for or assessed the effects of sex, age, and Full Scale IQ when examining intellectual characteristics of children with behavioral problems, Dawdy studied the relationships among these variables by examining the records of 310 children seen at the University of Iowa Psychiatric Hospital. Criteria for subjects selected included a complete psychiatric evaluation and the administration of a WISC or WISC-R within one year of the psychiatric exam. Upon review of the subjects' records, each child was rated on a symptom checklist comprised of 67 items grouped according to 11 psy- chiatric diagnoses. Two of the 11 categories were found to be related to WISC-R sub- test patterns. Conduct disordered children had higher scores on per- formance subtests and the organic brain disordered group had higher scores on verbal subtests. No significant differences for age, sex, or Full Scale 10 were found. Whereas, the depressed verbal scores of the conduct disordered group is consistent with other studies of emotional- ly disturbed children, the performance of the organic brain disordered children was not. The inconsistency may, however, be a function of the group which may be atypical of children diagnosed emotionally dis- turbed and placed in public school programs. Since the sample was drawn from a psychiatric population and may not be representative of l8 emotionally disturbed children in schools, the generalizations from this study are limited. The preceding studies are important in their attempt both to de- scribe the intellectual characteristics of children who are experi- encing school-related difficulties and to identify specific strengths and weaknesses which are relevant for prescribing educational treat- ments. Two lines of research were generally followed. One examined the WISC-R discrepancy scores, and the other studied WISC-R subtest patterns. The analysis of subtest patterns was consistently found to be more useful in describing the handicapped groups, with children ex- periencing emotional and/or behavioral problems having more subtest scatter and children with learning difficulties scoring lowest on sub- tests measuring attentional and concentrational factors. Since the populations were not clearly defined and sample sizes rather small, the results have limited generalizability. Differential Diagnosis of Educationally Handicapped Keogh, Wetter, McGinity, and Donlon (1973) investigated the use of the WISC subtest pattern for differentiating children with intellectual limitations and learning problems coupled with serious behavior diffi- culties and hyperactivity. The sample included 26 males in public school classes for educable mentally retarded students, 24 private school children with serious learning and behavior problems, and 26 males referred to a pediatric learning disability clinic for evaluating hyperactivity and learning problems. The groups did not vary signifi- cantly in age (7'= 10 years, 6 months); however, the educable mentally retarded children had a significantly (p<:.01) lower 10. The average 19 IQ of the educable mentally retarded group was 74.3, whereas the average for the other two groups were 103 and 104. The data analyses showed that the educable mentally retarded chil- dren had the lowest mean scores on subtests measuring what the author referred to as verbal comprehension and had the highest means on sub- tests which were labeled analytic. Both the children with serious learning and behavior problems and those identified as hyperactive and experiencing learning difficulties had average scores on subtests as- sessing verbal comprehension and analysis and the lowest scores on subtests measuring attention-concentration factors. Whereas withina group differences were significant at the p < .01 level, the between- group differences were not significant. The researchers concluded that the study's results provided evi- dence which links attentional deficits and learning problems. The study gave support to the contention that children with learning prob- lems demonstrate a unique WISC subtest pattern, with adequate scores on verbal and analytic subtests and lowest scores on subtests measuring attention and concentration. However, there are a number of limita- tions such as the failure to administer certain WISC subtests, drawing the sample from a private facility and not having clearly defined emotionally disturbed and learning disabled populations for comparative purposes. Keogh and Hall (1974) conducted a similar study in a public school special education program. The subjects included 157 learning disabled and hyperactive children and 83 educable mentally retarded children. The learning disabled and hyperactive group consisted of 125 males with an average 10 of 96 and 32 females with an average 10 of 91. The 20 educable mentally retarded group had 47 males with an average IQ of 71 and 36 females with an average IQ of 70. Using a series of ANOVAs, the data were analyzed to look for dif- ferences between the groups and sexes and within the groups' patterns of WISC performance. The Verbal-Comprehension score for the educable mentally retarded females was significantly lower (p< .025) than the Attention-Concentration or Analytic factor scores. The educable men- tally retarded males had analytic scores which were significantly high- er (p‘<.Ol) than Attention-Concentration or Verbal Comprehension. The learning disabled and hyperactive males scored significantly lower (p41 .01) on the Attention-Concentration factor.whereas the females had sim- ilar scores on all three factors. Although no significant differences were evidenced among the factors, the females‘ Attention-Concentration factor scores resembled that of the males. The researchers concluded that evi- dence was shown for the role which attentional deficits play in learning difficulties and that females diagnosed learning disabled and hyperactive demonstrate broader and more pervasive problems which include attention de- ficit and difficulty comprehending verbal stimuli and organizing spatially. The study was important for examining cognitive patterns of spe- cial education school children with school-related problems, however, like the Keogh et al. (1973) study, a clearly designated learning dis- abled population was absent. The results are restricted further by its use of an urban sample of children. Hughes (1977) investigated the usefulness of the WISC-R factor loadings to differentially diagnose special education students. The subjects consisted of 300 children who had been referred and/or placed in special programs for emotionally disturbed, educable mentally 21 retarded, language disability and learning disabled, minimal brain dysfunction, and a group considered ineligible for placement. Each subject's WISC-R was factor-analyzed, yielding factors labeled "perceptual organization," "verbal comprehension," and "free- dom from distractibility." The subjects in all the groups had patterns of relatively high perceptual organization and low verbal com- prehension and freedom from distractibility. Since no significant dif- ferences were found across the various groups, Hughes concluded that the three factors were of no benefit in differential diagnosis. The study allowed for the examination of disconfirming evidence by using the ineligible group in the design; however, the results are dif- ficult to interpret in light of the combining of language disabled and learning disabled children and separating the minimal brain dysfunction children into a discrete group. The failure to differentially classify the subjects may, in part, be due to the manner in which the children were grouped. Webster (1977) reviewed the case files of 1,596 children who had been referred by classroom teachers for psychoeducational evaluation because of learning and/or social problems. The subjects ranged in age from 6-17 years with an average 10 of 87.3. Subjects were classi- fied as learning disabled, emotionally disturbed, educable mentally re- tarded, physically handicapped, and ungrouped special education stu- dents with no clear diagnostic label. All subjects had been identified by a diagnostic team and placed in appropriate educational programs. From case records the researcher drew information pertaining to cultural background, socioeconomic status, geographic locale, history _ of parents experiencing learning problems, neurological or medical 22 difficulties and the use of a second language in the home. Diagnostic test results including the WISC-R and Wide Range Achievement Test (WRAT) were also taken from the files. A series of discriminant func- tion analyses was performed in an attempt to delineate the diagnostic usefulness of norm-referenced, objective test data and personal infor- mation in discriminating these special education children. The results showed that the analyses consistently failed to di- scriminate learning disabled from the other children. Even the most powerful discriminant function analysis indicated that 84% of the mis- classifications (27%) were assigned to the learning disabled group. However, the three most significant variables which contributed to the group differentiation were social responses, as measured by the Picture Arrangement and Comprehension subtests of the WISC-R; estimate of in- tellectual ability, indicated by the WISC-R Full Scale 10: and the ability to analyze phonemically and synthesizing words, as assessed by the WRAT. The author concluded that a child who has an average IQ but func- tions below grade level is generally labeled learning disabled, whereas a child who falls in the borderline defective range is labeled edu- cable mentally retarded. When a child with average intelligence func- tions close to grade level but still has problems learning, emotional disturbance is assumed. Therefore, Webster proposed a more appropriate diagnostic model that accounts for a child's manner of processing in- formation relative to his/her content base and ability to perform asso- ciated behaviors. The study has a number of limitations. Although the physically impaired and unlabeled special education children were initially 23 included in the design, they were excluded in the later analyses in order to minimize external sources of variance. In addition, only sub- jects with an IQ of at least 75 were included in the analysis, theore- tically eliminating the educable mentally retarded group. For these reasons, the study has limited generalizability. Bernard (1978) attempted to develop a formula that would maxi- mally differentiate children classified as emotionally disturbed (n = 83), educable mentally retarded (n = 163), learning disabled (n = 320), otherwise-impaired (n = 70), and non-impaired (n = 333); and, second, to determine the degree of overlap in their characteris- tics. Data were collected from the records of 1,129 children (866 males, 263 females) who had been referred for psychoeducational as- sessment and assigned to special education categories by Educational Planning and Placement Comittees (EPPC) in 45 Michigan school dis- tricts. The variables in the study included WISC-R test scores, WRAT, Peabody Individual Achievement Test (PIAT), grade placement, sex, socioeconomic status, EPPC recommendations, and the judgments of school psychologists and learning disability consultants. The data analyses sug- gested that no single characteristic or cluster of characteristics differentiate learning disabled children from emotionally disturbed, educable mentally retarded, otherwise-impaired, or non-impaired. The study did show, however, that by increasing the number of predictor variables, the degree of discrimination was increased. When Verbal- Performance IQ discrepancy alone was considered, 37.8% of the learning disabled group was correctly identified, whereas when Full Scale IQ, sex, age, and subtest patterns were combined, the rate of correct 24 identification increased to 67.5%. The researcher concluded that prac- titioners need to guard against simplistic description of these mildly handicapping conditions and consider a variety of characteristics. Becker (1978) examined the learning characteristics of a com- bined group of learning disabled and emotionally disturbed children with a group of educable mentally retarded children. The subjects in- cluded 60 children who ranged in age from 9 to 13. They were randomly selected from special education classes in a large metropolitan school district in Southern California. There were 40 children in the learn- ing disabled/emotionally disturbed group and 20 in the educable men- tally retarded group. There was an even sex distribution in the latter group, whereas the learning disabled/emotionally disturbed group con- sisted primarily of males. There were significant differences in mental age (LO/ED = 107, EMR = 91). IO (LO/ED = 83, EMR = 59).WISC-R performance (LO/ED 15 points higher than EMR), and Raven Progressive Matrices (LO/ED = 17, EMR = 13). The researcher concluded that substantial differences existed between the two groups; therefore, the assumption of character- istic overlap was questioned. However, the only characteristic which was examined was IQ. It is quite reasonable to assume that educable mentally retarded children, by definition, will be lower on 10 mea- sures. Nicholls (1979) studied the diagnostic validity and descriptive utility of a factor analytic approach to WISC-R profile interpretation among children with school-related problems. The study had a sample of 145 males and 61 females, all of whom were in the elementary special education programs. The children ranged in age from 6 to 14 years and 25 had been identified as learning disabled (n = 144), emotionally dis- turbed (n = 30), or mentally retarded (n = 32). Data included WISC-R scores, achievement test scores on the WRAT or PIAT, and information pertaining to sex, age, and socioeconomic status. These data were obtained from school records or direct test- ing. Univariate and multivariate analyses of variance revealed sig- nificant differences among the groups' factor score profiles. These differences were not, however, attributed to any specific WISC-R fac- tor. The learning disabled children were characterized by low freedom from distractibility scores, the mentally retarded children showed high perceptual organization skills, and emotionally disturbed children did not have a characteristic factor score pattern. Demographic variable effects were noted only for males in the various groups. Discriminant analysis procedures did not support the predictive validity of the factor score approach to differential diagnosis. Whereas the factor score approach to interpreting WISC-R profiles did not show diagnostic utility when used alone, it was seen as valuable in contributing information for the diagnostic process. The study showed strength in its use of a relatively large and representative sample of mildly impaired children, as well as control- ling for variables such as sex, age, and SES. The study did not, how- ever, indicate the effect of 10 on the results. Waits (1979) conducted a study of the cognitive functioning of learning disabled and behavior disordered children. The sample con- sisted of 325 children who ranged in age from 6 to 16 years. There were 186 subjects in the learning disabled group and 139 in the group having behavior disorders. 26 Variables of race, sex, and WISC-R data were used to determine their diagnostic utility in differentiating mildly handicapped chil- dren. The results revealed that the learning disordered group scored significantly higher than the behavior disordered group on WISC-R sub- tests and had a larger Verbal-Performance IQ discrepancy. When race, sex, and certain WISC-R subtests (Comprehension, Picture Arrangement, Arithmetic) were used as classifying variables, the two groups were differentiated more clearly. Although the study included a large sample of special education children, the failure to define more clearly the behavior disordered group limits the study's usefulness in differential diagnosis of these children. ‘ Waldman (1979) studied the intellectual characteristics of chil- dren classified as learning disabled, emotionally disturbed, and nor- mal. The subjects included 192 children. The researcher obtained WISC-R test information from the case records to determine the effi- cacy of using subtest patterns and discrepancy scores to differentially diagnose these mildly impaired students and normal students. Results indicated that the learning disordered and emotionally disturbed males scored significantly higher on Performance IQ and that both learning disordered males and females scored lower than the normal males on certain WISC-R subtests (Information, Arithmetic, Digit Span, Coding). The researcher concluded that WISC-R data alone could not be used to discriminate among learning disabled, emotionally disturbed, and normal children. Downey (1979) studied the effectiveness of differential diagnosis between emotionally disturbed and learning disabled children. From a 27 list of 278 children who were waiting to be placed in classrooms for the mildly handicapped, 80 children were selected for the study. There were 40 children in each of the special education groups and 40 chil- dren in a normal control group. The subjects were in grades 2 to 5 and had 105 between 83 and 110. 3 The three groups were compared on cognitive and affective charac- teristics. The cognitive measures included the Stanford-Binet and the Stanford Achievement Tests. The affective measures consisted of the Devereux Elementary School Behavior Rating Scale for adaptive behavior and the Quay and Peterson Behavior Problem Checklist for social skills. The data analyses revealed that no significant differences existed among the special education groups and the normal group on the basis of specific intellectual skills. The normal group did, however, score higher than the emotionally disturbed and learning disabled groups in terms of Full Scale IQ. The emotionally disturbed group scored higher than the learning disabled group on the Stanford reading and total scores, whereas the normal subjects scored higher on reading, math, and total score. There were no sigificant differences between the learning disabled and emotionally disturbed groups on the Devereux, which measured adap- tive behavior; however, the learning disabled children did show a more positive attitude toward the teacher and school. The normal group scored higher than both special education groups on all adaptive mea- sures except anxiety. The learning disabled group scored higher than the emotionally disturbed group on the Quay and Peterson Problem Check- list total score, conduct problems, and socialized delinquency. The normal subjects scored higher on all the social skills measures. 28 Segal (1979) conducted a similar study of learning disabled and emotionally disturbed children. The researcher likewise used a variety of measures to determine whether these mildly impaired groups could be differentially diagnosed. The sample consisted of 89 white, middle class males, ranging in age from 8 to 11 years. The IQ range for the total sample was between 85 and 115, as measured by the WISC-R. There were 49 children in the learning disabled group and 40 in the emotion- ally disturbed group. Measures included the WISC-R, the Coopersmith Self Esteem Inven- tory, the Child Behavior Scale, and the Keystone School Vision Screen- ing Test. Inner control was measured by the distractibility factor of the WISC-R, intersensory integration was assessed by the discrepancy between digits forward and backward on the WISC-R Digit Span subtest, abstract reasoning was measured by the Similarities and Block Design subtests of the WISC-R, practical reasoning was assessed by the WISC-R Comprehension and Picture Arrangement subtests, emotional' adjustment was assessed by the interpersonal maturity factor of the Child Behavior Scale, and vision was measured by the Keystone. Discriminant analyses were used and revealed that the two groups could not be predicted by the variables included in the study. The age and IQ effects were not reported; therefore, it is unknown the extent to which these variables affected the results. Gajar (1979) analyzed cognitive, affective, and demographic char- acteristics of children identified as learning disabled, emotionally disturbed, and educable mentally retarded. The purpose of the study was to determine the relevance of certain variables in defining and identifying these three groups of exceptional children. The subjects 29 consisted of 378 children who had been diagnosed and placed in special education programs. The cognitive measures included the WISC-R and standardized achievement tests. The affective measure included ratings on three factors: conduct disorder, personality problems, and immaturity-inadequacy. Demographic data included sex, race, and socio- economic status. Results showed that all three groups were lower than the popula- tion mean for IQ; were underachievers in reading, math, and spelling; had the highest mean scores on the immaturity-inadequacy factor; and had a disproportionately large number of black males and low SES chil- dren. The significant differences among the groups included educable mentally retarded's being lower than learning disabled or emotionally disturbed on IQ; learning disabled's being lower than emotionally dis- turbed on reading and having greater subtest scatter on the WISC-R than either emotionally disturbed or educable mentally retarded groups; and emotionally disturbed's being lower than educable mentally retarded on immaturity-inadequacy factor and lower than both learning disabled and educable mentally retarded on math scores, conduct disorders, and per- sonality problems. Review of these studies leaves little doubt as to the difficulty in discriminating among children with learning disabilities, intellec- tual impairment, and emotional/behavioral disturbances. The studies are, however, important in addressing the issue of characteristic over- lap and should be commended for their relatively large sample sizes and the consideration of demographic variables of IQ, sex, age, and socio- economic status. In general, the research did not support the pre- dictive validity of the cognitive or affective measures used to 3O differentially diagnose the children. The studies did, however, show that the rate of discrimination increased with the number of variables considered. A problem encountered in interpreting the results from the studies was the lack of consistency in defining the special education categories as well as a tendency to combine groups of children exhibit- ing difficulties in school. PIC Development and Description The Personality Inventory for Children (PIC) is an objective, mul- tidimensional personality inventory which seeks to provide comprehen- sive and clinically relevant descriptions of children and adolescent personalities (Wirt, Lachar, Klinedinst, & Seat, 1977). The PIC con- sists of 600 items and can be scored for 16 profile scales and 17 sup- plemental scales. A list of the 33 scales and their abbreviations is presented in Appendix A. The inventory is completed by a secondary respondent, generally the mother. The informant's perceptions of the child under study are intended to aid in the diagnosis and treatment of the child, as well as the early identification of developing patterns of problem behavior. Although the Minnesota Multiphasic Personality Inventory (MMPI) has been widely used for similar purposes as the PIC, the authors of the PIC were concerned that the child's level of motivation and/or cogni- tive abilities for reading and conceptual understanding may preclude a valid assessment using the MMPI. Wirt and Broen (1958), therefore, chose to develop an instrument which used parents as respondents. Personality dimensions and inventory items were selected on the basis of one or two methodologies, empirical and rational (Wirt, 31 Lachar, Klinedinst, & Seat, 1977). The empirical scale development was based on the use of appropriate criterion groups and normal contrast groups. Items were submitted to criterion and normal groups, and those items that differentiated between the two were included. In addition, the Darlington and Bishop (1966) method of scale construction was used to obtain optimum scale validity. This was accomplished by making pos- sible the use of all items in the item pool and adding new items ac- cording to iteration validity comparisons. The empirically derived scales consist of two validity scales (F and Defensiveness), one screening scale (Adjustment), and five clinical scales (Achievement, Intellectual Screening, Delinquency, Psychosis, and Hyperactivity). The rational scales were constructed using content-oriented and internal consistency methods. Twelve experienced judges nominated items from the PIC item pool, with each judge's choosing items for three scales. The criteria for item inclusion were that the item had to be selected by three-fourths of the judges as measuring the content of the specific scale and the item had to be keyed in the same direc- tion by at least two-thirds or three-fourths of the judges. The ra- tionally derived scales consist of one validity (Lie) and seven clini- cal scales (Development, Somatic Concern, Depression, Family Relations, Withdrawal, Anxiety, and Social Skills). Wirt, Seat, and Broen (1977) have pointed out that the profile scales which include three validity, one screening, and 12 clinical scales have greater usefulness for clinical application. The supple— mental scales have been shown to be less psychometrically-sound and require further research. The supplemental scales consist of Adoles- cent Maladjustment (AGM), Aggression (AGN), Asocial Behavior (ASO), 32 Cerebral Dysfunction (CDY), Delinquency Prediction (DP), Ego Strength (ES), Excitement (EXC), Externalization (EXT), Internalization (INT), Infrequency (INF), Introversion-Extroversion (I-E), K (K), Learning Disabilities Prediction (LPD), Reality Distortion (RDS), Sex Role (SR), Social Desirability (SD), and Somatization (SM). For a description of the PIC supplemental scales, the PIC manual should be consulted (Wirt, Lachar, Klinedinst, & Seat, 1977). The profile scales are briefly described below. Validity.Scales The Lie (L) scale is a 15-item scale intended to identify an informant's tendency to deny commonly occurring childhood problems and ascribe the most virtuous behaviors to the child (Seat & Wirt, 1973). The F (F) scale is a 42-item scale developed to target de- viant response sets such as the exaggeration of symptoms or randomness in responding (Seat, 1971). The Defensiveness (DEF) scale is a 23-item scale written to determine a respondent's tendency to be defensive about the designated child's behavior. Interscale correlates suggest that the DEF scale is negatively related to the informant's expressing negative attributes, particularly those that are interpersonal (Myers, 1974). Screening.$cale The Adjustment (ADJ) scale is a 76-item scale constructed as a screening device to identify general adjustment problems and serve as an indicator of those children in need of further psychological evaluation (Seat, 1969). Clinical Scales The Achievement (ACH) scale consists of 31 items. The scale was designed to identify children who are significantly below age expectancy in their academic achievement, regardless of their potential to achieve at an age-appropriate level (Lachar, 1974). 33 The Intellectual Screening (IS) scale is a 35-item scale in- tended to identify a child with impaired intellectual func- tioning and in need of further evaluation (Froman, 1973). The Development (DEV) scale is a 25-item scale designed to identify weaknesses in intellectual and physical development which may be reflected in poor classroom performance (Kline- dinst, 1972,1975). The Somatic Concern (SOM) scale is a 40- item scale which identifies reoccurring concern with physical symptomatology (Klinedinst, 1972,1975). The Depression (0) scale is a 46-item scale which reflects childhood depression and measures its importance as a com- ponent of psychological disturbance (Froman, 1971). The Family Relations (FAM) scale consists of 35 items which measgre family effectiveness and cohesion (Klinedinst, 1972, 1975 . The Delinquency (DLQ) scale is a 47-item scale intended to measure tendencies toward delinquent behavior (Lachar, Abato, & Wirt, 1975). The Withdrawal (WDL) scale is a 25-item scale designed to identify children who are withdrawn from social interactions (Klinedinst, 1972, 1975). The Anxiety (ANX) scale is a 30-item scale which was de- veloped to measure symptoms of anxiety, including a low threshold for frustration, exaggeration of problems, irra- tional fears and worries, nightmares and behavioral as well as pgychological correlates of anxiety (Klinedinst, 1972, 1975 . The Psychosis (PSY) scale is a 40-item scale constructed to discriminate psychotic children from normal, behavior- ally disturbed, nonpsychotic, and retarded children (Lachar, 1971). The Hyperactivity (HPR) scale is a 36-item scale intended to identify children whose behaviors are frequently asso- ciated with the "hyperkinetic syndrome" (Hegeman, 1976). The Social Skills (SSK) scale consists of 30 items designed to reflect an effectiveness in social relations and the rea- sons)for a lack of successful interaction (Klinedinst, 1972, 1975 . lia ret mot ser bet The frc 1111 the 011‘ V3 ma 0C 25 34 Reliability Studies of the PIC Wirt, Lachar, Klinedinst, and Seat (1977) have documented the re- liability of the PIC within clinical and normal populations. Test- retest reliability was estimated in a clinical population by having mothers of 34 children being evaluated at Lafayette Clinic's outpatient service complete PICs on two separate occasions. The time interval between the test administrations was betwen 4 and 72 days (Y'= 15.2). The clinical sample included 22 males and 12 females ranging in age from 5.2 to 14.7 years (7'= 9.7). The correlations between the two ad- ministrations yielded an average reliability coefficient of .86 for the 16 profile scales and an average of .89 for the clinical scales only. Two reliability studies, one in Michigan and the other in Pennsyl- vania, were conducted to obtain estimates of scale reliability in nor- mal populations. The Michigan study involved 46 mothers who on two occasions completed the PIC on their children. The sample consisted of 25 males and 21 females between the ages of 4.4 and 16.11 (7'= 9.4). The testing interval was between 13 and 102 days (7'= 50.9). The aver- age test-retest reliability coefficient for the profile scales was .71. The Pennsylvania study consisted of a sample of 55 children, 34 males and 21 females. The age range was between 5 and 11 (7'= 7.9). There was a two-week interval between the two occasions in which the mothers completed the PIC on their children. The average reliability coefficient for the profile scales was .89. These studies suggest that the PIC profile scales demonstrate suf- ficient stability across time to allow for the inventory's use in in- dividual assessment. The lower reliability coefficient obtained in the 35 Michigan study of nonpatients may be attributed to the extended test- retest interval, just as the higher correlations found in the Pennsyl- vania study may reflect a shorter time interval. Another factor which may have contributed to the higher correlations between tests in the Pennsylvania study was the data collection procedure, which did not assure that the informants would not refer to the inventory which they completed first while responding to the inventory for a second time. Studies Using and Examining the validity of the PIC Wirt et a1. (1977) document in the PIC manual that validity has been shown for a number of scales (ADJ, ACH, IS, PSY, and HPR) in the initial studies of scale construction. Patterns of PIC responses were effective in normative-criterion group separation and have shown con- sistency in cross validation using standardized intelligence tests and teacher ratings. Since the publication of the manual, several studies have been conducted which provide further support for the validity of the PIC. Most of these studies sought to determine the clinical util- ity of the instrument and, while doing so, showed findings of PIC scale construct validity. The evidence for validity is presented in conjunc- tion with previous studies which used the PIC and are relevant to the questions posed in this study. The following studies are organized according to the evidence which they provide for construct validity and criterion-related (predictive) validity, as defined in the APA (1966) standards for educational and psychological tests: Construct validity is evaluated by investigating what psy- chological qualities a test measures; i.e., by determining the degree to which certain explanatory concepts or con- structs account for performance on the test. 36 Criterion-related (predictive) validation compares test scores, or predictions made from them, with an external variable (criterion) considered to provide a direct mea- sure of the characteristic or behavior in question (pp. 12-13). Construct Validity Lachar, Butkus, and Hryhorczuk (1978) investigated the diagnostic potential of the PIC in a child psychiatric setting by determining ex- ternal correlates of the profile scales. Mothers of 79 children (55 males, 24 females) who had received outpatient evaluation at Lafayette Clinic in Detroit completed PICs. The average age of the children was 9 years, 8 months. The sample consisted of children with varied symptomatology, with primary diagnoses including hyperkinetic reaction; unsocialized, aggressive reaction; specific learning disturbance: depressive neuro- sis; adjustment reaction; overanxious reaction; mental retardation/ organic brain syndrome; withdrawing reaction; seizure disorder; and those with no psychiatric illness. No psychotic children were included. Two clinicians each reviewed half of the sample's medical records and completed a 94-item problem behavior checklist for each subject based on his/her medical records. Interclinician agreement was de- termined initially to be greater than 85%. The clinicians made no reference to PIC data when completing the checklists. The problem checklist contained items pertaining to self-concept, affect, cognitive functioning, interpersonal relations, physical development and health, family relations, and parent description. The PICS were scored for the 16 profile scales, and correlations were computed for each scale and behavior checklist item. The data 37 analysis showed that each PIC scale obtained an average of 11.9 cor- relates. The following PIC correlates were found to be the most ro- bust: "at least one year of achievement delay" with Achievement; "be- low average intellectual functioning" with Intellectual Screening; "at least one year of achievement delay and below average intellectual functioning" with Development; "places blame on others" with Somatic Concern: "few or no friends, complaints of peer hostility and discrimi- nation, and fights with siblings" with Depression: "father is strict disciplinarian, uses excessive physical punishment, is alcoholic or substance abuser, and is emotionally disturbed" with Family Re- lations: "places blame on others and disobeys parents" with Delinquen- cy; "unrealistic fears and has few or no friends" with Withdrawal: “manifests anxious, tense, nervous, and restless behaviors" with An- xiety; “seldom communicates“ with Psychosis; "prior stimulant therapy, overly active or agitated" with Hyperactivity; and "suicidal thought and/or self destructive behaviors and has few or no friends" with Social Skills. The authors suggested that the scales with few external correlates may reflect infrequent symptomatology in the sample (SOM, PSY), insuf- ficient number of certain types of items on the checklist (ACH, IS, DVL), and a lack of reliable documentation of parent attitudes in medi- cal records for relating information pertinent to the scale (DEF). In addition to providing support for the clinical utility of the instrument, examination of the relationship between the evaluation of the quality of interaction between mother and child, as described in medical records, and the profile scales yielded theoretically consis- tent results which demonstrate PIC convergent validity. Mothers who 38 were described as inconsistent in setting limits, using substantial physical punishment, strict disciplinarians and defensive about discus- sing their personal lives were more likely to describe their chil- dren as displaying externalizing, acting-out behaviors. This was evi- denced in scale elevations (higher mean scores) on the DLQ, HPR, and SSK scales. The authors hypothesized that the mothers were likely to model aggressive and impulsive behaviors and reinforce the child for them. Mothers described as overly concerned or protective, extremely rejecting or critical, and displaying psychopathology severe enough to require treatment generally described their children as demonstrating internalizing behavior. This was seen in elevations on the D, ANX, $0M, WDL, and PSY scales. These data were interpreted as suggesting that mothers who display anxious, depressed, or withdrawing symptoma- tology reinforce their children's tendencies to display similar behavior by modeling and characteristic interaction with the child. Lachar and Gdowski (1979a) studied the relationship between em- pirically derived dimensions of problem behaviors and the PIC profile scales. The study sought to extend the Lachar, Butkus, and Hryhor- czuk (1978) study by using a larger and more representative sample of both children and adolescents, by examining behavioral dimensions that are likely to be viewed as more stable than individual checklist items, and by studying both convergent and discriminant validity. A sample of 430 children and adolescents (272 males, 159 females) between the ages of 2 and 17 was used for the study. The majority of the subjects (78%) were in the lowest two categories of the Hollings- head Redlich socioeconomic classification system (Hollingshead, 1957) which estimates socioeconomic status from data on the education and 39 occupation of the household's highest income earner. There was a rela- tively even distribution of black (42%) and white (57%) subjects. The subjects had been referred for psychiatric evaluation at Lafayette Clinic by various sources such as parents, school personnel, mental health agencies, private psychiatrists, courts, and family physicians. The subjects' symptomatology varied, with primary GAP (Group for the Advancement of Psychiatry) diagnoses including healthy responses, re- active disorders, developmental deviations, psychoneurotic disorders, personality disorders, psychotic disorders, psychophysiologic disord- ers, organic brain syndrome, and mental retardation. Psychiatric residents with two to four years of training completed checklists which consisted of 100 items reflecting problem behaviors in affect, physical development and health, family relations, and maternal characteristics. Their judgments about the children were based on in- formation obtained on the initial parent and teacher questionnaires, parent and child interviews, and supervisory contact made with other child psychiatrists. The senior author of the study and a fourth-year psychiatry resident independently evaluated 24 cases on a preliminary version of the evaluation form. The interrater agreement was 87%. Factor analysis of the checklist yielded 16 interpretable factors, with 78.5% of the common variance accounted for. These factors in- clude hostility/dyscontrol, cognitive and neurological impairment, an- tisocial behavior, somatic symptoms, maternal criticism, psychotic dis- organization, parent conflict, sleep disturbance, perfectionistic, an- xiety/depression, emotional lability, suiCide intent, social with-_; drawal/limited social skills, maternal overconcern, poor school adjust- ment, and fearfulness. A relatively‘large amount of common variance 40 was represented by the first three factors (33.6%) which reflected problems with self-control, hostility, intellectual and neurological functioning, and antisocial acts. Correlation of the PIC profile scales T scores with the derived problem behavior factor scores provided substantial evidence for scale convergent and discriminant validity, as well as support for the diag- nostic potential of the PIC. The significant correlates suggest cer- tain interpretive guidelines for each scale: L (males tend to be more controlled and thoughtful in planning and antisocial: impulsive beha- viors would be unlikely as presenting problems), F (females are likely to show delinquent, hostile, and impulsive behavior; males show suici- dal ideation; and male and female children evidence sleep disturbance and social withdrawal), DEF (no significant correlates found), ADJ (similar to the F scale), ACH (children showed cognitive and neurolo- gical impairment, limited social skills, and poor planning), 15 (im- paired cognitive and neurological functioning), DVL (impaired cognitive and neurological functioning, limited social skills, withdrawal, and uncontrolled behavior), SOM (male children and adolescents evidenced somatic concern and response to stress, with symptoms of fatigue and headaches), 0 (males showed sleep disturbance, social withdrawal, and suicidal ideation; female children and male adolescents showed anxious and depressed behaviors), FAM (children were inconsistently disciplined by mothers and had parents experiencing marital discord; adolescents showed hostile and impulsive behaviors), DLQ (children showed social withdrawal and limited social skills; adolescents evidenced lack of deliberation and control; and both groups showed hostile, impulsive, and antisocial behaviors), WDL (children had impaired cognitive and 41 neurological functioning, antisocial behaviors; males showed an absence 0f hostile and impulsive behaviors; and all showed limited social skills and withdrawal), ANX (similar to the 0 scale; children showed withdrawal and limited social skills; males had sleep disturbance and suicidal ideation and behavior), PSY (all showed sleep disturbance, so- cial withdrawal,and limited social skills; children showed cognitive and neurological impairment; males evidenced suicidal ideation and be- havior: females evidenced hostile, impulsive behavior; and male adoles- cents showed psychotic disorganization), HPR (hostile and impulsive behaviors; absence of maternal overconcern, overpermissiveness, and overprotection; and a lack of perfectionism), and SSK (social with- drawal and limited social skills). In this study, discriminant validation of the PIC was conducted by examining problem behaviors that did not correlate with PIC scales. ~An- example of this is the author's finding that higher scores on scales reflecting externalizing, aggressive behaviors (DLQ and HPR) were clearly not related to factors on the problem checklist which repre- sented symptomatology of internalizing behavior (anxiety/depression, suicide intent, and fearfulness). 0n the other hand, convergent va- lidity was demonstrated by the high correlations of the internalizing scales (0 and ANX) with factors representative of internalizing sympto- matology (sleep disturbance and social withdrawal). Although the re- sults of the study are important in providing support for the clinical utility and validity of the PIC, the authors recognize the need for further research in varied settings, such as the schools. 42 Criterion Related Validity Gdowski (1977a) studied the ability of the PIC scales to discrimi- nate among relatively homogeneous groups of disturbed children. The sample consisted of children and adolescents who had been referred by a number of sources to Lafayette Clinic in Detroit. The 307 subjects (190 males, 117 females) ranged in age from 2.6 to 17.11 years (7': 12.5). Psychiatric residents were asked to evaluate each subject on a 66- item problem behavior checklist which assessed self-concept, interper- sonal relations, affect, physical and cognitive development, and achievement. PICs were completed primarily by the biological mothers - of the subjects. When the biological mothers were not available, other females with whom the subject had resided for at least six months served as informants. The first seven factors which were obtained in the factor analysis of the problem checklist were peer relationship difficulties, psychotic symptomatology, authority conflict, somatization, cognitive disturbance. social withdrawal, and limited intelligence. These seven factors,which accounted for 69.7% of the common variance, were subjected to a cluster analysis, which allowed for the construction of eight homogeneous groups of subjects according to the patterns of disturbed behavior shown in the factor analysis. To determine whether the PIC was sensitive to the symptom patterns, an analysis of variance was used to compare the cluster groups with PIC profile scale scores. Post hoc comparisons allowed for the examination of cluster mean differences on the individual PIC scales. With the exception of the SOM scale, all clinical scales of the PIC differed 43 significantly across the cluster groups. The validity scales (L, F, DEF) and screening scale (ADJ) did not vary significantly across groups The results indicate that the PIC scales were sensitive to varied patterns of symptoms in a clinic population and were able to differen- tiate relatively homogeneous groups of behaviorally disturbed children and adolescents. DeHorn, Lachar, and Gdowski (1979) studied the ability of the PIC profile code type classification system to discriminate among homogene- ous groups of behaviorally disturbed children. The sample of 2,946 children included the normative sample and six of the criterion groups used in the scale construction. The criterion groups included children ages 6-16 who had been identified as delinquent (119 males, 32 females), cerebral dysfunction (63 males, 10 females), hyperactive (63 males, 17 females), somatizing children (35 males), retarded (85 males, 53 fe- males), and psychotic (56 males, 23 females). The PICs were completed by the subjects' mothers and then classi- fied by two systems, normals and code type, and factor cluster. Both systems were based on the relative T-score elevations (high mean scores) of the clinical scales. The profiles with all clinical scales having T-scores less than 70 were placed in a "normal limits" group. A code type system, where the highest or two highest scores were ex- amined, grouped the remaining profiles (those with T-scores greater than 70 on scales). A factor cluster method was also used to further group the profile code types into six possible combinations of three main factors (conduct disorder, cognitive development, internalization), which were suggested by the initial factor analysis of the PIC (Wirt et a1., 1977). 44 To determine if each profile classification system significantly differed across the six criterion groups, a chi-square statistic was applied. The results indicated that the code type classification sys- tem effectively separated the profiles of the six criterion groups. Although 17 code types classified 80% of the total sample, only six of these were statistically significant, therefore reducing the classi- fication of the sample to 55%. The researchers interpreted the results to indicate that individual code types of the PIC can identify concep- tually meaningful differences across relatively homogeneous clinical samples. For example, the FAM/DLQ code type was obtained only by adjudicated delinquents. Although two-point code types were infrequent for the normative sample, 19-20 code types classified 70-78% of the heterogeneous clinic sample. Since the sample employed in the study was used in the PIC scale construction, further research using another sample is needed to determine the utility of the code type classifica- tion system in differential diagnosis. Cichon (1979) studied the potential usefulness of the PIC scales to screen for personality disturbances in young children. PICs were completed by the mothers of 199 normal children and 65 children, aged 2% to 6%, evaluated for emotional difficulties. Mean scale scores for the 16 profile and 17 supplemental scales were obtained for both groups. The ANOVA resulted in significant dif- ferences between groups on 28 of the 33 scales. Although SES effects were significant on 14 scales, the effect proved to be less powerful than group membership. Data which were calculated from the normal sample showed SES effects, but greater main effects were found for age and sex. 45 To determine the ability of the PIC to differentiate among indi- vidual children, hit rates or the number of children correctly identi- fied as disturbed or normal according to a given cutting score were determined for each scale. The ADJ scale had an 80% hit rate, whereas DVL and DEF scales performed less well. The author concluded that the remaining scales might also prove useful in discriminating among groups; however, further research with specific criterion groups of disturbed preschool children would be needed. Cross validation of the PIC on amore representative sample would also be necessary to further investigate the usefulness of the ADJ scale to screen preschoolers for emotional disturbance. Berman (1979) evaluated the validity of the Spanish translation of the PIC (Harris, 1976) and studied its diagnostic potential in a school setting. The sample consisted of 23 public school children (14 males, 9 females) and 23 parochial school children (13 males, 10 fe- males) identified by teachers as displaying significant behavioral problems and/or learning difficulties. Another group of 21 children (12 males, 9 females) who, according to teachers, exhibited no signi- ficant school problems was also studied. All of the children were be- tween the ages of 6 and 10 and were in the lowest socioeconomic level of the Hollingshead Redlich classification system. Parents of the subjects completed the PIC, and teachers completed classroom behavior checklists. With the exception of the L, DEF, and HPR scales, the PIC profile scales significantly differentiated the two groups of children exhibiting learning and/or behavioral problems from the normal comparison group. Using the F, ACH, and IS scales to construct a regression equation, the discriminant function analysis 46 correctly identified 89% of the parochial school children. Socioeco- nomic effects were noted on the FAM, WDL, and PSY scales, with the lower SES group's obtaining higher scale elevations. By correlating the PIC scales with the teacher checklist data, an average of 8.8 significant correlates were obtained for each scale. Although the correspondence among these correlations and those of the English version of the PIC scales is poor, Berman's study is important in providing evidence for the stability of the PIC across cultures and the diagnostic potential in differentiating children with academic- related problems. Webb (1977) had conducted an earlier study to examine the ability of the PIC to discriminate groups of exceptional children according to differences in personality and adjustment. Although a total of 374 children (200 males, 174 females) were included in the study, only 221 subjects between the ages of 6 and 16 were used for the phase of Webb's research which explored the relationship between PIC content scales and teacher classifications. The identified groups of children in- cluded emotionally disturbed (29), learning disabled (23), mentally re- tarded (6), culturally deprived (10), gifted (34), and normals (119). Mothers completed the PIC, a 68-item problem behavior checklist, and a family information sheet pertaining to the occupational and edu- cational levels of the parents. Teachers filled out the problem beha- vior checklist and a teacher classification form. Other academically- related data were drawn from school records. The results indicate that within a number of categories of excep- tionality, as classified by teachers, one or more PIC scale elevations were present. Although no scales were significantly elevated in the 47 learning disabled group, the DVL scale was elevated in the mentally retarded group; and in the emotionally disturbed group, the interna- lizing scales, WDL and ANX, and the DVL, FAM, ASO (asocial), and RD (reality distortion) scales were elevated. The ASO and RD scales are among the ten content scales which Klinedinst (l972) initially con- structed and later revised (Klinedinst, 1975). The culturally de- prived group had elevations on the A50, FAM, and DVL scales. No sig- nificant elevations were found on any scales for the gifted and normal children. The PIC content scales were significantly correlated with the home behaviors identified by mothers and school behaviors reported by teachers. Mothers did, however, report more problem behaviors than teachers, and both mothers and teachers noted fewer difficulties for females. The author interpreted these findings as demonstrating the value of using the PIC in classifying exceptional children, particu- larly when usedirlconjunction with school classification systems. The combination of data provided a more complete description of the pat- terns of problem behavior. This study provided important information regarding the potential utility of the PIC with school populations of exceptional children; however, the gross discrepancy in the cell sizes and the procedure for classifying the children necessitate further investigation with a lar-- ger and more representative sample of children whose classifications are based on criteria other than teacher nomination. Culbert and Gdowski (1982) sought to determine whether the PIC scales could effectively differentiate be tween a reading-disabled group of children and a normal comparison group, as well as assess ’ 48 whether the scales were sensitive to variations in cognitive abilities. The reading-disabled group consisted of 12 males, aged eight and nine years, who had been identified as dyslexic, as defined by the World Federation of Neurology. The boys were recruited from reading dis- ability programs in the metropolitan Detroit schools. The control group of 12 males were from the same area and were matched to the dys- lexics according to the Peabody Picture Vocabulary Test IQ (PPVT), SES, age, race, and handedness. All subjects were right-handed Caucasians whose families' minimal annual income was $12,000. The demographic characteristics for the control and reading-disabled groups, respec- tively, were: age (9.4, 8.7 years), PPVT IQ (113, 107), and the Pea- body Individual Achievement Test reading standard score (111, 83). Subjects were excluded if interviews revealed information suggesting a history'of medical, neurological, or emotional problems that could interfere with the development of reading skills. One father and 23 mothers completed PICs, and each child was ad- ministered the Weschler Intelligence Scale for Children--Revised (WISC-R). The results showed that the reading-disabled group had high- er PIC scale elevations on ADJ, ACH, IS, and DVL. Correlation of PIC scales and WISC-R subtest scores and Verbal, Performance, and Full Scale IQs revealed significant relationships that clustered around the ADJ, ACH, IS, and DVL scales. Elevations on ADJ, ACH, and DVL scales were correlated with Kaufman's (1975) "freedom from distractibility" factor on the WISC-R, defined by the child's scores on the Arithmetic, Digit Span, and Coding subtests. Elevations on the ACH scale were also related to Kaufman's "verbal comprehensiod'factor (Information, Simi- larities, Vocabulary). The IS scale was sensitive to variations in all 49 these factors defined by Kaufman: "verbal comprehension" (Information, Vocabulary, Comprehension), "perceptual organization" (Block Design, Picture Completion), and "freedom from distractibility" (Arithmetic, Digit Span). The authors concluded that the data reflected the ability of the PIC to discriminate the two groups according to cognitive and academic dimensions and that the three scales constructed to identify cognitive and academically-related skills are related to variations in academic achievement by children of average intelligence. Further Applications of the PIC In a study intended to examine the differential treatment response and long-term adjustment of children being treated with methylpheni- date, Voelker (1979) found that the PIC profile scales were able to predict response to stimulant medication for children diagnosed as hyperkinetic. The PIC was chosen as an independent variable, since multiple externally-validated dimensions of child functioning are as- sessed by the inventory. Subjects were drawn from a population of 691 children and adoles- cents evaluated at Lafayette Clinic. Initial subject selection of a potential subject pool of 131 was based on the resident psychiatrist's recommendation for stimulant medication treatment. However, to pre- vent the inclusion of cases with difficulties which could produce simi- lar symptoms, further exclusionary criteria were imposed. Subjects whose medical records indicated possible brain damage, psychosis, men- tal retardation (IQ< 70), or stimulant medication treatment at the time 50 of the initial evaluation were excluded. Of the 88 remaining subjects, only 49 were chosen for the initial phase of the study. The selection was based on the availability of progress notes which documented medication response for the first two months of treatment (minimum daily dosage = 10mg. methylphenidate). The sex dis- tribution was in favor of males, with 39 males and 10 females. There was a relatively even distribution of black (20) and white (29) sub- jects. The ages ranged from 4.8 to 15.2 years (7'= 8.9), and IQs were between 72 and 139 (7'= 93.39). The modal socioeconomic level was the lowest of Hollingshead Redlich's classification system (Hollingshead, 1957). Using information taken from the medical records, improvement rat- ings for home and school behavior and academic achievement were made by two judges. Interrater agreement was sufficiently high to allow for the combination of the two sets of ratings. The adult who came in for the psychiatric interview for each subject being evaluated at the clinic was asked to complete the PIC. Informants were primarily biological parents; however, responses were in some cases obtained from step- parents, adoptive parents, and grandparents. The judges' ratings of improved behavior were correlated with the PIC scales, yielding eight significant correlates. Initial HPR eleva- tions were (positively correlated with behavior improvement at home and school. Improved home behavior was negatively correlated with the L, DVL, WDL, and ANX scale elevations. School behavior improvement was negatively correlated with ACH and 15 scale elevations. While evidence of cognitive deficits, developmental delays, and internalizing sympto- matology were counterindicative of improvement, behaviors frequently 51 associated with an attentional disorder with hyperactivity (HPR) served as the best predictor of medication response. The author concluded that the PIC profile scales were able to predict response to stimulant medi- cation for children diagnosed as hyperkinetic. Pipp (1979) studied the relationship between PIC profile scales and adolescents' self reports on the MMPI. The design followed that of Lachar and Gdowski (1979a) in determining the levels of the PIC scales which would accurately predict MMPI self reports. The sample consisted of 247 adolescents (140 males, 107 females) seen at Lafayette Clinic for psychiatric evaluation. During the eval- uations, the adolescents were required to complete the MMPI, and their parents were asked to respond to PIC items. The MMPI items and the PIC scales were correlated. MMPI corre- lates supported the antisocial and impulsive indications of the DLQ and HPR scales, the internalizing symptomatology as measured by the WDL and ANX scales, the physical concerns and worries of the SOM scale, family interactional difficulties as seen on the FAM scale, and problematic social relations of the SSK scale. Pipp (1979) concluded that the scales were useful in predicting the self report of adolescents, providing further support for the interpre- tive intent of the PIC. These investigations have provided substantial support for the validity (construct and criterion-related) and diagnostic utility of the PIC. The PIC was able to differentiate across relatively homogene- ous groups of children; however, the samples were drawn primarily from psychiatric populations. The few studies which were conducted using school populations likewise provided evidence for the potential use of 52 the PIC in school settings. The researchers, in general, used large samples and controlled for the effects of age, sex, and socioeconomic status, making interpretation of the results more clear. A summary of these studies is presented in Appendix B. Summary of the Research To summarize the literature reviewed in the preceding pages, past studies using traditional standardized intelligence and achievement tests and self report psychological inventories have shown that emo- tionally disturbed, learning disabled, and educable mentally retarded children appear more similar than different, with no single character- istic or group of characteristics associated with any of these special education categories. Although Bernard's (1978) research confirms this finding, he has shown that the inclusion of multiple variables in the diagnostic process will increase the degree to which these mildly handicapped children can be discriminated. The majority of studies involving differential diagnosis, however, have included mostly cogni- tive variables, which have been few in number. As seen in these studies, a frequently employed method in discrim- inating among emotionally disturbed, learning disabled, and educable mentally retarded children has been the analysis of WISC-R profiles. Although the results of these studies have often been confusing and contradictory, there is little doubt that these children share a number of intellectual characteristics which make differential diagnosis based solely on cognitive measures extremely difficult. Researchers who have examined personality as well as cognitive variables have been faced 53 with a similar problem when attempting to classify these mildly handi- capped children. According to the previously cited studies, under the current classification system the problem of misclassification seems inevi- table. With the rise of mandated services for the handicapped and the subsequent necessity of assessment, a more effective means to evaluate children with special needs is warranted. Review of the literature has shown that the PIC has potential to be a diagnostically useful instrument (Lachar, Butkus, & Hryhorczuk, 1978; Lachar & ‘Gdowski, 1979a). However, as seen in the PIC studies, the population from which the subjects were drawn was primarily that of children being seen for psychiatric evaluation at Lafayette Clinic in Detroit. Webb (1977) and Culbert and Gdowski (1982), however, provided support for the diagnostic utility of the PIC with children referred for psychological evaluation within the public schools rather than at psychiatric facilities. Both studies demonstrated that patterns of profile scale scores varied according to special education categories. Furthermore, the PIC was shown to be sensitive to variations in cogni- tive functioning (Culbert & Gdowski, 1982) and provided a more com- prehensive description of problem behaviors when used irlconjunction with a classroom behavior measure (Webb, 1977). CHAPTER III METHOD Subjects The subjects included 141 children and adolescents who range in age from 6 to 16 years. Each subject had been identified by school psychologists and placed by Educational Planning and Placement Com- mittees in special education programs for the learning disabled, emo- tionally disturbed, or educable mentally retarded. The special edu- cation programs were in a primarily white, middle class Detroit suburb of 70,000. With the exception of the learning disabled children whose place- ments in the schools were solely in academic support programs, the subjects were in either self-contained classes or mainstreamed in regular education while receiving support from the special education teacher. The subjects were grouped according to their diagnostic category and were not subgrouped according to their participation in resource programs versus self-contained placements. The learning disabled group included 70 subjects (51 males, 19 females), the emotionally disturbed group consisted of 47 subjects (36 males, 11 females), and the educable mentally retarded group was comprised of 24 subjects (10 males, 14 females). The actual breakdown of subjects according to diagnosis, sex, and age is presented in Table 3.1. As noted in the table, the subjects are grouped as children (aged 12 and younger) or adolescents (l3 and older). 54 55 Table 3.1 Age and Sex Distribution Children 512 yr. Adolescents > 12 yr. Learning Disabled (N=70) Males (N=51) 31 20 Females (N=l9) l6 3 Emotionall Disturbed (N=47) Males (N=36) 18 18 Females (N=11) 7 4 Educable Mentally Retarded (N=24) Males (N=lO) Females (N=l4) \IU'I N01 Initial subject selection was based on whether the child was identified according to one of the three categories of special edu- cation and was within the 6 to 16 year age range. Subjects were ex- cluded if either parent or teacher refused to participate in the study by completing the necessary forms. As seen in Table 3.1, the sex distribution was quite uneven, with considerably higher numbers of males than females. This is consistent with the research on sex-related differences which indicate that males are three times more likely than females to exhibit behavior problems (Gilbert, 1957; Becker, 1978). In light of the incidence figures pro- vided by the Joint Commission on the Mental Health of Children (Glide- well 3 Swallow, 1968) and other researchers (Hobbs, 1975), a 2 to 1 ratio of learning disabled to other educationally handicapped students is not unexpected. According to the subjects' scores on the Wechsler Intelligence Scale for Children--Revised (WISC-R), the group means for the learning 56 disabled and emotionally disturbed fall within the average range, whereas the mean IQ of the educable mentally retarded group is in the mildly retarded range. The WISC-R Verbal IQ, Performance IQ, and Full Scale IQ means and standard deviations are presented in Table 3.2. Table 3.2 WISC-R IQ Scores .L2 .52 Eli: Mean 5.0. Mean 5.0. Mean S.D. Verbal IQ 90.76 11.83 90.72 14.49 66.88 11.03 Performance IQ 97.76 13.16 97.85 13.45 67.74 12.99 Full Scale IQ 93.47 11.29 93.45 13.48 65.26 12.20 Instruments Personality Inventory for Children The Personality Inventory for Children (PIC) is a 600-item, objec- tive, multi-dimensional personality measure (Wirt, Lachar, Klinedinst, & Seat, l977). screening, and 12 clinical scales. sented in Appendix C. The profile scales consist of three validity, one The complete PIC inventory is pre- The PIC was constructed in the same manner as the Minnesota Multiphasic Personality Inventory (MMPI); however, the PIC is completed by a secondary informant, generally the child's mother. The profile scales were developed using one of two methods, em- pirical and rational (Wirt et a1., 1977). The method of contrasting groups was used in the empirical scale construction, as was Darlington 57 and Bishop's iteration of validity comparisons for developing the scales in stages to gain optimum validity. Content-oriented and in- ternal consistency methods were used to construct the rational scales. Studies have provided substantial evidence for the reliability (Wirt et a1., 1977) and construct validity (Lachar, Butkus, & Hryhorczuk, 1978) of the PIC. The reader is advised to refer to the literature review for further description of these and other relevant PIC studies. Wechsler Intelligence Scale for Children--Revised The Wechsler Intelligence Scale for Children--Revised (WISC-R) is a revised and restandardized version of the original scale (Wechsler, 1974). According to the manual, the standardization was based on stratification from the 1970 U. S. Census. A total of 2,200 cases (100 boys and 100 girls at each of 11 age levels from 6% through 16%) were used for the standardization. The scale has twelve subtests, six verbal (Information, Similarities, Vocabulary, Comprehension, Arithmetic, and Digit Span) and six performance (Picture Completion, Picture Arrangement, Block Design, Object Assembly, Coding, and Mazes). Ten subtests are mandatory; while Digit Span and Mazes are optional. Each of the subtests is equally weighted to yield three 105: Verbal, Performance, and Full Scale which is a composite of the first two. The WISC-R is intended to be used with children between the ages of 6 and 16. The potential IQs in the normative range are from 45 to 155, with a mean IQ of 100 and a standard deviation of 15 (Y subtest scale score a 10). 58 Kaufman (1975) examined the factor structure of the WISC-R at eleven age levels, ranging from 6% to 16% years. Using the standardi- zation sample reported in the WISC-R manual, Kaufman factor analyzed the scale, yielding three factors: Verbal Comprehension (Information, Similarities, Vocabulary, and Comprehension); Perceptual Organization (Block Design, Object Assembly, and Picture Completion), and Freedom from Distractibility (Arithmetic, Digit Span, and Coding). According to Kaufman, the "verbalcomprehension" factor measures verbal knowledge and comprehension generally obtained from formal edu- cation, the "perceptual organization" factor measures non-verbal di- mensions which reflect the ability to interpret and/or organize visual input, and the “freedom from distractibility" factor measures the ability to attend or concentrate. No age effects were found; there- fore, these factors can be applied to any child. A number of researchers have conducted similar factor analytic studies of the WISC-R by using data collected from mildly handicapped children, such as those diagnosed as emotionally disturbed, learning disabled, and educable mentally retarded. With few exceptions these groups demonstrate similar factor loadings, with verbal subtests load- ing on the "verbal comprehensiod'factor, the performance subtests loading on the "perceptual organization" factor (Schooler, Beebe, & Koepke, 1978), and factor loading on the "freedom from distracti- bility" factor (Peterson & Hart, 1979). The WISC-R manual has documented that the test is acceptably re- liable, as measured by test-retest and internal consistency methods. The reliability coefficients for the subtests range from .70 to .86 and for the IQs, .90 to .96. However, the reliability coefficients 59 are a function of numerous variables; therefore, the reader should refer to the manual (Wechsler, 1974). Several studies that have ex- amined predictive validity have shown that the WISC-R is a valid predictor of achievement as measured by standard achievement tests such as the Peabody Achievement Test (DeBell & Vance, 1977). In ad- dition, factor analytic studies have provided substantial evidence for construct validity (Kaufman, 1979). Teacher Rating_and School Information The Teacher Rating and School Information form (TRSI) is a 109- item checklist which rates classroom behaviors and academic achieve- ment. The form is generally completed by the child's teacher. Fifty items are drawn from a published behavior rating form, the Walker Problem Behavior Identification Checklist (Walker, 1970). The Walker checklist was developed for use by the classroom teacher in the identi- fication of children with behavior problems and was intended to be used as a supplement in the total diagnostic process, rather than an in- strument to classify children. The remaining 59 items on the TRSI ask teachers to evaluate a child's level of academic achievement, provide possible reasons for observed learning difficulties, and make recommen- dations for further educational programming. The TRSI form is pre- sented in Appendix D. As the TRSI is a newly developed instrument, few studies using the form have been conducted. There are no studies specifically designed to estimate the reliability and validity of the TRSI. Support has, however, been provided for the reliability and the predictive validity 60 (Walker, 1967) and construct validity, as estimated by a factor analy- sis (Gdowski, 1977b), of the first 50 items of the TRSI or the Walker checklist. Gdowski (l977b) submitted the Walker checklist to a factor analy- sis, yielding seven factors which accounted for 79.2% of the common variance. These factors include emotional'hbility/peer relationship difficulties, social withdrawal, distractibility/impulsivity, anxiety, poor self concept, autism, and excessive internalized standards. Procedures Researchers (Gdowski, Lachar, & Butkus, 1980) have shown that ob- taining data directly from secondary respondents, such as parents, re- duces errors of omission and distortion that often occur when informa- tion is taken from children's case records. Parents and teachers were, therefore, asked to provide information pertaining to the behavior of the child under study. Special education teachers for the learning disabled, emotionally disturbed, and educable mentally retarded in the school district were initially contacted to determine whether they would agree to participate by completing a problem behavior checklist, the Teacher Rating and School Information form (TRSI). Upon receipt of the teacher's consent to be involved in the study, the parents of potential subjects were sent a letter describing the study and requesting their participation by completing "a personality inventory" (PIC) and giving their permission to release psychological records. Parents interested in participating were required to sign consent forms and return them to the researcher in care of the Psycho- logical Service department of the school district. The PIC was then 61 mailed to the parents for them to fill out on their children, with directions to return the completed inventories within 10 days. Addressed, postage-paid envelopes were provided to assist in the re- turn of the consent forms and test materials. Following receipt of the parental consent forms, participating teachers were sent the TRSI forms through inter-school mail. A re- quest was included with the TRSI that the forms be completed and re- turned by the end of six weeks in the same manner they were sent. The school information was collected during the second half of the school year to assure that teachers would be familiar with the children's classroom behavior and achievements. While waiting for the return of the PICs and TRSIs, the researcher examined psychological case files of the children to obtain test re- sults from the Wechsler Intelligence Scale for Children--Revised (WISC-R). Cases with WISC scores only were to be eliminated since re- search has shown that WISC test scores are consistently higher than WISC-R scores (Paal, Hesterly, & Wepfer, 1979). When the WISC-R, the revised form of the WISC, was published in 1974, the school district's psychologists stopped using the WISC. Since legislation requires re- evaluation of special education students every three years, no cases were excluded because of this criterion. Since the major portion of research on the PIC has used the mother as the informant and the norms are based on responses from mothers (Wirt et a1., 1977), the present study included only those subjects whose biological mothers, step-mothers, adoptive mothers, or female guardians completed the PICs. Thirty-three percent of the 533 parents 62 contacted agreed to participate in the study: however, only 141 ac- tually completed and returned the inventory. The PICs were hand scored using the 16 templates for scoring the profile scales. These scores were converted to T-scores, using the conversion tables in the manual (Wirt et a1., 1977). The profile scale scores were then subjected to the analyses intended to test the hypotheses of the study. It should be noted that clinical scale ele- vations, denoted by higher numerical scores, indicate the presence of problem behavior. The TRSI form is scored according to the presence or absence of problem behavior, with higher scores denoting more dis- turbed characteristics. Unlike the PIC and TRSI, higher scores on the WISC-R indicate higher levels of intellectual functioning. Hypotheses and Data Analyses Preliminary descriptive statistics such as means, standard devia- tions, and correlation coefficients were computed to explore the rela- tionship of the PIC profile scales to each other and to determine characteristic group profiles. Variables included group membership (learning disabled, emotionally disturbed, educable mentally retarded), age (children, 6-12 years; adolescents, 13-16 years), sex (male, fe- male), and test scores (PIC, WISC-R, Walker factors taken from the TRSI). The status of the variables, whether dependent or independent, varied according to the data analyses. The Statistical Package for the Social Sciences, SPSS (Nie, Hull, Jenkins, Steinbrenner, & Brent, 1975), was used for all the computer analyses. In order to determine the diagnostic potential of the PIC with a special education popula- tion, the following hypotheses were tested. 63 Hypothesis 1: The Personality Inventory for Children (PIC) will discriminate among three groups of special education children, learning disabled, emotionally disturbed, and educable mentally retarded, according to their profile scale patterns. Analysis of Hypothesis 1 The SPSS program MANOVA was used to perform a profile analysis which determined whether the learning disabled, emotionally disturbed, and educable mentally retarded groups differed according to their PIC profile scale patterns. The dependent variable was the PIC profile scale scores, and the independent variable was diagnostic category. Given a statistically significant difference (p<<.05) among the groups, univariate analyses of variance across the 16 profile scales were planned to identify the scales which were significantly different. To increase the power of the statistical test, planned comparisons were performed and tested at the .05 level of significance. Since considerable research suggests that learning disabled and emotionally disturbed children are more difficult to distinguish, one apriori-planned test examined the difference between these two groups according to PIC scale scores. The second planned comparison was be- tween the educable mentally retarded group and a combined group of learning disabled and emotionally disturbed subjects and was conducted in light of indications in the literature that educable mentally re- tarded children are more easily distinguished from the other two diag- nostic categories. Since previous research with the PIC suggests that sex and age differences exist (Lachar & Gdowski, 1979b), three-way analyses of 64 variance among age, sex, and diagnosis were conducted and tested at p (.05. Hypothesis 2: A $1 nificant discriminant function using multiple variables IPIC profile scales, WISC-R IQ, Walker factors, sex, age) will predict membership in the learning disabled, emotionally disturbed, and educable mentally retarded groups. Analysis of Hypothesis 2 In order to maximally discriminate among the three special educa- tion groups, the SPSS program DISCRIMINANT was used to perform a mul- tiple discriminant analysis. The dependent variable was group member- ship (learning disabled, emotionally disturbed, educable mentally re- tarded), and the independent variables were the PIC scale scores, WISC-R IQ, Walker factor scores, sex, and age. In addition to examin- ing the predictive efficacy of the discriminant function, the relative contributions associated with each variable were examined. This was done by performing a series of four separate discriminant function analyses, with particular attention given to the ability of the PIC to correctly classify subjects. Hypothesis 3: There will be significant negative correla- tions between the cognitive scales of the PIC (Achievement, Intellectual Screening, Development) and the Wechsler In- telligence Scale for Children--Revised (WISC-R) subtests and 105. Analysis of Hypothesis 3 A correlation matrix was computed to determine whether the cogni- tive triad (Achievement, Intellectual Screening, Development) of the PIC was sensitive to cognitive variations, as measured by the WISC-R. Pearson product moment correlations were obtained among the 65 Achievement, Intellectual Screening, and Development scales of the PIC and the WISC-R subtest scores, Verbal IQ, Performance IQ, and Full Scale IQ. Further examination of the construct validity of the cogni- tive scales, by means of discriminant validation, was established through the correlation of the WISC-R subtests and IQs with the non- cognitive scales of the PIC. The significance level was set at p«<.001 because of the large number of correlation coefficients tested. Limitations and Assumptions of the Study The subjects consisted of a sample of special education students in a suburban school district who were identified by school psycho- logists as being either emotionally disturbed, learning disabled, or educable mentally retarded and placed by Educational Planning and Placement Committees in special education. The process involved in identifying these children and the possible pressures from teachers, parents, and administrators for or against a particular placement were not examined in this study. It was, however, assumed that all subjects were appropriately classified. By this it was meant that the psycholo- gists diagnosed the children according to federal guidelines, Public Law 94-142, and the Educational Planning and Placement Committees simi- larly followed the law in placing the children in their respective programs. Beyond meeting the age criteria of 6 to 16 years and placement in one of the three special education programs, subjects were included in the study on the basis of both their teachers' and parents' willingness to participate. Since the sample was not randomly selected, 66 generalization of the results is limited; however, the accuracy of the completed measures should be greater due to the voluntary nature of the informants' involvement in the study. Since the Wechsler Intelligence Scale for Children--Revised was not administered to the subjects during the study, rather the data were taken from case records, inaccuracies of test information due to errors in administration and scoring are not controlled for. With regard to the Teacher Rating and School Information form (TRSI), the accuracy and comprehensiveness of the observations of the special edu- cation teachers completing this form may vary according to vari- ables such as the number of hours spent with the child and the variety of situations in which they have encountered the child. Since both the PIC and TRSI were administered after the children were diagnosed and placed in special education programs, it is possible that the parents' and teachers' knowledge of the diagnostic classifi- cation may have influenced their responses. It was, however, assumed that most parents have rather limited information about the specific category under which their children are placed; therefore, the effect should have been minimal. CHAPTER IV RESULTS Included in this chapter are the descriptive statistics, results from the testing of hypotheses and supplementary analysis. The sup- plementary analyses include data on the reliability and validity of the Personality Inventory for Children (PIC). Descriptive Statistics Personality Inventory for Children Learning disabled group. The mean PIC profile scale scores, as seen in Table 4.1, show that the learning disabled group had its higha est score on the Intellectual Screening (15) scale and second highest on the Development (DVL) scale. An elevated Achievement (ACH) score collaborates these results which indicate that parents of learning disabled children described them as having their greatest weakness in areas of cognitive functioning. The learning disabled subjects had a relatively high score on the F scale and low scores on the Lie and Defensiveness (DEF) scales. This suggests that parents of these children are aware of their children's difficulties and are less apt to deny commonly occurring problems or be defensive about their children's behavior. The learning disabled group's profile was basically defined by peak elevations on the IS and DVL scales and secondary elevations on 67 Table 4,1 Score Means and Standard Deviations Across PIC Profile Scales SOM MEAN S.D. DVL MEAN S.D. IS MEAN S.D. ACH MEAN S.D. ADJ MEAN S.D. DEF MEAN S.D. LIE MEAN S.D. MEAN S.D. qummmc FOQNNNM CNQSOLOmN Fannie-mix meemmmv \DLDQF-N‘Dm NF:(\.IO;NOC l_l'—l— r-r-N (tsp—[~50 NQ'MNQ‘SDQ’ \DACOtONNSD COCONOSM commemosco O‘NOitONNM NVNNwQND momentous QCYGCOC Dede—AND l—r— l—l—l—N mgommoomo memmxoux’ori were-ecce- NmNNr—Nfi' éééfiééé l-l-‘I—l—NNF- Uzdzfil'NNKDN omuimrxcow tomLGKONONOO mmomooo VG’LOQ'l-Omm ‘UF-P'UF- 0r- 0 gr- 0 PST! :6 Q U '0 U U GU E H- 68 I-:<:r-N<:NO éNddmfi'N F—l— VOFOLOSDO NMMMOO‘N mmmmms-rm LOONr-OMN MQ'NOSNOLD MONFNOO LOLOO‘NQ’OF: O‘DSO‘OSDOQ mr-ziod'Q'NDO «id-dr-mmu; NNNNNNI— NGM‘OPNLD d-«icshihiaiua NNmon‘D Q'Or-CNQN OOTO'SO'SQ'O‘M I- r- N LONOQ'MI-Tw PéNONMC ‘0‘0‘0‘0‘0‘0‘0 @NVCVCN SOLOQ‘LDF-NN r-F-F-I—NNN GNOMOECONO SONSMMF:§'1~D NNQNNNO equiv-ecu Q'erflafi'SON SONNLDMMLO oKdéé§d VMMQ’LOLOQ' govmrxsor-q: QNONNNG 0—:— MCDr-Q'SOMN dOVNfi'mm LOADSOLOLDLDLD Q'l—OMFOM Q'Q'Q'Q’LDLOLD 'UF-F-‘Ul— Ul— O gl— 0 F-S'U .C'U CU U CO 0) U (U CE ‘4- LLJ NNMPNLON l—l—‘F'l—‘l— l— \OQSOQOTCSD mongasoom Q'Nl-OQLONO; LDI-Ol-OQ'LDLDLO meONGC NNQ‘I—I—NSO r-F-r-F-g—r- VONOO‘r-Q' I—QOOTQ’Q'OOW QNNNmGN N‘DCONI-zma‘l «iuiésomeu; F-F-l—l— OQVNMOO gorxommtocr COiOOH—I—v— !— F- l—I-F' enhance-moo NNNNNLéLn Officer—03¢ nmmgooomm QQOQQNO O‘NQMQQ’N SDOQ’QG C39; l—NNF-l—I-l— NNQ‘OOMN NO'; QNPN NN QNNN r—NQ'oFl-nm NmmNomN r—r— F‘F' OQONFGM 0300611511501“: «moped-mm ”NI—.mQMI— mNmNmLDo— l—NN l— LDND‘O‘OOOM LOOSNONOLO commonest) coo—hmf‘OMQ' mind-minnow 'DF-r-‘UF- d)!— O GUI- O F-S'U E-C'U O: ‘5 U f5 0 U ‘6 55 Q- 1nued l, cont Table 4. SSK MEAN S.D. HPR MEAN S.D. PSY MEAN S.D. ANX MEAN S.D. WDL MEAN S.D. DLQ MEAN S.D. FAM MEAN S.D. MEAN S.D. NVNVCDVN mange-sow.— manor-mo NNLAQSOSQSC LDLOLOLOLDLDO I—I—O‘NLDONI— . O O O O . f-l—‘mMI—l—F' NO‘LDMr-mo r-:I-—CVl—I-:O LOLDLOLDLDLDIO N‘DN‘DNNIO OPP-:gtDNO’; l—F-l—l— NO‘NMPmN ”MFNMNV mmmmmmm O‘OMMVNLO VNu—VONOO r-II—t—r-I-NI— VLO‘OLO‘DLDN \D‘Ole-DVI— mmmmmmm VNI-OI—ch oommmmm l—F- F- QNMNOEMM NI—to;V‘D‘OC mmemm‘D \OQLDF-NNO LDLOLOOI-DLDLO Uv—r-‘UF- UF 0 EF- 0 l—S‘U 2.0 a U (U a) U ‘U Q ‘0- _.I 69 MLDVwam MFONGDNN eon—000mm VVOSONNM $000000“) LOVLOLO‘OTOS VVVVMVM r-t—v—r—I—r-t— ”NVOQGM NdFNI—Md mmtommmd' VNNNLO‘O‘O ONOSMVLDI— l—‘I—O—l—t—f—‘F- coeoomcomoo NGMVQNN LDLOLOLDLOCOLD VF-VNONF- MNNNQNO‘I F-l—‘l—l-‘f—F-F- VGMLONNQ mmeNr-O LOLDQLDLDOLD ~00}:qu NF-No—‘moa; FPPFFFF VMQNNOSCD mmdchV mmmmmmm NNmmmmm NwdeéN r—r-r—I—t—t—l— mazommoao NNNMMON \OQDOIOOLO v-zmanNl-zN émmO‘DMNQ NMQQKONLD IONNONMO l-OLOLOLOLDLOLD F-r—VO‘mF-N uncommon) 'UI—F-‘Ut— d)!— 0 gr— 0 l—SU CU E U (U (D U ‘6 a '4- NNCDMLDOC NVLDNOQN FMNVQNP NNMPOOM \DkDNfiD‘ONKD cameos-eon OOVAOOC‘60 F-l—F- GOer—mm oooomohfioooa VVVVVVV memOfiN OMOLONVO Nle-‘l—F—N mmeMVl— éVMVOmr-z NNQON‘DN WQNQQMN l—r—F-Nl—O;V I—r-NCOSVM N‘DQVNCLO mmmmmmm QNMMI—MV ouoncufim F-f-l— TONmI—‘Q‘D oénmcdmo OO‘DLDI-Ol-DO can—oatme- NVNr-dml— c—r-N r- CO‘OQVO‘O compose-o mmmmmmm TNOVNF-V O;I-:LI;VNKSQ l—F- VOCDNOO‘F: t—chnu—‘mw mmmemV l—l—mml—F—‘F- commence 'UF-l—‘Ur- G)!— 0 f6!— 0 I—S'U ECU a: CO U M 0) U6 EE 9- 70 the ACH and Adjustment (ADJ) scales. Other than this, the profile was relatively flat, suggesting that these children demonstrate little psychopathology. Emotionally disturbed group. The emotionally disturbed group had its highest scores on scales measuring psychological adjustment (ADJ) and intellectual functioning (IS). Although it is not unexpected that emotionally disturbed children would be described by parents as being psychologically maladjusted, below average intelligence is surprising. However, as seen in Table 4.l, there was marked variability with the emotionally disturbed group on the IS scale which suggests that emo- tionally disturbed children cannot be characterized by deficient in- tellectual functioning. Relatively high scores on ACH and DVL scales indicate that these children exhibit delays in their development and academic achievement. Emotionally disturbed subjects were further described as having difficulty with both internalizing and externalizing behaviors. This was evidenced by elevations on the Depression (D), Delinquency (DLQ), and Social Skills (SSK) scales. There were also relatively high scores on Anxiety (ANX), Psychosis (PSY), and Hyperactivity (HPR) scales, which again identify acting out and internalizing psycho- pathology. The lowest scores were on the Lie and DEF scales, suggesting that it is unlikely that parents of emotionally disturbed children are de- fensive about their children's behaviors or attempt to describe them in a more virtuous manner. The Somatic Concern (SOM) scale was the 71 lowest clinical score, indicating that emotionally disturbed children are not perceived to have excessive somatic complaints. Educable mentally retardedggrgup. Educable mentally retarded sub- jects had their peak elevation on the IS scale. Table 4.l shows that there was little variability on this scale and the score far exceeds other scales. The perception of cognitive deficiency was collaborated by the group's having its second highest score on the DVL scale and a relatively high score on the ACH scale. Besides cognitive limitations, parents of educable mentally re- tarded subjects described them as having significant behavior problems. Relatively high elevations on the ADJ, PSY, and SSK scales suggest that educable mentally retarded children are psychologically maladjusted and have poor social relationships. The educable mentally retarded subjects had their lowest scores on the Family Relations (FAM) and HPR scales, suggesting that parents describe them as being less apt to evidence disturbed family relations and overly active behavior. Examination of the mean scale scores across all three groups in- dicates that the subjects were described as having their greatest dif- ficulty in cognitive functioning and psychological adjustment. Al- though the groups had different degrees of elevation on the scales which identify these behaviors (ACH, IS, DVL, ADJ), each group had its primary or secondary elevations on these scales. Similarly, the groups had their lowest scores on scales which predict family disturbances and somatic concerns (FAM, SOM), suggesting that the subjects, regardless of group membership, have the least problems in these areas. 72 In terms of group variability on the scales, the learning dis- abled, emotionally disturbed, and educable mentally retarded subjects varied little on the DEF, SOM, and FAM scales. The groups were, how- ever, quite different on the DVL, ADJ, IS, and PSY scales. Wechsler Intelligence Scale In terms of intellectual functioning, as measured by the Wechsler Intelligence Scale for Children--Revised (WISC-R), the learning dis- abled group had scores which fall within the average range. There was minimal discrepancy between the subjects' verbal and performance skills, although there were lower scores on subtests measuring verbal abilities. The IQ means and standard deviations for each group are presented in Table 3.2. Emotionally disturbed subjects also scored within the normal limit of intellectual functioning. Again, little discrepancy was noted between the verbal and performance areas. The emotionally disturbed subjects did, however, perform slightly better on tasks requiring non- verbal skills. Educable mentally retarded subjects had IQ scores in the mildly retarded range. There was basically no difference in their verbal and visual-spatial skills, as the group demonstrated marked limitations in each area. As seen by the mean WISC-R 105, both the learning disabled and emotionally disturbed groups were of average intellectual strength, whereas the educable mentally retarded group was significantly below average. There was little discrepancy between the verbal and 73 performance areas across all three groups, suggesting rather even de- velopment of intellectual abilities for learning disabled, emotionally disturbed, and educable mentally retarded subjects. Hypothesis Testing Hypothesis l Hypothesis l: The Personality Inventory for Children (PIC) will discriminate among three groups of special education children, learning disabled, emotionally disturbed, and educable mentally retarded, according to their profile scale patterns. This hypothesis was supported since the difference among the three groups' mean profiles was statistically significant at the p< .0001 level. The mean profiles for all three groups are shown in Figure 4.l. Although a profile analysis confirmed the significant difference among the mean profiles of the learning disabled, emotionally disturbed, and educable mentally retarded groups, further analyses were needed to identify the scales which were significantly different across the groups. In order to increase the power of the statistical tests, apriori- planned comparisons were performed. Since the literature indicated that learning disabled and emotionally disturbed children are the most difficult of the educationally handicapped to differentiate, one planned comparison was conducted to examinewhich PIC scales were sig- nificantly different. A second apriori-planned test analyzed the difference among PIC scales of the educable mentally retarded group and the learning disabled and emotionally disturbed groups taken together. The comparison was set up to determine the accuracy of prior conten- tions that educable mentally retarded children are the most easily 74 e. A“? 9 \‘0 *6 T Q d“ S) I —_ _ — us; _ : no; _ 106— _ - ' .00.} - _ - 95-. _ 90-I ' I as—; ‘ 80+ _ - - : O 75-: - ‘ - o'. ro-i ‘ - : I .°. 1 : 2 . V - : ao—: ' ‘ — - 3'30 l’La:oIsFAouAcnus-oVLsomosAmoLowomePsvmanssxT Figure 4.1. PIC profiles for L0, ED, EMR subjects. Groups designated as Aare learning disabled, as o are emotionally dis- turbed, and as I are educable mentally retarded. 75 differentiated from learning disabled and emotionally disturbed chil- dren. The two planned comparisons are shown in Table 4.2. Planned comparisons between the learning disabled and emotionally disturbed groups showed that six PIC scales significantly differen- tiate the two groups at p<(.05 level of significance. These scales include the Lie, DEF, ADJ, DLQ, ANX, and SSK. As seen in Figure 4.2, the profile of the learning disabled group compared to the emotionally disturbed group was, in general, less elevated and, therefore, least suggestive of psychopathology. Although the learning disabled group did not produce the highest score on the Lie scale, the score was significantly more elevated than the emotion— ally disturbed group's Lie scale score. The learning disabled subjects, likewise, scored significantly higher than the emotionally disturbed subjects on the DEF scale. This suggests that compared to parents of emotionally disturbed subjects, parents of learning disabled children are more likely to deny commonly occurring problems and to be defensive about their children's behavior. Though not statistically significant, the learning disabled group had higher elevations on the F, ACH, and DVL scales. This suggests that the parents of the learning disabled group describe their chil- dren as having their greatest difficulties in areas of academic achievement and evidence marked developmental delays, both intellec- tual and physical. Subjects in the emotionally disturbed group produced the highest elevations, compared to the learning disabled group, on the ADJ, IS, 50M, 0, FAM, DLQ, Withdrawal (WDL), ANX, PSY, HPR, and SSK scales. Though the emotionally disturbed group was not significantly more 76 Table 4.2 Effects of Diagnostic Group on Profile Scale Elevations gigggoggggs C°CB7EBt sfifi?fBa§tso Pooled Variance Pooled Variance LQ_ §Q_ .Eflfl t Probabilities t Probabilities LIE 49.6 44.0 53.1 .008 * .001 * F 60.9 59.3 65.5 .597 .141 DEF 45.3 40.6 43.0 .044 * .999 ADJ 62.9 76.9 72.7 .000 * .444 ACH 63.1 61.5 67.0 .377 .033 * IS 72.2 74.7 108.0 .496 .000 * DVL 67.4 65.3 81.4 .385 .000 * SOM 51.0 52.4 54.3 .499 .313 D 56.5 61.0 61.5 .073 .357 FAM 52.8 56.2 51.4 .075 .182 DLQ 56.3 62.9 58.0 .022 * .654 WDL 53.7 56.4 60.1 .201 .043 * ANX 5l.7 59.4 57.1 .001 * .561 PSY 57.5 59.3 72.0 .549 .000 * HPR 53.9 57.8 48.9 .142 .028 * SSK 59.4 64.4 67.1 .033 * .060 * p (.05 77 Q r 0‘" 9 0‘3qu ‘9) 9+¢¢€é¢o€w a?! 0‘3 4 V 00' 6 v1SP 940 +36%? T ' 120-.__ - - - __ am ns-S _'-;_... 110—?- _-‘—::_"‘_:-§-no we; "“"-3: -§-ws ... -—:§.... _- .. 90.5 ‘:_‘ ‘-_':‘:--:_§..o so-fi‘ - -Z§... ..-z 1-}. f F‘(/: - - - g“ : - 2.. 3..., is; _ _ €40 I : _ ‘ - - - 3-45 - — - - _ - _ :45 _ _ - i—ao E' ousow 0 pm DLOWDLANXPSYHPFISSK .T Figure 4.2. Profiles for planned comparisons: LD and E0. Groups designated A are learning disabled, and as o are emotionally disturbed. 78 elevated on the IS, SOM, 0, FAM, WDL, PSY, and HPR scales, the scores on the ADJ, SSK, DLQ, and ANX scales were significantly different from the learning disabled group. This suggests that relative to parents of learning disabled children, emotionally disturbed children's parents perceive their children as more poorly adjusted, having their greatest difficulty interacting with same-aged peers, exhibiting delinquent-type behaviors, and being highly anxious. The emotionally disturbed subjects, compared to the learning dis- abled subjects, scored lowest on the Lie, F, DEF, ACH, and DVL scales. The emotionally disturbed group was significantly lower on the Lie and DEF scales, suggesting that these parents are least likely to deny com- monly occurring problems and be defensive about their children's be- havior. Apriori-planned tests between the educable mentally retarded group and the learning disabled and emotionally disturbed groups combined revealed seven PIC scales which separateithe groups at the p <.05 level of significance. These scales are Lie, IS, ACH, DVL, WDL, PSY, and HPR. The educable mentally retarded subjects' mean scale scores showed that in comparison to the pooled mean scores of the learning disabled and emotionally disturbed subjects, the educable mentally retarded group had significantly higher elevations on scales measuring cognitive abilities (ACH, IS, DVL). This suggests that parents of educable men- tally retarded children view them as having marked intellectual defi- cits, developmental delays, and below average academic achievement. Parents of educable mentally retarded subjects also described their children as exhibiting internalized symptomatology. as evidenced by significantly elevated PSY and SSK scale scores. 79 The educable mentally retarded group was also significantly higher on the Lie scale which suggests that these parents are more likely than parents of learning disabled or emotionally disturbed subjects to deny behavior problems which occur normally in childhood and adolescence. This may be a result of the parents' tendencies to ascribe greater psychological adjustment to their children who they recognize as having serious intellectual deficiencies. Though the difference between the educable mentally retarded group and the combined learning disabled and emotionally disturbed group on the F, 50M, 0, and SSK scales was not statistically significant, the educable mentally retarded subjects had higher elevations. The educable mentally retarded group had lower scores on the FAM and HPR scales. The HPR scale, however, was the only one which was significatnly dif- ferent. This suggests that educable mentally retarded subjects are least likely to be described as having behaviors characteristic of the hyperkinetic syndrome. These profiles are shown in Figure 4.3. As seen in the analyses, the first hypothesis which states that the PIC can discriminate among learning disabled, emotionally dis- turbed, and educable mentally retarded children, according to their profile patterns, was supported. Since recent research with the PIC indicates that sex and age differences exist (Lachar & Gdowski, 1979b), a series of three-way analyses of variance were performed to determine the effects of these variables on the performance of subjects on each of the PIC scales. The results of the analyses are presented in Table 4.3. Examination of the table reveals that few age-specific and sex-specific differences 80 ‘3 q. ‘9 9' 5‘ o ‘ O 0 -& 4‘ 3 ““9 as?" f” 6523’ or“: so" a“? 99" 9‘0 3’4, 55$ «39" a a? e fee: .e e «we ,f’fy .o a e T ‘ « 79> v9 \9 e c at‘ e9 ,6 eP° T has; _ - - - _ - Jar ns-j— _ : " ‘ ‘ : - ‘ _ _ - §.ns 110—: - ’ - _ ' ' - Z—iio 5 - - _ 9. _ 7 7 - - 3 105+ ‘ " °"_ " - - : 1.105 100-: - - — .' ‘ ‘ - I I - ' - ‘ 2 340° I 9 - _ _. - - _. _ I 95—: _ _ . — _ _ - ‘ j—gs so—f ' : _ - ‘ .' _ ‘ - ‘ - f-so as—j ‘ - - z _ ' ‘ ' j—as eo-f - : - - - - f—oo 75—3 - : ‘ - " — - - 345 ro-: - j ' _ L - - - f—ro : - _ __ - .o, : as—j - _ - '. - 3 j—oa 2.5—} 3.55 50-2 3-50 4o—j 1.“, 35.; 3.35 30-3 :-30 T 0 FAM oco WDL mx PSY HPR ssx T Figure 4.3. Profiles for planned comparisons: EMR and LD/ED. Groups designated 0 are educable mentally retarded, as A are learning disabled and emotionally disturbed. Diagnosis (A) Agei(B) Sexg(C) .07 .29 .55 ACH FAM DEF ADJ DVL SSK LIE IS ANX PSY SOM WDL DLQ HPR *p< .05 1 .80* .26* .16* .40* .20* .12* .16* 7.66* NNNN .40 .70 .87 .32 .83 81 Table 4.3 Effects of Diagnosis, Sex, and Age on Profile Scale Elevations .30 .08 6. .01 8.45* 12. .79 3. 2.70 5 5.96* .65 3. 2.31 2 .28 1.59 .51 1 .11 2. .58 .70 1.75 .12 88* .40 14* 92* .90* .22 40 .06 .15 .03 .48 67 .23 .01 .24 Ass. 2. 56 .42 .66 .47 .21 .34 .53 .14 .76 .23 .07 .11 .35 .27 .54 .50 M ,4.49* .89 3.50* 3.46* .28 2.43 1.29 1.47 .30 .59 .33 .55 .03 as .52 .57* .02 .33 .06 .83 .40 .10 .07 .96 .46 .01 .07 .08 .01 .04 AxBxC 3.56* .07 .36 .88 .68 .39 .10 .15 .82 .01 82 exist. The sex and age differences which are present, however, are not consistent enough to require separate analyses. Significant main effects for age were found on the DEF and SSK scales. DEF scale means suggest that parents of adolescents tend to be more defensive about their children's behavior. Lachar and Gdowski (1979b) had a similar finding in their study. These results may be due to the fact that adolescents tend to exhibit fewer problem behaviors (Quay & Werry, 1979). Mean scale scores on the SSK scale, however, suggest that the difficulties which adolescents have frequently in- volve social interactions. A significant main effect for sex was found on the ADJ scale. According to scale means, males demonstrateimore difficulty in their overall adjustment. This is consistent with the literature in child- hood psychopathology (Achenbach, 1966; Quay a Herry, 1979), as well as previous research with the PIC (Lachar & Gdowski, 1979a). The scale designed to identify developmental delays (DVL), both intellectual and physical, reflected sex differences. Mean scale scores indicate that females are described more often as developmen- tally immature. This finding, however, may be related more to the fact that more females in the study have been placed in programs which focus on intellectual rather than emotional difficulties. The significant two-way interaction between sex and age on the ACH scale suggests that female children, 12 years and younger, are more likely to be described as having poor academic achievement. This is not surprising since academic difficulties are more often the reason for children, particularly females, to be referred for special educa- tion services. 83 The interaction between diagnosis and sex on the FAM scale indi- cates that female learning disabled subjects and male emotionally dis- turbed subjects are more likely to exhibit problems in family rela- tionships. There was a significant two-way interaction between diag- nosis and sex on the DEF scale. Examination of the mean scale scores suggest that parents of emotionally disturbed and educable mentally retarded subjects are more likely to be defensive about the behavior of female children than males. This may be interpreted as an indica- tion that females present fewer behavior problems than males (Quay & Herry, 1979). A significant three-way interaction was found on the F scale. Examination of the mean scale scores indicated that parents of learning disabled males are more likely to exaggerate their children's behavior- al difficulties when they are adolescents rather than children. The opposite was true for parents of learning disabled females who tended to ascribe more severe symptomatology when the child was young. Al- though there was little difference in the way parents of educable men- tally retarded subjects emphasize the problems which their female chil- dren and adolescents have, this was not the case for male educable men- tally retarded subjects. Of particular interest was the finding that parents of male educable mentally retarded children, ages 12 and under, were most apt to exaggerate their children's difficulties. Significant main effects for diagnosis were shown on the Lie, ADJ, IS, DVL, ANX, PSY, and SSK scales. Planned comparisons, as shown in Table 4.2, indicate that while parents of learning disabled subjects, compared to emotionally disturbed subjects, were significantly more apt to deny commonly occurring problems, parents of educable mentally 84 retarded subjects, compared to the combined group of learning disabled and emotionally disturbed subjects, were the most likely to deny daily problems and ascribe more virtuous behaviors to their children. This finding was demonstrated by the subjects' performance on the Lie scale. The ADJ, ANX, and SSK scales showed significant main effects for diagnosis. According to apriori-planned tests, emotionally disturbed subjects, compared to learning disabled subjects, were described more often as psychologically maladjusted, anxious, and demonstrating poor social relationships. Hypothesis 2 To further examine the usefulness of the PIC to differentially diagnose mildly handicapped children and adolescents, a discriminant function analysis with multiple variables was performed. Since spe- cialists have access to WISC-R data for decision making, but not the PIC or Walker, a second hypothesis was formulated concerning the abil- ity of these measures to validate the diagnostic decisions of the specialists. Of particular interest was the diagnostic power of the PIC in relationship to other related measures. Hypothesis 2: A si nificant discriminant function using multiple variables ?PIC profile scales, WISC-R IQ, Walker factors, sex, age) will predict membership in the learning disabled, emotionally disturbed, and educable mentally retarded groups. This hypothesis was supported by a set of discriminant function analyses which correctly classified 78.6% of the total subjects, based on 16 PIC scales, WISC-R Full Scale IQ, and Walker factors,sex and age. A series of four separate discriminant function analyses were performed to determine whether the discriminatory power of the analysis 85 was increased by the addition of certain variables, beyond the PIC. It was of particular interest to assess the ability of the PIC to accu- rately predict group membership without the assistance of related mea- sures such as the WISC-R and Walker checklist items. The first discriminant function which used the WISC-R Full Scale IQ, sex, and age as variables demonstrated significant group mean dif- ferences for two of the three variables, WISC-R IQ and age. Table 4.4 shows the relative contribution of the two discriminating variables. In order of effectiveness, the WISC-R IQ and age significantly separ- ated the three groups at the pi<.00001 level. As seen in Table 4.5, the WISC-R IQ had a far greater discriminative power than age. Application of the equation to the three groups resulted in the correct classification of 43 learning disabled subjects, or 61.4%. Ef- fective assignment for the emotionally disturbed group was 20, or 42.6% and for the educable mentally retarded group, 22 were correctly as- signed, or 95.7%. The total percent correctly classified was 60.7%. The results of this analysis are presented in Table 4.6. The second discriminant analysis used the PIC profile scales, sex, and age as variables to discriminate among the three groups. Ten of the 16 profile scales (IS, ADJ, PSY, Lie, ACH, ANX, D, SSK, HPR, WDL) and age contributed significantly to the total discrimination of 73.1% of the total subjects. As seen in Table 4.4, each of these variables was significant at the p< .00001 level. Examination of Table 4.4 reveals that two significant functions are present, one reflecting intellectual development, and the other psychological adjustment. As noted, however, by the eigenvalues in 86 Table 4.4 Contribution of Variables in the Discriminant Function Analyses Stand. Coefficient Variables Wilks Lambda Sig. Level Func. 1 Func. 2 Discrim. Anal. #1 WISC-R IQ .5719 .0000 .9931 .1233 AGE .5566 .0000 -.1598 .9878 Discrim. Anal. #2 £19_ IS .6835 .0000 -.8288 .1569 ADJ .5867 .0000 -.1291 .8479 PSY .5502 .0000 -.4465 .7770 LIE .5232 .0000 -.3703 .1834 AGE .5029 .0000 -.3384 .2768 EIQ_ ACH .4876 .0000 .1276 .5350 ANX .4717 .0000 -.0657 .7908 D .4497 .0000 .5077 .9313 SSK .4278 .0000 -.4390 .5841 HPR .4190 .0000 .3211 .0419 WDL .4094 .0000 .2475 .3102 Discrim. Anal. #3 WISC-R I! .5701 0000 - 7667 .0271 £l§_ DJ .4766 .0000 .3838 .8709 IS .4282 .0000 .6053 -.0771 DVL .3966 .0000 -.4359 .5333 PSY .3827 .0000 .1802 .5543 ANX .3680 .0000 .3578 .7077 D .3432 .0000 -.6917 .7981 AGE .3326 .0000 .2285 .2028 £19_ SSK .3207 .0000 .2492 .0271 LIE .3145 0000 .1887 .1712 HPR .3090 0000 - 1938 .1127 Discrim. Anal. #4 WISC-R I L .5701 .0000 -.8304 .0441 PIE. DJ .4766 .0000 .3552 .8149 IS .4282 .0000 .5927 .1470 DVL .3966 .0000 -.4386 .4937 PSY .3827 .0000 .2500 5768 ANX .3680 0000 .3298 7060 D .3432 0000 -.7168 7623 WALKER ACT OUT .3314 0000 -.1164 .3380 ‘AQE .3209 0000 .2439 .1416 PIC SSK .3096 0000 .1437 .5810 'WAEKER DISTRACT. .3026 0000 .1461 .2018 £19. HPR .2968 0000 -.l966 1394 Discriminant Analysis 87 Table 4.5 Relative Ability of the Discriminant Function to Differentiate Groups —J (a) (b) (C) (d) Ev U A U in Q—v ‘U A d) '4- : Cm A O V G) S- 0 CU ‘PC .0 Q) C V 0‘0 U 52 IV ‘28 «98 «a 3' 85 '4': 5.4-, :0) 0'5 44> W‘U U) G‘U ’f-l— DU 0): 0'!- Un— XD I SO :0) U): OF- S-& CS- PS “I- ma) 05> 00-: 00-66 mm 3Q) 0P!!! a: 01$- '1'“, DLL LLI> 0-) LLD Z—l 0 DM- (13—1 0 .5566 79.97 4 .0000 1 .7678 97.93 1 .9840 2.20 1 .1383 2 .0162 2.07 0 .4094 118.79 22 .0000 1 .7278 63.75 1 .7073 46.061 10 .0000 2 .4139 36.25 0 .3090 153.84 22 .0000 1 1.3091 76.53 1 .7135 44.22 10 .0000 2 .4015 23.47 0 .2968 158.53 24 .0000 1 1.3164 74.33 1 .6874 48.91 11 .0000 2 .4547 25.67 number of discriminant functions used relative ability of each function to differentiate groups (greater power = higher value) relative percent by which each function separates the groups number of functions that contribute significantly to group separation decreasing contribution made by each function (less dis- criminating power = higher values) 88 mm.mn o.mm m.w m.¢ om N F ¢.m m.¢m F.mp mm m m m.e _.NP e.me m N, mm .mam .mm .mm macaw umuuFumca Ama< .xmm .oHa .maumH3 .mux4<3y oo.ow m.pa m.a m.a _N F _ m.N_ N.o~ o.ep 8 mm m m.N m.ap m.~m N o. mm mam. .mm “a“ naoew cmuupumga Axum .mu< .oHa .m-um_zv mo.m~ m.nm o m.- pm 0 m m.~. ~.Pc m.m~ a mu N_ m.NF 4.,” “.me a m mm mam. .mm .mm macaw cmuuwumga Axum .mw< .uHav —~.oo «.mm m.¢ o mm p o 8.0. o.~a w.ea m cm NN p.“ e.Fm a.pe m NN me Mafim wa Nwfl macaw umuuwuoga Axum .mua .m-8m23v um_mwmmmpo Appumcgou acmogma Peach em mzm me am as a; macaw Pmauo< mommu co Longs: mmwxpmc< cowuuczu pcmcwswgumpo an mcovumopmvmmmPo uomggoo mo «pack xgmsszm o.¢ mpnmh 89 Table 4.5, the function which represents intellectual development has far greater power to separate the groups. Table 4.6 shows that this discriminant function equation correctly classified 75.7% of the learning disabled group or 53 subjects, 61.7% or 29 emotionally disturbed subjects, and 87.5% or 21 subjects in the educable mentally retarded group. Compared with the first discriminant function analysis, this equation was superior in classifying emotion- ally disturbed subjects. The third discriminant analysis included as independent varibles the PIC, WISC-R IQ, sex, and age. Nine of the 16 PIC profile scales (ADJ, IS, DVL, PSY, ANX, D, SSK, HPR, Lie), the WISC-R IQ, and age form a function that was significant at the p‘(.00001 level. Examination of the standardized function coefficients in Table 4.4 reveals two signi- ficant functions, one reflecting intellectual potential and the other psychological adjustment. As seen in Table 4.5, the function which represents intellectual functioning accounts for a far greater per- centage of the variance. Eighty percent of the total subjects were correctly assigned to their respective groups. In terms of predicted group membership, the equation effectively classified 58 or 82.9% of the learning disabled group, 33 or 70.2% of the emotionally disturbed group, and 21 or 91.3% of the educable mentally retarded group. These results are presented in Table 4.6. Prior to performing the four discriminant function analysis which included factors from the Walker as one of its variables, it was neces- sary to factor analyze the Walker. Six interpretable factors which 9O accounted for 99.9% of the common variance were found. Items with loadings of .35 or greater on the six factors are presented in Table 4.7. Factor one which accounts for 40.7% of the variance was charac- terized by variables such as argues, openly strikes back, disturbs other children, displaces physical aggression, and does not obey. The content of the factor clearly reflects acting-out behavior, along with disturbed peer relations. The second factor which accounts for 19% of the variance had sig- nificant loadings from items which suggest social withdrawal. This factor includes items such as does not initiate relationships with children, does not engage in group activities, and has no friends. The third factor extracted accounted for 13.6% of the variance and is characterized by items like is easily distracted, has difficulty concentrating, and is underachieving. Since the item content suggests problems with concentration, the factor was labeled distractibility. The fourth factor had significant loadings from items such as refers to self as dumb, when teased or frustrated takes out frustration on inappropriate person or thing, and comments that nobody likes him. This factor which accounted for 9.9% of the variance was named poor self concept. Factor five accounted for 9.4% of the variance and was labeled de- pression since items included listless and continually tired, habitu- ally rejects the school experience, somatic reaction to stress, and does not complete tasks attempted. The sixth factor was characterized by items which reflected an- xiety. The factor accounted for 7.3% of the variance and consisted of Table 4.7 Item Composition of Walker Problem Behavior Factors ITEM LOADINGS Factor 1: Acting,out[0isturbed Peer Relations (40.7%) Argues and must have last word in verbal exchanges .79 Openly strikes back with anger when teased by peers .79 Disturbs other children by teasing, fighting, etc. .79 Does not obey until threatened by punishment .74 Displays physical aggression .72 Does not conform to limits without external control (.72) Distorts truth by contrary statements .72 Has rapid mood shifts .67 Becomes hysterical if intents are thwarted .66 Reacts with defiance to instructions or commands .57 Continually seeks attention .56 When teased or irritated, displaces anger (.53) Complains of others' unfairness or prejudice toward him (.49) Apologizes repeatedly for self .44 Makes distrustful or suspicious remarks of others' actions (.40) Has temper tantrums . Expresses concern that something terrible will happen (.36) Factor 2: Social Withdrawal (19.0%) Does not initiate interaction with peers .81 Does not engage in group activities .66 Has no friends .62 Tries to avoid calling attention to self .61 Does not protest when hurt by others .60 Shuns or avoids heterosexual activities .57 Other children act as if he were tainted .44 Comments nobody likes him (.43) Utters nonsense or babbles to self (.35) Factor 3: Distractibility (13.6%) Easily distracted from task by ordinary stimuli .68 Has difficulty concentrating for length of time .68 Underachieving .64 Is overactive, restless, constantly shifting position .48 Utters nonsense or babbles to self (.41) Repeats one idea, thought or activity over and over (.37) Does not conform to limits without external control (.37) Frequently stares blankly into space and unaware .37 92 Table 4.7,ycontinued Factor 4: Poor Self Concept (9.9%) Refers to self as dumb When teased or irritated, displaces anger Comments nobody likes him Expresses concern about being lonely and unhappy Is hypercritical of self Complains of nightmares Repeats one idea, thought or activity over and over Expresses concern that something terrible will happen Makes distrustful or suspicious remarks of others' actions Comments that nobody understands him Complains of others' unfairness or prejudice toward him Factor 5: Depression (9.4%) Is listless and continually tired Habitually rejects the school experience Reacts to stress by somatic complaints Does not complete tasks attempted Factor 6: Anxiety (7.3%) Weeps or cries without provocation Will destroy own product rather than show it Expresses concern that something terrible will happen Steals from other children Has nervous tics ( ) = item loads on other factors 93 items like weeps or cries without provocation, will destroy things he has made, expresses concern about something terrible happening, and has nervous tics. The subject to variable ratio was less than four to one; therefore, the factor weight may be relatively unstable. However, these factors are very similar to those derived by Gdowski (l977b), using a larger psychiatric population. The factor scores which were generated for all subjects were used in the final discriminant function analysis which was quite effective in separating the learning disabled, emotionally dis- turbed, or educable mentally retarded groups. In addition to the Walker factors, the analysis included the PIC, WISC-R IQ, sex, and age as in- dependent variables. The results indicated that the WISC-R IQ, eight of the PIC scales (ADJ, IS, DVL, PSY, ANX, O, SSK, HPR), two of the six Walker factors (Acting Out, Distractibility), and age significantly discriminated 78.6% of the subjects at pi<.0001 level. These results are shown in Table 4.4. _ As in the second and third analyses, there were two significant functions which reflected cognitive skills and emotional adjustment. The function associated with cognition, however, had the greatest discriminative power, as determined by the eigenvalues. These values are shown in Table 4.5. This discriminant function correctly assigned 78.6% or 55 learning disabled subjects, 47.5% or 35 emotionally disturbed subjects, and 87% or 20 educable mentally retarded subjects. The classification re- sults are summarized in Table 4.6. The discriminant analyses disclosed that two functions, cognitive and emotional, consistently differentiated among the special education 94 groups across all four analyses. The use of successive analyses de- monstrated that prediction, particularly with emotionally disturbed, was markedly improved by the use of the PIC. Inspection of Table 4.6 shows that despite the frequent misclassification of learning disabled subjects as emotionally disturbed subjects, and vice versa, accurate prediction of group membership for the emotionally disturbed group could be increased from 42.6% to 70.2% by the use of the PIC, in ad- dition to the WISC-R. The fourth discriminant function analysis confirmed the second hypothesis which states that with multiple variables, including the PIC, Walker factors, WISC-R IQ, sex, and age, a discriminant function will accurately predict membership into learning disabled, emotionally disturbed, and educable mentally retarded groups. Hypothesis 3 A third hypothesis was formulated concerning the construct valid- ity of the PIC profile scales. Although testing of the first two .hypotheses revealed that the PIC is useful in differentiating groups of mildly handicapped children, further information pertaining to the validity of particular scales, specifically the cognitive triad (ACH, IS, DVL), was desired. Hypothesis 3: There will be significant negative correla- tions between the cognitive scales of the PIC (ACH, IS, DVL) and the Wechsler Intelligence Scale for Children-- Revised (WISC-R) subtests and IQs. The hypothesis was confirmed by negative correlations among the ACH, IS, and DVL scales of the PIC and the WISC-R subtests and 105, significant at the pi<.OOl level. As previously noted, greater score elevation on the PIC clinical scales denotes problem behavior, whereas 95 higher WISC-R scores indicate higher levels of intellectual function- ing. A correlation matrix is presented in Table 4.8. Inspection of the table reveals that the significant negative correlations primarily cluster around the PIC scales designed to measure cognitive function- ing (ACH, IS, DVL). There are, however, a number of significant corre- lations with the PIC scale constructed to identify poor reality test- ing (PSY). The relationship between the WISC-R data and the PSY scale may be explained by the fact that the PSY scale has been shown to reflect developmental retardation and cognitive confusion (Lachar & Gdowski, 1979b). As seen in Table 4.8, the significant correlations are be- tween WISC-R subtests which measure spatial organization and the PIC scale. The only other WISC-R subtest which was significantly related was the Comprehension subtest. The relationship with this subtest is not surprising since the PSY scale was intended to identify poor com- prehension of reality. The WISC-R Verbal IQ, Performance IQ, and Full Scale IQ were all negatively correlated at the pi<.OOl level with the ACH, IS, DVL, and PSY scales. As seen in Table 4.8, there were also significant correla- tions among the eleven WISC-R subtests and the IS and DVL scales. These correlations suggest that the two PIC scales are related to the three WISC-R factors, verbal comprehension, perceptual organiza- tion, and freedom from distractibility, which Kaufman (1975) extracted from the standardization sample. There were substantially fewer significant correlations between the ACH scale and the WISC-R subtests. Unlike the other two cognitive 96 ~00. v 04 00.: _F. 400.: 00.: 5F.: 00.: 00.: 0p.: 0—.: 400.: 400.: 400.: 00.: 00. 5F.: 0p.: d0.00<0m 0000 F0.: 00. 400.: 00. 0P.: 00.: p0.: 00.: 0—.: 400.: 450.: 400.: 00.: 00. 0P.: 0P.: 00 002<2000000 00.: mp. 400.: 00.: 0F.: p0.: P0.: 0~.: 00.: 400.: 400.: 400.: 00.: —p. 0~.: 0P.: m0 0<000> 400.: 00. 400.: 00. 0P.: 0F.: 00.: FP.: 0p.: 400.: 400.: 0P.: 05.: 00. 0~.: 00.: 000000 00.: 50. 400.: 00.: 09.: 00.: 00.: 0p.: 00.: 4P0.: 400.: 55.: 00.: 09. 00.: 00.: .Fnsmm0 pumwno 0F.: 0p. 400.: 00.: 0F.: 50.: 00.: 0P.: 0P.: 400.: 400.: mp.: FF.: 0p. ~0.: 00.: cmwmmo xuopm 00.: 0F. 55.: 0F. 00.: 00. 00. PF. 00. 400.: 400.: 00.: op. 50.: 00.: 00.: .0:gg< mgauuPa 0P.: 00. 400.: 00.: 00.: 00. F0. 00.: 00.: 400.: 400.: 400.: 50.: P0. 00.: 00.: .0500 0000000 mummunzm mocmseowema 00.: 00. 00.: 0_.: 0p.: 00.: 00.: 00.: 05.: 400.: 400.: 0p.: pp.: 00. 00.: 00.: 0000 00000 400.: 0F. 400.: 0F.: 00.: 05.: 50.: P0.: 0~.: 400.: 400.: 400.: 5p.: 5p. 0~.: 00.: cormcmcmgasou P0.: 05. 00.: P0.: ~0.: 00. 00.: 0p.: 00.: 400.: 400.: 400.: 00 00. 0P.: 0—.: agapznmoo> 00.: 00. 0~.: 00.: 5F.: 00. 00. 50.: 00.: 400.: 400.: 0p.: 00. 00.: 0P.: 00.: ovum50uwg< 0F.: 0p. m—.: 00. 00.: 00. 00. 00.: 00.: 450.: 400.: 0p.: 00. 00. 00.: 0—.: mmwuwgmpwswm 00.: _0.: 00.: 00.: 0_.: 00.: 00. 0p.: p~.: 4P0.: 400.: 400.: 00. 50. 00.: 0p.: 00000500000 mummunam Fmagm> :11: :II: :: 0:000: 2mm ma: >ma xz< do: can z00 xz< 00: 040 meowpmpwggou A5505 ..Fo pm 00030 0mpgm>o on «:0 covampmegou 00. 00. 00. F0. 00. 00. 00. 00. 00. 00.: 00. 00. 00. 00. 00. .0<|0 00. 50. 05. 05. 50. F0. 00. 00. 00. 00. 00. 0F. 00. 0p. 00. 0 00. —p. 00. 00. 00. 00. 00. 50. 00. 00. 00. 00. 00. 0F. 00. 001...... 00. 50.: 00. 0F. 00. 0F. 00. 00. 00. .04. 00.: 00.: 00. 00. 50. 50. 00. F0. 00. 00. 50. 00. 00.: 00.: 00. 0:0. 00. pp. 50. 00. 00. 00. 50. 00. 00. 05. 00. 50. 50. 00. 00. ad 00. 50.: 00. pp.: 00. 00.: 00. p0.: 50. 0p.: 00. 00.: 00. pp.: 00. 00.: 00. 00.: 00. 0p.: 00. 00.: 00. 00.: 50.: 00.: 00. 00. 00. 00. mag mupgm>0 2000 cu 0:0 mcovpmpmggoogmucm umpum0xu 0:0 mcopumpmsgoogmucu mpuum mpwmog0 00000000 0.0 m—nmh 00000000 u umgum0xm n 00. 00.: 00. 00.: 00. 00.: 00. 50.: 00. 00.: 00. 00.: 50. 00.: 00. 50.: 00. 00.: 00. 00.: 00. 00. 00. 00.: F0. 00.: 00.: 00. 00.: 00. m 001 wpm; 003c— mpm; 0000: x00 00: >00 xz< 003 000 z<0 200 0>0 00 :0< 00< 000 000 100 improbability of obtaining reports of externalizing and internalizing symptoms when the Lie scale is elevated. The correlations of the F with other PIC scales showed that the scale represents a wide variety of problems, as well as reflecting the informants' tendencies to exaggerate their children's difficulties. Interscale correlations suggest that the DEF scale is negatively re- lated to indications of poor psychological adjustment and the need for further assessment. With the exception of the cognitive scales, ACH, IS, and DVL, the ADJ scale was moderately correlated with all other PIC scales. The re- lationships indicate that the ADJ scale reflects acting out behaviors, as well as internalizing psychopathology. Although there was a rela- tively large number of shared items between the ADJ scale and the 0 and DLQ scales, it is unlikely that the high correlations are due to this item overlap. As seen in Table 4.9, the obtained correlations markedly exceed that which is expected from such overlap. It should be noted that the expected correlations are taken from the PIC manual (Wirt et a1., 1977). These correlations are given here to enable a comparison among the obtained intercorrelations. Since the expected correlations among the PIC scales are based on the number of items shared by the scales, the number of common items are presented in Table 4.10. ACH correlations indicate that the scale primarily reflects cog- nitive and developmental tasks. In addition to being moderately cor- related with the IS and DVL scales, the ACH scale was related to the SSK scale. This suggests that there is a secondary emphasis on social adjustment, a finding which is consistent with prior studies (Wirt et a1., 1977). 101 Table 4.10 Number of PIC Items which Overlap the Profile Scale Unique Scale x00 00: >00 xz< 003 0.5 z<0 :00 0>0 00 :0< 00< 000 Li.) H Item Number ..: 0 15 10 LIE DEF ADJ ACH IS DVL 3 -2 0 0 2 3 20 8 7 0 -1 -2 4 2 3 13 9 1 1 3 4 O 4 -1 11 0 O 0 0 0 O 34 10 26 17 12 10 20 15 10 SOM 1 -2 0 2 0 -1 0 5 2 12 2 -1 0 D O 1 0 D -2 0 O 9 0 0 1 0 FAM 0 3 -1 DLQ WDL ANX 3 -1 O 3 O 6 O 0 0 17 O 2 0 1 1 -1 1 -6 0 O 0 4 2 0 -1 l o -3 -2 2 -3 o 11 002020042 0 6 0 0 3 1 O 9 PSY HPR 7 0 -1 O 0 -1 SSK 102 IS correlations with other PIC scales suggest that the scale re- flects poor academic performance (ACH) and developmental delays (DVL). Again, the number of shared items was relatively large, but the ob- tained correlations exceeded that expected. The relationship between the IS and PSY scales may also reflect delayed development since, ac- cording to the manual (Wirt et a1., 1977), the PSY scale is related to delays in the development of pragmatic skills. The IS scale was nega- tively correlated with the scale constructed to identify hyperactive behavior (HPR). This suggests that with IS elevations, it is less likely that overly active, distractible behavior will be reported. As with the IS scale, the DVL scale was related to the PSY scale. This further suggests that developmental emphasis of the PSY scale, in terms of poor reality testing. The DVL scale was also correlated with the SSK scale, reflecting the scale's emphasis on social skill develop- ment as well as intellectual and physical development. Interscale correlations between DVL and other profile scales in- dicate that DVL primarily represents cognitive abilities (IS) and aca- demic achievement (ACH). SOM correlations suggest that the scale is related to depressive symptomatology (D), delinquent behaviors (DLQ), withdrawal (WDL), anxiety (ANX), psychotic symptoms (PSY), and problems with family relationships (FAM). These interscale correlations indi- cate that the SOM scale reflects both internalized, as well as exter- nalized symptomatology. The correlation with the F scale may suggest that considerable problem behaviors exist, and there is a need for fur- ther evaluation, rather than an exaggeration of difficulties. Correlational data suggest that the 0 scale reflects an overall poor psychological adjustment. Primarily the scale was related to 103 anxiety (ANX), psychotic symptomatology (PSY), and poor social skills (SSK). The 0 scale was also correlated with symptoms of withdrawal (WDL), delinquency (DLQ), and family relationship difficulties (FAM). This suggests that the scale reflects acting out, as well as interna- 1ized discomfort. FAM correlations with other PIC scales showed that the scale em- phasizes psychological maladjustment (ADJ), delinquent behaviors (DLQ), and poor social relationships (SSK). The DLQ correlations were related to a wide variety of problems. This includes externalized (FAM, HPR) and internalized (D, ANX) symptomatology. WDL correlations suggest that this scale is strongly related to internal discomfort. WDL ele- vations were associated with elevations on scales such as PSY, SSK, and 0, all of which reflect psychological and social withdrawal. ANX cor- relations with other 'PIC scales suggest that the scale reflects poor social relationships (SSK), somatic complaints (SOM), psychotic symp- toms (PSY), delinquent acting out (DLQ), and depression (0). Correlations among the PSY scale and other scales reflects inter- nalized psychopathology (D, ANX, WDL, SOM, SSK). To a lesser extent, it reflects acting out phenomenon, such as delinquent behaviors (DLQ). The PSY scale was also moderately correlated with the IS scale. This reflects the cognitive nature of the scale in terms of possible con- fusion and poor reality testing. HPR correlations indicate that this scale reflects general psychological maladjustment (ADJ) and delinquent-type behavior (DLQ). The SSK scale was highly correlated with the D, WDL, ANX, and PSY scales. As Wirt et al. (1977) noted, the association with the D, WDL, and ANX scales may represent the ef- fects of inadequate social skills, whereas the association with the 104 PSY scale suggest deficits which interfere with the development of ap- propriate social skills. Although few scales had significant overlap with other PIC scales, a number of significant relationships among these scales were found. Many of these relationships reflect the intent of the scales, while others provide reasons for the subjects' difficulties in particular areas. For instance, the moderately high correlation between the So- cialSkills scale and the scales measuring withdrawal, depression, and delinquency hwy reflect the consequences of poor social relation; ships, whereas correlations with scales identifying psychotic sympto- matology and developmental delays are likely to be related to the rea- sons for the lack of success in social interaction. These intercorrelations are very similar to a previous study by Wirt and his colleagues (1977). The results from both of these studies suggest that further grouping of the scales would not be ad- vantageous. Reliability and Validity In addition to testing the third hypothesis which examined the construct validity of the three PIC cognitive scales (ACH, IS, DVL), further validity information pertaining to the profile scales was de- sired. However, prior to examining the construct validity of the PIC by use of a behavior checklist, the Teacher Rating and School Informa- tion (TRSI) form, the reliabilities for each measure were calculated. Although earlier studies with the PIC (Wirt et al., 1977) had provided data regarding its reliability, obtaining an estimate of total PIC 105 scale reliability within a special education population was considered important. Using Cronbach's alpha, a measure of internal consistency, a re- liability coefficient of .83 was obtained for the PIC. The same pro- cedure was used to estimate the reliability of the TRSI. A coefficient of .88 was obtained. Since the measures were both found to be suffi- ciently reliable, the relationship between the PIC profile scales and teacher ratings of academic performance and school behavior were ex- amined by correlating the PIC scales and the items of the TRSI check- list. Significant correlations between the PIC profile scales and TRSI checklist items are presented in Table 4.11. The relationships found between the two measures provided further support for the diagnostic utility of the PIC profile scales in school settings, as well as sup- port for the PIC profile scale's construct validity. Where significant correlations were found, the dimensions tapped by the PIC appear to be particularly relevant to the school experience and lend support for the convergent validation of the PIC. On the other hand, the nonsig- nificant correlations may be explained by the absence of relevant items on the TRSI to correlate with the PIC scales. The lack of significant correlates may be viewed as a reflection of the poor performance of the PIC; however, other studies have shown that most of these scales which have few or no significant school correlates in this study have a num- ber of significant relationships using correlations obtained from other sources such as parents and psychiatrists(Lachar & Gdowski, 1979). As noted in Table 4.11, all significant school correlates of the Lie scale were inversely related to scale elevations. Minimal Lie 106 mpmm op cow“ -eauaa 0:000ae m00e>< .m_ mm. mm. om. 0m. om. 0N. mm.. mm. e~.- . 0:00:000 .mewmaau >0 0000.0;u m00=0m00 .0— 00. 00. 50. 0cw>mwgumgm000 .0. 00. 00:0» mg» mugoumwo .NP 000m 0cm >>0mumw0mg mm~00opo0< .Fp >0. 00. 00. 00.: mmmpummg .m>wuom apgm>0 .0_ 00. 00.: 00.: 0cpumgacmucoo >u_=uww000 .0 >0. 0000000 a.ce 00 we “ea mupx .m up 0cpzosm 0o ummpm:_ 00°: :30 m>ogum00 .5 umwcopuum0gm0 .0 8232000: 93 o: 930 .0 00. 00. 00. 000003;“ 00 _muwgmumxg mmeoumm .0 00. 00. 00. 00.: 00. 00.: «00200 on Egomcou 0o: mmoo .0 eacwp uea aaa_ua_0 .N 00. 00.: mmmcgwmmcs uzona m:_m>0eo0 .F 0mm 00: >2 :2 a: as 2,: 0 .zolm 49.0. fl :2 22 0.3 0 0: _m00 aea mapaem a000oca 000 eaazuam mee00apaeeoo 0=a8000e00m 0_.e a.aa0 107 00. 00. 00. >0. 00. 000cm mmum>0 :m>u .cmmmmu :00: .00 00. 00. mm>u>>>uum 0:000 c> m0m0cm 0.0mmo0 .00 00. 05:0 mm 0>mm op m0m000 .00 00. >0. m5:0ucma 0m05¢> .50 0m>o 0:0 0m>o you 00 mmc> mumo000 .00 5>0 mox>> >uoao: mucmEEoo .00 co>ucmuum mxmmm >>>m:c>uco0 .00 00. 00. 50. 00. 00.: 00. mo>0m>>>m mmcmmco: m0mupa .00 00. 000 mum: .00 mucm>0m0xo >oogum muumnm0 >>>m:u>am: .>0 mu>a m:o>0w= mm: .00 eu.:mu 0e 00>: mxmmu 30: mmsumo000< .0> 00o: «mo> mumw:::mm:00< .0> 00. mmzuo >060 00>: mmm0um o» macaw: .5> >0. m0oguo 0:600 m00o5¢0 >:0um:0um>u mmxoz .0> 0mm 00: 50 :5 5: g .20 0 am. a m. =2 22 to :a3e00eoe .>>.e a>aa0 LLI LU H _J 108 00. 00. 00. cm. mm. 00. cm. _0. 0p. 00. 00. 00. 00. m0. xmm «a: >m0 xz< 0a: 300 z_0um _0:xmmogmum0 00:00 00. 00:0 0003 ammuoga 0.00000 00mm» mumpaeou 0.0mmoo mpmm mo Pauppvgugmqaz m>memg00m xppmuwmx00 mm. om. 0P.u 00. 00.- P0. Fo>ogaam acmpmcou muwmz mucmwg$ 0: mm: :0. cmggmg __*z m_0000m0 9.. 239.25 Ufl:L@U:OQ 0cwmumu op xum0 mmxrgum apcmao 00000:: .apmcop mp m>0m 0000500000 00 0000,0500 o0. ~0.n newscmwc:a 00_3 000000000» Ppug: 0000 0.00000 00.:m woos :_a~m zu< do< mun 0 NH :m:=_0=ou ._..¢ mpamh .me .00 .me .00 .pe .oe .0m .00 .00 .mm .mm .00 .mm .00 .pm 109 00. 00. 00. 00. 00. mm. 00. 00. m0. 00. 00. 00. 00. 00. 00. 00. 00. 00. 00. 00. 00.- mm. 00. xmm mm: >m0 .mmw .000 0| xz< 0a: 000 z<0 00. 00. 00. 00. ON. mm. 00. 00. 00. 00. 00. 00. 00. 00. 050000000 000:0000 .00 000000000 000 0.00000 .00 00. .00000000 00000>0 30000 .00 02020>m010<00020 000 0200000 00000000x0 00000> .mm 000000:00 00000000 .00 0000050000: .00 00. 0000002000: .00 00000000 .00 0000000000500 0000000 .00 000000 00000000 .00 000000 000:0000\0000000 .00 0m>m0 02020>0000< 00000 0000 000000 00000:0000 .00 00.- 0000000000 000000 .00 00. 000000 00 00000:0m .00 000000 -0>000 0:00003 00000 .00 00.- 00005500 00 0000000 .00 zu< 0o< 003000000 .00.0 00000 “J Lu H _1 mmo 110 00. mm. 00.- 00. 00. 00. 00. mm. 00. 00. 00. 00. 0mm 00: >m0 xz< 003 .a00 z<0 0 00.- 00.- 00. 00.- 00.- 00. NM» 00.- 00. 00. mm.- m0.- m0.- 00.- mm.- 00. .000 .000 000 000 000 000 0 003000000 .00.0 00000 0000000:0 0000\0000:: 00:000000 000000 00000 00000 0000000000000\00000 00005 0000 0000 00005 00000 0000 000000:00000 00000000 -0000: 0003000000 00000> 0003 05000000 0000000 000: 05000000 00.- 0>000:050 000000 00:00 0000 00000 0000:000000\00000000 00000000 00.- 000000000000000 000000 0: 00>00 00000 00000 00000500 0.00000 00030500 00000500 0.00000 0000000000 00000 0000 00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .00 .mm .00 .mm .00 111 00.:. 00. 00. 00. 00. 0000:0 00 5000000 00 00000>0m .00 00000000 00000 0000000 .00 00000 0 0000500\00005000 .00 00000000 000000:00 0000000 .00 00. 0000000 000500000 00000 .00 02000<020220000 0000000 05000000 000000 .00 0>0000000 .Nm 000000:00 00 0000000000000 000500 .00 00. 00. 00. 00. 00.- 00.- 05000000 000000050 .00 000 00: 000 020 :00 000 000 0 .000 .400 000 :00 000 000 0 00: 00:000000 «00.0 00000 112 scale elevation suggests that the presence of non-conforming, disrup- tive behaviors which generally motivate a teacher to refer a child for psychological evaluation are less likely to occur. F scale correlates suggest a low tolerance for frustration and a tendency toward physical aggression. This scale not only identifies unusual informant responses, but reflects the presence of various ex- ternalizing complaints. The DEF scale, like the Lie scale, was inversely related to indi- cations of disruptive behavior. Slight elevation suggests a decreased probabilityof difficulty concentrating and controlling anger toward other children. ADJ scale elevations reflect a wide range of acting-out behavior which motivates teachers to refer a child for further assessment. The scale correlates also suggest distractibility. ACH scale correlates reflect poor school performance in areas of reading and spelling. No significant school correlates were found which suggest reasons for this underachievement. For instance, Lachar and Gdowski (1979b) found that the ACH scale elevations were related to limited intellectual skills and the presence of a negative self con- cept which seemed to be associated with the lack of involvement in academically-related activities. The IS scale correlates reflect below average intellectual and academic functioning. Teachers described children who had elevations on this scale as disinterested in school activities and lack the per- severance for adequate performance. In addition to reflecting re- tarded development in the cognitive domain, correlates represent teachers' perceptions of marked lags in physical and social 113 development. Elevations on the scale make it unlikely that the child would be described as having poor self control, but s/he would be de- scribed as avoiding peer interaction. DVL, like the ACH and IS scales, clearly reflects below expectant academic functioning. Poor school performance was evidenced in the basic areas of achievement and perceived to be caused by delays in cognitive, perceptual, and emotional development. DVL was the only cognitive scale which was significantly related to difficulties in fine motor coordination and slow progress in physical education. The SOM scale was intended to identify children with excessive somatic concern. The lack of meaningful correlates may reflect the relative absence of TRSI items which describe somatic phenomena. Few correlates may also indicate that teachers are either more toler- ant of somatic symptoms or are less sensitive to their presence. The correlates that were present suggest the presence of more externalized than internalized symptomatology. D correlates suggest poor self concept and low tolerance for frus- tration. Few correlates may indicate that depressive symptomatology is less likely to be observed in school settings since the demonstra- tion of depression in childhood is often characterized by acting out behaviors. The D scale was, however, associated with teachers' judg- ments of the severity of problem behavior. Correlates for the FAM scale reflect problems conforming to limits. Poor school performance, lack of perseverance, aggressive and attention-seeking behaviors are likely with elevations on the FAM scale. 114 The DLQ correlates revealed a significant relationship between scale elevations and impulsive and aggressive behaviors. Elevations indicate an inability to conform to limits and tolerate frustration. The teacher's report of the severity of the problem was clearly as- sociated with the degree to which this scale was elevated. Beyond the social withdrawal which the WDL scale reflects, the scale correlates were suggestive of aggressive behaviors such as Steal- ing from other children. Teachers described the child with WDL scale elevations as passive, not protesting when hurt. The isolated nature of the correlation with items reflecting achievement in mathematics and handwriting made the interpretation unclear. ANX correlates reflect teachers' observations of a child's preoc- cupation and disturbed peer relations. There was an indication that children with ANX scale elevations tend to openly strike out against other children, despite their frequent passivity. Significant PSY correlates reflect problems in verbal expression and motorical coordination. Elevations were also related to preoccupa- tion and the reluctance to protest when hurt. This was the only PIC scale which was related to teacher recommendations that behavioral problems were severe enough to require a change in the child's place- ment. The HPR scale had the greatest number of correlates, indicative of its meaningfulness in reflecting school-related difficulties. HPR clearly suggests distractibility, impulsivity, restlessness, hostility, and social difficulties. Attention-seeking behaviors and academic un- derachievement were also related to HPR scale elevations. The HPR and ADJ scales shared a number of correlates; however, HPR correlates were 115 more indicative of a particular etiology for behavior problems, such as difficulty with control and overly active behavior. The SSK scale, like the DLQ, ANX, and HPR scales, reflected dis- turbed relationships and poor social skills. Scale elevations were suggestive of restlessness and problems in speech development. The correlates clearly reflect the severity of emotional problems. These results provide substantial support for the construct validity of the scales. The relationships among the TRSI items and the PIC profile scales were, in general, consistent with the interpretive intent of the scales. For example, the scales constructed to identify acting out behavior problems were significantly correlated with TRSI items which described aggressive and disruptive behaviors; whereas, the internalized scales correlated significantly with items which reflected internal discomfort. Similarly, the cognitive scales were related to TRSI items which identified academic achievement difficulties. CHAPTER V DISCUSSION Studies pertaining to the differential diagnosis of learning dis- abled, emotionally disturbed, and educable mentally retarded children have, in general, shown that discrimination among these special educa- tion categories is extremely difficult, if at all possible. Most of the studies have relied on measures which are administered directly to the child and/or are completed by professionals such as teachers and psychologists. An accumulating body of research has shown that parents are a valuable resource for obtaining comprehensive diagnostic informa- tion despite the fact that they are seldom consulted during the diag- nostic process. Previous research with the Personality Inventory for Children(PIC) has demonstrated that parents can relay diagnostically useful informa- tion about their children by responding to the individual items on the inventory. To date, most of the studies using the PIC have been lim- ited to psychiatric populations. This study examined the diagnostic potential of the PIC in school settings, specifically special educa- tion, and attempted to determine whether there was a meaningful dif- ference in the way the PIC profile scales represent problematic beha- vior of’ educationally handicapped children. The present study showed that through responses on the PIC, par- ents can provide reliable and valid information about children diag- nosed as learning disabled, emotionally disturbed, and educable 116 117 mentally retarded. The results further suggest that the PIC has the po- tential to be used as a screening device for identifying children in need of individual psychological testing and educational intervention. Differential Profiles Researchers have repeatedly shown that learning disabled and emo- tionally disturbed children are more difficult to differentiate, where— as educable mentally retarded can be more easily distinguished from the others. These contentions were examined in this study by planned com- parisons among the learning disabled and emotionally disturbed subjects and among the educable mentally retarded and a combined group of learn- ing disabled and emotionally disturbed subjects. The results showed that the PIC was equally effective in making these two comparisons; however, the particular profile scales which discriminate among the groups did vary. Emotionally Disturbed/Learning Disabled The PIC profile which best differentiated emotionally disturbed from learning disabled children was one which had significantly higher scores on the Adjustment, Delinquency, Anxiety, and Social Skills scales, and significantly lower scores on the Lie and Defensiveness scales. The Adjustment scale which reflects poor adjustment has been a valuable scale for identifying children in need of psychological evaluation. Elevated scores on the scale are characteristic of general clinic populations (Seat, l969); therefore, it is consistent with the definition of emotional disturbance that these subjects had signifi- cantly higher scores than children whose primary difficulty is learn- ing. 118 Although the Delinquency scale measures psychological adjustment, it represents to a greater extent delinquent tendencies. The items reflect a wide range of anti-social behaviors and a general disregard for the feelings of others and limits which are imposed. It is unlike- ly that many of the emotionally disturbed subjects are adjudicated de- linquents; therefore, the group's elevated scores on the scale are probably a function of their disruptive behaviors and problems with im- pulse control which are interpreted as a disregard for limits. Factor analysis of the PIC indicated that both the Anxiety and Social Skills scales represent internalized psychopathology (Lachar, 1975). Examination of the scales' content and the findings of Kline- dinst (1972, l975) which show that children seen in guidance clinics obtain significant elevations on these two scales suggest that an ex- ternalizing syndrome is being reflected. Since emotionally disturbed children more often exhibit acting-out behavior, these findings are consistent with the diagnostic classification. Elevated Social Skills scale scores reflect the social problems which emotionally disturbed children experience as a result of their inappropriate and often intrusive behaviors. The Anxiety scale ele- vation most likely reflects the sensitivity of these children to the frequent rejection by peers and reactions of adults. The learning disabled subjects' significantly higher scores on the Lie and Defensiveness scales indicate that the parents of these chil- dren, compared to parents of emotionally disturbed subjects, are more apt to deny commonly occurring behavior problems and respond to the inventory in a defensive manner. This finding may be related to some reasons why children are placed in learning disabilities programs. 119 Compared to the diagnosis of emotional disturbance, there seems to be less stigma attached to the learning disabled classification. It may be that many of the learning disabled subjects were placed in the pro- gram as a result of parental defensiveness and resistance to other pro- grams such as emotionally disturbed. Given this is the case, it is reasonable to assume that these parents will continue to maintain their defensiveness by ascribing more virtuous behaviors to their children. However, it is more likely that elevated Lie and Defensiveness scale scores are related to the actual low occurrence of problem behavior with learning disabled children. The actual T-scores are not high enough to be interpretable (Lachar & Gdowski, l979b) and the relatively flat PIC profile of the learning disabled children does suggest that fewer cognitive and emotional/behavioral problems are present. Educable Mentally RetardedLEmotionally Disturbed and Learning Disabled The PIC profile scales which best differentiated educable mentally retarded children from a combined group of learning disabled and emo- tionally disturbed were the Achievement, Intellectual Screening, Devel- opment, Withdrawal, Psychosis, Hyperactivity, and Lie scales. Signifi- cantly higher elevations on the cognitive scales (Achievement, Intel- lectual Screening, Development) suggest that educable retarded children achieve below age expectancy, have intellectual deficits and show de- velopmental delays. This is a particularly important finding in that it demonstrates the strength of these scales in reflecting the charac- teristics which define the educable retarded classification. The Withdrawal scale elevation may be a reflection of the re- sponse of retarded children to the rejection by peers. According to 120 Klinedinst(l972, l975),the scale represents uncommunicativeness, emo- tional distance, shyness and fear of strangers, as well as social and physical isolation. The discomfort and social withdrawal which the scale reflects may be a function of the retarded child's inability to develop appropriate social skills that would enhance peer interaction. Elevation on the Psychosis scale, like the Withdrawal scale, is most likely related to the intellectual limitations of the educable retarded child. The mean score of the educable group was much lower than that of children diagnosed as psychotic (Lachar, 197l); therefore, the difference between these subjects and those classified as emotion- ally disturbed and learning disabled suggests that educable retarded children have difficulty making judgments about reality due to their cognitive deficits. Since the literature has shown that educable re- tarded children often exhibit severe psychological problems (Hallahan & Kauffman, l976), it is reasonable to assume that scale elevations also reflect emotional problems. The educable mentally retarded subjects, compared to the combined group of learning disabled and emotionally disturbed, scored signifi- cantly lower on the Hyperactivity scale which suggests that they de- monstrate fewer behaviors which are characteristic of the hyperkinetic syndrome. Although certain groups of retarded children demonstrate overly active and distractible behaviors, such as those whose etiology is organic (Cromwell & Foshee, 1960), in general, hyperactivity is not characteristic of mental retardation. 0n the other hand, both learning disabled (Gaddes, l980) and emotionally disturbed (Quay & Werry, 1979) have been shown to exhibit behaviors which are characteristic of the hyperkinetic syndrome. These behaviors include attention deficits, 121 restlessness and problems inhibiting response to external stimuli. These findings are quite consistent with the intent of the Hyperac- tivity scale. Significantly higher scores on the Lie scale suggest that parents of educable retarded children, compared to parents of a combined group of learning disabled and emotionally disturbed children, are more like- ly to deny problems which commonly occur and describe their children as more virtuous. It seems that for most parents of retarded children, the initial acceptance of their children's intellectual limitations is extremely difficult. It may be even more difficult for these parents to acknowledge the presence of emotional and behavioral problems. The tendency to deny problem behaviors may be a function of a feeling that emotional problems are more a result of poor parenting than are intel- lectual weaknesses. It is quite reasonable to assume that guilt feel- ings may be preventing them from accepting the difficulty which their children experience in adjusting to daily demands. Sex and Age Differences The present study revealed few sex- and age-specific differences on the profile scales; however, the findings are generally consistent with the literature and provide support for the interpretive intent of the scales. For instance, the main effects for sex indicataithat.males are more often described as being poorly adjusted, a finding which is congruent with numerous studies which show that males are referred to guidance clinics more often and exhibit more maladaptive behaviors than females (Quay & Werry, 1979; Lapouse & Monk, 1958). Although this may not indicate that males experience greater psychological discomfort, it 122 does suggest that they are externalizing their problems. It is, there- fore, likely that the elevated Adjustment scale for males is reflecting the acting-out dimension of that particular scale. Females had significantly higher scores on the Development scale which indicates that they experience greater problems that are of a developmental nature. Since developmental delays, both cognitive and physical, often serve as the impetus for females' being referred for special educational intervention, this finding is not unexpected. Neither is it surprising to observe that females, l2 years and younger, had a significant elevation on the Achievement scale. Again this sex and age interaction most likely reflects the intent of the scale to identify poor school achievement since academic difficulties more often serve as the referral foci for children, particularly females. This finding supports that of an earlier study by Lachar and Gdowski (l979b) and further supports the ability of the scale to differentiate among children with learning problems. The only main effect for age was found on scales which identify parental defensiveness (Defensiveness) and poor social relationships (Social Skills). The results indicated that parents of adolescents are more apt to be defensive about their son or daughter's behavior; how- ever, they described their adolescents as having considerable problems caping with social situations. The finding that these parents were more defensive about their children may appear to be the opposite of that which is expected since adolescence is a time of considerable family stress as the adolescent strives to individuate and separate from parents. Elevations on the Defensiveness scale may be a function of the parents' anxiety about their children's continuing problems 123 which are requiring special education services. It is likely that par- ents become increasingly upset by their children's problems as they grow older; therefore, they tend to deny problems more than they would when their children were young. The finding that adolescents were described as having social prob- lems is consistent with their stages of psychosocial development. Since adolescence are seeking an identity, they are quite involved socially. This often results in conflict with peers, something which parents may interpret as poor social relationships. Sex and diagnostic group interactions indicated that female learn- ing disabled and male emotionally disturbed children and adolescents were more often described as having family-related problems. This is demonstrated by elevations on the Family Relations scale. Although it is reasonable to find that emotionally disturbed males have greater difficulty relating within the family, since males tend to exhibit more maladaptive behaviors, particularly disturbed males, it is somewhat surprising to find a similar report for learning disabled females. This finding may, however, suggest that a number of females, classified as learning disabled, are actually more emotionally disturbed. Again, with the stigma often attached to the category of emotional disturbance and the fact that females evidence problem beha- vior in less of an aggressive and acting-out manner, their learning disabilities' placements may be inappropriate to their primary needs. Parents of emotionally disturbed females were significantly more defensive about their daughters' behavior, as evidenced by elevations on the Defensiveness scale. This may reflect their feelings of guilt about their children's problems or may be a function of the actual 124 representation of their children's behavior. Since females tend to manifest their psychological discomfort in more of an internalized way and actually exhibit fewer problem behaviors (Werry & Quay, l97l), this finding is not unexpected. Another interesting finding was that parents of educable mentally retarded male children, 12 years and younger, tended to ascribe numerous problem behaviors to their children. This was shown by a significant elevation on the F scale. Although the scale is intended to identify parents who exaggerate their children's symptoms, the F scale can also reflect the relative intensity and severity of the symptomatology($eat, 197l). This seems to explain why parents of young, mentally retarded boys who demonstrate considerable cognitive and emotional problems (Hallahan & Kauffman, l976) appeared as if they were exaggerating be- havior problems. Cognitive Triad of PIC Although most of the findings across the scales were explained by sample characteristics, whether age, sex, or diagnostic grouping, the actual performance of the individual profile scales remained a question. Since the study is focused on the diagnostic potential of the PIC in school settings, it seemed reasonable to pay special atten- tion to the scales which purport to identifycognitive-related skills. These scales include the Achievement, Intellectual Screening, and De— velopment scales, and without exception each group's primary or secon- dary elevations were on one or more of these scales. There were differential degrees of elevation on the Achievement, Intellectual Screening, and Development scales across the three groups; 125 however, the groups' primary or secondary elevations were on these scales. This finding suggests that parents of learning disabled, emo- tionally disturbed, and educable mentally retarded all perceive their children as having considerable difficulty with cognitive tasks. This is particularly relevant since most children referred for special edu- cation services have experienced problems learning in regular education Not only does this provide support for the intent of this cognitive triad to assess cognitive-related problems, it also indicates how simi- lar parents and psychologists are in their perceptions of these mildly handicapped children. Perhaps the most notable finding, however, was the consistent peak elevation of the Intellectual Screening scale across all three groups. The educable mentally retarded group's mean scale score was signifi- cantly higher than that of the learning disabled and emotionally dis- turbed subjects. According to the PIC interpretive guidelines (Lachar & Gdowski, l979b), however, the scores of all three groups were suf- ficiently high to raise concern about possible limited intellectual en- dowment. One explanation has to do with the item content of the scale. Since the Intellectual Screening scale items reflect an inability to retain and apply basic skills, the high score of the emotionally dis- turbed subjects may be a function of the emotional disturbance which has markedly interfered with their ability to attend and learn basic skills. The learning disabled group's score may, on the other hand, reflect some type of neurological involvement which causes the child problems attending and processing information. The fact that parents of emotionally disturbed and learning dis- abled described their children in a similar way as parents of children 126 with mild intellectual deficits may suggest that parents of these sub- jects share a common misperception about special education. Special education has been viewed by many to be synonymous with mental retar- dation; therefore, it is not surprising that these parents may hold such views. Besides, it was assumed that the parents of the subjects have very limited information about the differential characteristics of their children's categories. Since both the emotionally disturbed and learning disabled groups' mean IQ scores, as measured by the Wechsler Intelligence Scale for Children--Revised, fell well within the average range, this raises a question pertaining to the validity of the scale. However, an examina- tion of the construct validity of the Intellectual Screening scale show substantial support for the scale's validity. (The convergent and di- criminant validation of this, as well as other PIC profile scales, is discussed in more detail later in the chapter.) It may be that the cut-off score which is presented in the interpretive manual is too low; however, further study is necessary to determine this. The groups' performances on the remaining two cognitive scales (Achievement, Development) warrant further explanation. Again, there were differential levels of elevation, with the educable mentally re- tarded group's scoring significantly higher than the others on both scales. However, beyond significant score differences among the groups, it is important to note the relationship of the cognitive scales to each other and the remainder of the profile. The finding that the Achieve- ment scale score for the educable retarded subjects was markedly ex- ceeded by the Intellectual Screening scale score suggests that academic achievement problems are a result of intellectual deficiencies. 0n the 127 other hand, the fact that emotionally disturbed subjects had much high- er Adjustment scale scores than Achievement scores suggests that the poor achievement of these children is a result of emotional problems. The relationship of the Achievement scale to the Intellectual Screen- ing scale of these two groups reflects the definitions of the diagnos- tic categories and provides support for the constructs which these scales were designed to measure. The finding that the Development scale score of the educable re- tarded and emotionally disturbed groups was relatively high compared to other profile scales suggests that these subjects show developmental delays. This is consistent with research which demonstrates that re- tarded (Hallahan & Kauffman, l976) and emotionally disturbed children (Lucas, Rodin, & Simson, l965) exhibit a number of problem behaviors which are developmental in nature. The Achievement and Development scales of the learning disabled subjects are higher than the other profile scales, except Intellectual Screening. The Development scale reflects academic potential as well as attention problems and coordination difficulty. Items such as "my child seems more clumsy than other children his (her) age" and "read- ing has been a problem for my child" which comprise the scale often serve as indicators for the presence of a learning disability. It is, therefore, not unexpected that learning disabled subjects would have high scores on this scale. The finding that the Intellectual Screening scale exceeds these scales suggests that intellectual deficits are causing academic diffi- culties. This is inconsistent with the results of Johnson (l97l) and Shove (1972) which show that learning disabled children have their peak 12R elevations on the Achievement scale, followed by secondary elevations on the Intellectual Screening scale. These discrepant results may be interpreted as a function of the sample's characteristics; however, this raises a question about the performance or validity of the scales. Construct Validation of PIC The construct validity of the PIC profile scales was investigated by convergent and discriminant validation methods. The two indepen- dently obtained measures which were used to evaluate the scales were selected because of their relevance to academically-related behaviors. Whereas the Teacher Rating and School Information (TRSI) checklist was chosen specifically to examine scales which purport to assess problem behaviors and academic achievement, the Wechsler Intelligence Scale for Children--Revised (WISC-R) was selected as a means to study scales which were constructed to identify cognitive deficits. WISC-R as Criterion The findings clearly show that the majority of significant corre- lations cluster around four of the scales which discriminated between the educable mentally retarded subjects and a combined group of learn- ing disabled and emotionally disturbed subjects. These scales, as seen in Table 4.2, are the Achievement, Intellectual Screening, Development, and Psychosis. Although all four scales were significantly correlated with the WISC-R Verbal, Performance, and Full Scale 105, the Achieve- ment and Psychosis scales have fewer subtest correlates. Correlates obtained for the Intellectual Screening scale were con- sistent with previous studies which examined the relationship among the scale and WISC-R Full Scale IQ (Wirt et al., 1977) and WISC-R subtests 129 and IQs (Culbert & Gdowski, l982). Elevated scores on the Intellectual Screening scale were related to indications of intellectual limitations in verbal and performance abilities. These results provided support for the construct of intelligence which the scale purports to measure.. The Development scale was also related to subtests which measure both verbal abilities and performance skills. Since the Development scale was designed to assess intellectual, as well as physical, devel-' opment, this, too, is a significant finding in support of the validity of the scale. The fewer significant relationships between the Achievement scale and the WISC-R seem to be a function of the scale's intent to measure academic achievement, regardless of intellectual abilities (Wirt etal., l977). The only significant correlations were with WISC-R scales mea- suring verbal skills. This may be a result of the importance of verbal abilities in academic achievement. The relationship between the Psychosis scale and the WISC-R may be interpreted as a function of the scale's intent to identify poor real- ity testing. Since reality testing is to a certain extent a cognitive task, the significant correlation is not unexpected. It also explains why the scale was related to the Comprehension subtest of the WISC-R which assesses the child's ability to make practical judgments. In ad- dition to identifying cognitive confusion, the Psychosis scale also predicts internalized psychopathology (Lachar, l975). This most likely explains why the Psychosis scale was significantly correlated with WISC-R scales measuring attention and concentration, as well as spatial skills. Elevated scores on the scale suggest that an individual who 130 exhibits psychotic symptomatology has difficulty attending to external stimuli. The Social Skills scale had two significant WISC-R correlates, one with a subtest measuring comprehension and the other attention and sus- tained effort on a psychomotor task. The relationship between the PIC scale and the Comprehension subtest can be explained easily since both purport to assess the ability to make practical social judgments. Sim- ilarly, it is reasonable to find a relationship between the Coding sub- test which reflects freedom from distractability (Kauffman, l975), since children with social difficulties are often insensitive to ex- ternal cues. These significant relationships provided substantial support for the convergent validation of the Intellectual Screening and Development scales which obtained the most correlates. The fact that these scales were designed to measure intellectual development and functioning is extremely important. With the exception of the few significant cor- relations among the WISC-R and the Psychosis and Social Skills scales, there were no significant relationships between the WISC-R and the PIC scales which measure behaviors other than cognitive. In addition to noting the relative lack of significant correlations with non-cognitive scales, it is important to observe how extremely low the non-signifi- cant correlations were (Table 4.8). These findings provided substan- tial support for discriminant validation of the PIC cognitive scales. Further information about the validity of the PIC profile scales was provided by examining the correlations between the PIC and the Teacher Rating and School Information (TRSI) checklist. 131 TRSI as Criterion Examination of the relationships between the PIC profile scales and the TRSI provide considerable support for the construct validity of the PIC. With the exception of the Somatic Concern, Depression, and Anxiety scales, the PIC profile scales had a number of significant TRSI correlates. This was particularly true of the Adjustment and Hyperac- tivity scales, both of which significantly discriminated among the special education categories. As seen in Table 4.2, the emotionally disturbed, compared to the learning disabled, had significantly higher Adjustment scale scores.and the combined group of emotionally disturbed and learning disabled had a significantly higher elevation on the Hyperactivity scale when compared to educable retarded subjects. The finding demonstrates the importance of these two scales in representing problem behavior in school. Scales designed to identify the denial of commonly occurring prob- lem behaviors, the Lie and Defensiveness scales, were consistently re- lated to the absence of symptomatology. The F and Adjustment scales which are associated with indications of marked emotional and cognitive difficulties were significantly related to indications of maladaptive behavior in the classroom. These correlates for the three validity and one general adjustment screening scale were consistent both with the purposes for which the scales were designed and the findings of pre- vious research (Lachar & Gdowski, l979b). The three cognitive scales (Achievement, Intellectual Screening, Development) were significantly correlated with TRSI items which re- flect below average progress in the basic areas of school learning. Support for the convergent validation of the Intellectual Screening and 132 Development scales was found by examining the relationships among these scales and teacher reports of the reasons for a child's poor academic achievement. Teachers attributed academic difficulties to below aver- age intelligence and delays in physical development as evidenced by fine motor coordination problems. The scales' negative correlation with items which reflect emotional and motivational reasons for achievement's being below age expectancy provided further support for the construct validity of the two scales. This suggests that certain learning problems are attributed to cognitive and developmental delays. The absence of such correlates for the Achievement scale provided sup- portive data for discriminant validation of the scale. Specific scale correlates showed that the Intellectual Screening scale represents both academic and social difficulties. Elevations on the scale were related to teacher descriptions of socially withdrawn behavior. This relationship may be a function of the child's being rejected by peers as a result of the poor development of social skills caused by retarded intellectual functioning. Poor achievement in read- ing, math, verbal expression, and language and phonetic skills was at- tributed to below average intelligence, not a lack of concentration and perseverance. In addition to being correlated with the basic areas of achieve- ment that the Intellectual Screening scale was correlated with, the Development scale was significantly related to slow progress in spell- ing and physical education. Since the scale purports to identify de- lays in both intellectual and physical development, this is a very im- portant finding in support of the scale's validity. Examination of the external correlates of the TRSI indicated that poor achievement is a 133 result of problem behaviors which are generally recognized as develop- mental in nature. These include poor coordination and intellectual deficits. The Achievement scale had the fewest TRSI correlates. The scale was solely related to difficulties in reading, spelling, and phonetic skills. Although the correlates were reasonable in terms of the intent of the scale to measure achievement in school-related subject areas, there was minimal support for the scale's validity. Since Lachar and Gdowski (l979b) found considerable support for the scale using the TRSI and other independently obtained measures, the present results may be a function of the sample characteristics. Further research which ex- amines the relationship between the Achievement scale and measures of achievement such as a standardized achievement test would be beneficial Family Relations scale correlates reflected the child's need for adult supervision. Teachers described children with elevated scale scores as aggressive, defiant, and non-conforming. Although the TRSI was not designed to identify family-related problems, these correlates did reflect the scale's intent to assess a child's ability to cooperate. The scale was also significantly related to acting-out behavior (Kline- dinst, l972, l975); therefore, the correlation with the TRSI items is not unexpected. The correlates which attributed poor academic perfor- mance to high absenteeism, lack of perseverance,and emotional diffi- culties suggest that behaviors which interfere with school achievement may be caused by family-related difficulties. The Delinquency and Hyperactivity scales were correlated with items which suggest that these scales reflect an inability to conform to school limits. These scales were among the few that were 134 significantly related to the severity of the child's problems. Chil- dren with elevations on these scales were described by teachers as dis- turbing to other children because they tease and provoke fights, have a low frustration tolerance, and fail to conform to externally imposed limits. Although neither scale was significantly correlated with achievement items on the TRSI, teachers indicated that below average school achievement was caused by impulsivity, poor study skills, lack of sustained effort, and emotional difficulties. These correlates were quite consistent with the Delinquency and Hyperactivity scales' purpose to identify externalizing behavior. Both scales assess difficulty controlling acting-out behavior. Correlation with the teacher's judgment of the severity of the problem reflected the relevance of these scales in describing mildly handicapped children in classrooms. In addition to the correlates that the scales share, the Delin- quency scale is related to poor class attendance. The Hyperactivity scale also reflected overly active, distractible, and impulsive behav- iors. High absenteeism is consistent with delinquent-type behaviors, just as poor impulse control and restlessness are characteristic of the hyperkinetic syndrome (Quay & Werry, l979). The TRSI correlates with the Withdrawal scale reflected social isolation and apathy. Teachers indicated that children with high scores on this scale also have problems with stealing. Whereas the re- lationship between the scale and the items which represent withdrawal were expected, correlation with teacher reports of stealing was not. Acting-out, aggressive behaviors are often a result of social rejection However, these are seldom related to withdrawal and isolation. 135 Stealing, on the other hand, may be a more passive-aggressive response to feelings of peer rejection; therefore, a relationship between the Withdrawal scale and items indicating that stealing is occurring is more consistent than it appears. Elevations on the Psychosis scale reflected problems in motorical coordination and verbal expression, apathy, and preoccupation. This scale was also related to teacher recommendations for grade placement changes, suggesting that present educational programming is not suffi- cient. According to Lachar's (l975) factor analysis of the PIC, the Psychosis scale loads on the internalization factor. Since the problem behaviors which were reflected by the TRSI were also indicative of in- ternalized psychopathology, these findings were relatively consistent with the intent of the scale. Social Skills scale correlates suggested that the scale represents problems in verbal expression, restlessness, and disruptive behavior. The scale was also related to social problems ranging from social with- drawal to aggression toward peers. TRSI correlates indicated that teachers described children with high scores on the Social Skills scale as having severe emotional problems. The scale is comprised of items which measure both the lack of success in social relationships and the reasons for the situation. It is most likely that the TRSI correlates are those which suggest reasons for the lack of success socially. Poor communication and aggressive behaviors are certainly deterrants in forming social relationships. The teacher's report of severe emotional difficulty may be viewed as either a cause of the social difficulty or a result of such. 136 The lack of support for the Somatic Concern scale may be due to the relative absence of TRSI items which reflected these problems. It may also indicate that teachers' information is not sufficient to enable them to respond to items dealing with somatic difficulties or the teachers are more tolerant of this type of complaint from the stu- dent. The few number of Depression scale correlates may be a result of the fact that while depressed children display overt depressive symp- toms for definite periods, these children frequently defend against their unhappiness by aggressive, acting-out behaviors (Cytryn & McKnew, 1972). The finding of few TRSI correlates may be due to the fact that the Depression scale primarily represents depressed behaviors which are internalized rather than externalized.and the TRSI has few items which reflect internalized psychopathology. A similar situation may be true of the Anxiety scale since children often manifest anxiety by restless, distractible, and even aggressive behaviors (Silver, 1979). The An- xiety scale items, however, focus more on internalized symptomatology such as excessive worry and fear. These findings regarding the relationship of the PIC profile scales to the TRSI are, in general, consistent with similar research using psychiatric populations (Lachar & Gdowski, 1979b). The rela- tionships provided substantial support for the construct validity of the majority of profile scales. Where minimal or no support was found, further research with these scales in school populations, specifically educationally handicapped, is advised. 137 Diagnostic Validation The present study provided answers to a number of questions re- garding the Personality Inventory for Children. For instance, the findings indicated that the PIC is a highly reliable and valid instru- ment that can be used to describe cognitive and personality character- istics of children diagnosed as learning disabled, emotionally dis- turbed, and educable mentally retarded. The results also showed that, compared to the use of multiple measures, the PIC was nearly as effec- tive in differentially diagnosing these children. For a number of years, teacher-completed behavior checklists and standardized intelligence tests have been a part of the armamentarium to diagnose educationally handicapped children. The degree to which the PIC contributed to the differential diagnosis of these children was, therefore, judged in relation to the Walker factors which were ob- tained from the Walker items extracted from the Teacher Rating and School Information checklist and the Wechsler Intelligence Scale for Children--Revised. ' The discriminant function analyses which were used to assess the relative abilities of these measures, along with variables of sex and age, to discriminate among the three groups of children showed that the groups can be separated along two salient dimensions, one intellectual and the other emotional. An interesting finding was that across all the discriminant function analyses, the factor which represented intel- lectual functioning was the most powerful in separating the groups. This is particularly relevant since cognitive deficits are most often the impetus for special education referral and are generally the major focus of the diagnostic process. The importance of using measures of 138 intellectual ability was demonstrated not only for the educable re- tarded group which is defined by below average intelligence, but for learning disabled and emotionally disturbed as well where intellectual functioning is on the average below the population mean (Kirk & Elkin, 1975; Hallahan & Kauffman, 1976). The PIC, unlike the WISC-R and TRSI, has the ability to assess both intellectual and emotional features of the child. When compared to the WISC-R, the PIC was not as effective in identifying educable mentally retarded children; however, the PIC was clearly more powerful in classifying emotionally disturbed subjects. This may be a function of the Adjustment, Psychosis, Anxiety, Depression, Social Skills, Hyperactivity, and Withdrawal scales which describe characteristics of emotionally disturbed children that the WISC-R is unable to assess. Learning disabled children were also classified more accurately when the PIC rather than the WISC-R was used. However, the improvement in predictive accuracy was not as dramatic as that observed for emo- tionally disturbed children. The improvement may be due to the PIC scales such as Achievement, Hyperactivity, and Social Skills which assess characteristics generally associated with learning disabled children. The behaviors which the scales reflect included poor impulse control, restlessness, concentration problems, difficulties in school relationships, and poor academic achievement. The discriminant function which best classified the total sample was one which used both the PIC and WISC-R. When the Walker factors were added to the analysis, there was no gain in overall predictive accuracy; however, the greatest number of emotionally disturbed were correctly classified.. When an analysis with multiple measures 139 was used, the following variables contributed to the discrimination among the groups: WISC-R IQ, PIC Adjustment, Intellectual Screening, Development, Psychosis, Anxeity, Depression, Walker Acting-out/Dis- turbed Peer Relations, age, PIC Social Skills, Walker Distractibility, and PIC Hyperactivity. These variables are listed in decreasing order of their contribution to group separation. Regarding the PIC scales, with the exception of the Depression scale, the Adjustment, Intellectual Screening, Development, Psychosis, Anxiety, Social Skills, and Hyperactivity scales were the ones which were significantly different when analyses of variance were performed. As seen in the planned comparisons, the Adjustment, Anxiety, and Social Skills scales significantly differentiated the profiles of emotionally disturbed from learning disabled, whereas the Intellectual Screening, Development, Psychosis, and Hyperactivity scales were significantly different when educable retarded subjects were compared to combined learning disabled/emotionally disturbed. The results of the discrimi- nant function provided further support for the diagnostic potential of these scales to differentiate characteristics of educationally han- dicapped children. The finding that the Depression scale contributed significantly to group differentiation in the discriminant function analysis, but not in theanalysis of variance, may be due to the way the planned comparisons were arranged. For instance, the emotionally disturbed subjects were either grouped with or compared to learning disabled. Since no further comparisons were made for the emotionally disturbed subjects, the Depression scale effects which would be ex- pected for that group may have been masked by the choice of compari- sons in the planned tests. 140 The Walker factors were obtained when the Walker Behavior Check- list items from the TRSI were submitted to a factor analysis. The fac- tors which contributed most to group differentiation were the Acting- out/Disturbed Peer Relations and Distractibility factors. This sug- gests that these behaviors are relevant in discriminating among educa- tionally handicapped, a finding which is not unexpected since both in- attentiveness and disruptive behaviors frequently serve as concerns for. special education programming. The WISC-R IQ was particularly powerful in separating the subjects into their respective groups. This not only indicates how relevant in- tellectual functioning is in characterizing the groups, but also suggests that the WISC-R measured a somewhat different aspect of in- tellectual functioning than the PIC Intellectual Screening scale did. This is very important since the administration of the Intellectual Screening scale of the PIC is not intended to be used as a substitute for administering an individual intelligence test (Wirt et al., 1977). Although the sex of the child did not contribute to group differ- entiation, age did. This is somewhat unexpected since very few age- specific differences were found in the analyses of variance; however, the finding is not inconsistent with the literature (Werry & Quay, 197l) and the findings of previous PIC research (Lachar & Gdowski, 1979b) which show age effects and may be a result of the heterogeneity of the groups. The overall results of the discriminant function analyses indi- cated that there was strong agreement between parents' and psycholo- gists' perceptions of children with problem behaviors and that the PIC 141 was an effective instrument, along with other measures to predict diag- nostic classifications. Although cross-validation is necessary to de- termine whether the variables would have the same weight if used to predict group membership for another sample, this was not possible in the present study because of the small sample of educable mentally re- tarded children. CHAPTER VI SUMMARY AND CONCLUSIONS No later than 25 days after a child is referred for special educa- tion services, an Educational Planning and Placement Committee must convene and make recommendations based upon appropriate diagnostic information (Education for All Handicapped Children Act, 1975). Fur- thermore, the law states that the sources to be considered in the eval- uative process include parents, as well as teachers and psychologists. Not only are there few psychometrically-sound measures for ob- taining information from parents regarding a child's personality and cognitive abilities, but most methods are quite inefficient. A number of researchers have indicated that the Personality Inventory for Chil- dren (PIC) offers an opportunity to assess, via parent responses, both personality and academically-related abilities with increased validity and efficiency over currently used methods. If this is the case, the PIC may be a vital instrument for evaluating children who are referred for special education. The present study sought to determine whether the PIC could dif- ferentiate among l4l children (ages 6 to l6) identified as learning disabled, emotionally disturbed, and educable mentally retarded, ac- cording to profile scale patterns. Second, the study investigated the construct validity of the PIC by examining the relationships among the PIC profile scales and independently derived measures of cognitive and 142 143 emotional/behavioral adjustment. Since the studies to date have pri- marily been limited to psychiatric populations, this study examined the utility of the inventory in school settings, specifically, special edu- cation. Findings The results of this study showed that the PIC is a highly reliable instrument which has the potential to assist in the diagnosis of educa- tionally handicapped children and adolescents. The learning disabled, emotionally disturbed, and educable mentally retarded children had significantly (p<:.000l) different PIC profiles; however, consistent with the literature, the three groups shared a number of characteris- tics. The educable mentally retarded group compared to a group of learn- ing disabled and emotionally disturbed received significantly (p‘<.05) higher elevations on the Lie, Intellectual Screening, Achievement, De- velopment, Withdrawal, and Psychosis scales and significantly lower scores on the Hyperactivity scale. Emotionally disturbed children, compared to learning disabled children, obtained significantly lower scores on the Lie and Defensiveness scales and significantly higher scores on the Adjustment, Delinquency, Anxiety, and Social Skills scales. These findings indicate that the three special education groups are differentiated along relevant cognitive and personality di- mensions, as represented by the PIC. Without exception, the groups' primary or secondary elevations were on PIC scales which measure cognitive abilities (Achievement, In- tellectual Screening, Development); however, it was the degree of 144 elevation on these scales and the relationship to the remaining profile which defined each diagnostic group. The educable mentally retarded children had peak elevation on the Intellectual Screening scale with secondary elevations on the Psychosis, Adjustment, and Social Skills scales. This suggests that these children experience emotional, as well as cognitive difficulties. The adjustment problems which retarded children encounter, however, appear to be more a function of their in- tellectual limitations rather than a result of a primary emotional disturbance. Since reality testing and social skills are to a great extent cognitive tasks, it was not surprising to find that educable retarded children had elevations on the Psychosis and Social Skills scales. Although the emotionally disturbed children showed high scores on the cognitive scales, particularly the Intellectual Screening scale, these elevations were secondary to the elevations on the Adjustment scale. Thus, in the case of the emotionally disturbed child. Academic difficulties may be explained in terms of emotional problems rather than intellectual deficits. The emotionally disturbed child had rela- tively high scores on the Depression, Delinquency, Anxiety, and Social Skills scales which suggest that the PIC is able to detect both in- ternalizing and externalizing symptoms of emotionally disturbed chil- dren. With the exception of elevated cognitive scale scores, the PIC profile of the learning disabled group was relatively flat. Compared to the emotionally disturbed and educable mentally retarded groups of children, the profile of the learning disabled group was least sug- gestive of psychopathology. These findings across all groups are 145 consistent with the definitions of the categories as well as the liter- ature which describes classifications(rfembtionally disturbed, learning disabled, and educable mentally retarded children. Few main effects for age and sex were found in the study. The findings showed that adolescents, l3 years and older, are more often described as having difficulties in social relationships and that par- ents of this age group tended to be more defensive about their son's or daughter's behavior. The results also revealed that males are per- ceived by parents as being more psychologically maladjusted, whereas females are viewed as developmentally immature. These findings were not unexpected since males are more often referred to special education for behavior problems and females for academic problems or developmen- tal delays. A sex and age interaction indicated that female children, aged 12 and younger, are described more often as underachievers. Interactions between diagnosis and sex showed that learning dis- abled females and emotionally disturbed males have greater family- related difficulties. Interestingly, parents of emotionally disturbed females were more often defensive about their children's behavior. This was interpreted as a function of the more internalized manner in which females evidence problem behavior and may also be a result of the small number of females in programs for emotional disturbance. On the other hand, a three-way interaction showed that parents of educable mentally retarded male children, 12 years and younger, tended to de- scribe their children as having substantial problems. Although this may indicate an exaggeration of symptoms, the F scale elevation may also reflect the actual presence of problem behavior. 146 The results of the discriminant function analyses which further examined the diagnostic utility of the PIC suggest that the PIC is a very effective instrument for classification. The major advantage for using the PIC is that the PIC profile scales assess cognitive abilities and emotional adjustment, both of which are considered relevant for discriminating among mildly handicapped children. As the discriminant function analyses showed, there are two salient dimensions along which these groups can be separated, cognitive and emotional. Although the WISC-R was best able to assign children to the edu- cable mentally retarded group, compared to the PIC, it was least ef- fective in classifying emotionally disturbed and learning disabled. The PIC, though not as effective in predicting group membership for educable retarded children, when used in conjunction with the WISC-R, correctly classified the highest percentage of total subjects. Al- though the additional use of the Walker factors from the Teacher Rating and School Information form (TRSI) did not improve the overall predic- tive accuracy, two factors derived from the checklist items-Acting Out and Distractibility, were beneficial in assigning emotionally disturbed children to their respective groups. This finding was not surprising since acting out problems and distractibility are often characteristic of emotionally disturbed children. Cross-validation is needed prior to applying the discriminant equations to actual cases of referred children; however, the analyses clearly demonstrated that the degree of group differentiation can be improved by increasing the number of relevant variables. Since the factor which contributed most significantly to the differentiation of educationally handicapped children was a cognitive factor, the PIC 147 scales which presumably measure cognitive functioning (Achievement, Intellectual Screening, Development) were examined for their construct validity. Convergent and discriminant validation of the cognitive scales were sought by examining the correlations between the PIC profile scales and a standardized intelligence test, the Wechsler Intelligence Scale for Children--Revised (WISC-R). The results showed that the majority of significant (p( .OOl) correlations clustered around PIC scales which discriminated between educable mentally retarded children and a combined group of emotionally disturbed and learning disabled children. These scales included Achievement, Intellectual Screening, Development, and Psychosis. The only other significant relationships were among the PIC Social Skills scale and the Comprehension and Coding subtests of the WISC-R. Significant correlations among the Intellectual Screening and De- velopment scales of the PIC and all WISC-R subtests and IQs provided support for the construct which the scales are to measure, intellectual ability. Fewer significant relationships were found between WISC-R data and the Achievement scale. This is consistent with the intent of the Achievement scale to measure academic progress, regardless of in- tellectual abilities. The Achievement scale's correlation with WISC-R subtests measuring verbal comprehension is also congruent with the generally recognized importance of verbal skills for school achievement. The significant correlations among the Psychosis scale and the WISC-R 10s and subtests which measure spatial organization are likewise consistent with the scale's intent to identify poor reality testing, a perceptual and organizational as well as cognitive task. There were 148 significant relationships among WISC-R subtests measuring attention and concentration and the Psychosis and Social Skills scales, both of which are considered to identify internalized psychopathology. This suggests that children who are particularly sensitive to internal cues are also likely to have difficulty attending to external stimuli for any length of time. The significant correlation between the Comprehension subtest of the intelligence scale and the Social Skills scale of the PIC indi- cates that children with difficulty in social interactions are also weak in practical-social judgments. This was an expected finding in light of the scale's intent to assess social judgments as well as rela- tionships. The construct validity of the PIC profile scales was further studied by correlating the inventory with an independently derived mea- sure of problem behaviors, the Teacher Rating and School Information (TRSI) checklist. The TRSI, which was also found to be a highly re- liable measure, was chosen because of its relevance to school behavior. The scales designed to identify a tendency of the informant to deny common childhood problems (Lie and Defensiveness) were the only scales which were consistently related to the absence of symptoms. The scales which measured the exaggeration of symptomatology and general maladjustment (F and Adjustment) obtained correlates which reflect the varied symptomatology to which these scales are sensitive. With the exception of the Somatic Concern, Depression, and Anxiety scales, the PIC clinical scales received considerable correlate support The lack of significant external correlates for the Somatic Concern scale may be due to the absence of TRSI items which reflect these problems or the teacher's unawareness of or high tolerance for somatic 149 complaints, rather than the poor performance of the scale. The lack of significant Depression scale correlates may bea function of the scale's failure to take into account aggressive behaviors which are character- istic of childhood depression, as well as "traditional" depressive symptoms such as sadness, crying, loss of appetite, and sleeping dis- orders. The Achievement, Intellectual Screening, and Development scales reflected progress below age expectancy in the basic areas of school achievement. The Intellectual Screening and Development scales were related to reasons for underachievement, suggesting that limited cog- nitive functioning and delayed development cause learning problems. The Delinquency and Hyperactivity scales reflected an inability to conform to external limits. These scales had a number of corre- lates in common which reflect low frustration tolerance, disruptive behaviors, and impulsivity. Family Relations scale correlates re- flected the child's need for externally imposed limits and adult su- pervision. specifically, the scale reflected defiance, non-conformity, and aggressive behavior toward peers. Correlates of the Withdrawal, Psychosis, and Social Skills scales indicated that these scales measure dimensions of internalized psycho- pathology such as apathy, problems with verbal expression, social with- drawal, and isolation. In addition, the Psychosis scale reflected coordination problems and preoccupation, while the Social Skills scale indicated restlessness and disruptive behaviors. Interestingly, the scales which correlated significantly with the judged severity of the child's problems were the Adjustment, Depression, Delinquency, and Social Skills scales. With the exception of the Depression scale, 150 these scales were the ones which discriminated between the emotionally disturbed and the learning disabled children. The significant (p000 0000000> 000005000000 000 000000>000 000 00000 -0>0 .00000 0000 000 00000000 0000000000 00000000000 0:0005:z .0000>000 0000000> 000000>000 000 00000 -0>0 .0000000000 0.00 00 00000>0 00 000000 -00 00000 000 0000 0000 ”005000 0 0 0>0 .z<0 .om< H00 00 .om< z<0 .0>0 .xz< .003 "00 0>0 "a: "000000>000 00000 000 00000000000 .00000000 0000:0000 00 00:000 0:000000500 000>000000 0000000000000 00 0000 003 000 05000500 00 00000000 00000> 00 0>00 -00000 003 000 003000 000:000 .000000000 000>0000 5000000 0 5000 000000 000000 0003 000000 00000 000 0000000000 .05000 0000 -00000 000>0000 5000000 0003 000000 00000 000 0000000000 .0000000000 000>0000 .000 000 000:0000 00000000> .0<>0z< 00 000050x0 0003 00000000 00 00:000 0003000 00000000000 00000000000 . 000000000 000>0000 500000 0 00 00000000 000 -000 5000 00>0000 00000:000 00000000 000000 00 00:000 0:000000500 00 00000:00 0003 000000 000 0000000000 00:000000 .00000000 00000 000 00003 00 0000:0000000 00>0 .0000» 00-0 00000 00< .0000 -0:00>0 00000000000 000 00000000 00000500 000 .00005 0000 00000000 000 .0.0 u 000 00000>< .00000:00>0 00000000000 000 00000000 00000500 00 .00005 000 00000000 00 .000 00 0000» 00 000 0 0003000 .005000 000 000 000000 .0000 00>00000 0000030030 .0: .00 .00 00 0000000000 00000000 000 .00.00 -0.0 00000 000 .0000 -0:00>0 00000000000 000 00000000 00000500 000 .00005 0000 00000000 000 000500 000 0:0 00 >00000<> 0:0 wz0z0zm00050 000000 0003 0000000000 000>000000 00 000 =0< 00 000000>000 000050>000 -50 0500 0003 0000000 -000 000>000000 xz< .003 .0>0 .0 00 000000>000 m0o0>0000 000000 000 0500 00>00050 00 0000000 000000>000 00: 0000000 .00000 0000000 000 0000 000 0000000000 00000000000 0.0 00000> -00 x00005 00000000 -000 .00000 005000 000 5000 00000000 00 000000 00000000000 000 000000 -00000000 000000 000 000 .000000 00: .000 .0 00 0000000x0 0002 .0000000505 00000 00 00000 00 000 0:0 00000000000 0003 00000 -00 000 .000000 00 000 00 mm 00 000000 0003000 00000000000 00000000000 .0000000 00000 0000000 000 0000 0000 0000 000 00:00 000000 000000000 00050 00000 00000000000000 00 00000000000 00000000000 .0000000 0:00 00 00000000 0000000 00000 000 000 00 000500000 00 000000 000 0003 0000000 -000 0003 000500000 0000 000300000 000>0000 00>000 -50 00 0000000 .000000 .0000000000 000000x0 000500000 00 000000000 000>0000 0000000 0003 0000000000 0000 000000 000 .000000 00000 000 00 000 -000000 0000000505 00000 0000000 00 000: 00000000 00000000 000005000000 .000000 000005000000 000 0000000 000 00 000 -000000 000000 000 00030 -00 00000000000 0005000 -00 00 000030000 <>0z< .000000 000000000 000 0003000 000000050000 -00 000500000 00 005000 -000 000000000 000000-000 .00000000 000000 000 00000 0000 00 0000000000 0000 .000500000 0000 000005000 000 00000550000 00000000 00 .00.0 00030 -00 000000 0000 .0500 -0000 00000000 00x000 00000>0000 0000000000 00 0000000000 00000000 00 000 00000000 005000 0N .00000.»0-m~ 00030 -00 0000 .05000000 000000050 000 0000: -00>0 00000 00000000 00 .00000000 005000 000 .0000 -00000000 00000 000 0000 00:000 000000000 000 00 x00 000 000500 0>0005000 5000 000000 00-00 00000000 000.~ 000000 000000> 000000 005000 000000 000000 000000 0002000 0:0 000000 .000000 158 0000 .0>0 .000 .000 000000 000 000 0000000 0-0003 00030 -00 000000000000 0000 -0000000 .000 .00 .000 .00< 00 000000 00000 000000 0000000000000 000 00000 00000000 0000000 .00000 05000000 000000 000 000 .0z<00 00000000 000500 000 .0=omv 0000003 000 00000 -000 00000000 000 .0x2< .0030 05000500 00000000 -00000 000 .000: .0000 000000 000 00 00000 -0000050 0>0000050 000 0000000000 000 000000 -000 0000000000 0022 .000000 0.000: 0003 000000 00000 0000000 000 0000000000 .0000 .000000 030 000 00 00000000 0000000 .000000 00000 000 0003 0000000000 0003 00022 00 00000 0000000 0000 .00.0 n 000 My 00005 005000 N0 .00.0 n 000.00 00095 00000000 0000000 N0 .00000000>0 00000000000 000 000000 000000000 00 0000 00000500 000 .00005 0000 00000000000 00m 000000 0003000 0:0 0000030 000000 0000 APPENDIX C PERSONALITY INVENTORY FOR CHILDREN (PIC) PERSONALITY INVENTORY FOR CHILDREN ADMINISTRATION BOOKLET by ROBERT D. WIRT, Ph.D. PHILIP D. SEAT, Ph.D. WILLIAM E. BROEN, Jr., Ph.D. Published by WESTERN PSYCHOLOGICAL SERVICES PUBLISHERS AND DISTRIBUTORS 12031 WILSHIRE aouuwuto Los ANGELES. CALIFORNIA was A DIVISION OF MANSON WESTERN CORPORATION This inventory consists of statements about children and family re- lationships. DIRECTIONS: First till in the information requested on the answer sheet; then read each of the statements in this booklet and decide whether it is true or talse as applied to your child. Section of answe, sheet correctly marked Look at the example of the answer sheet shown Y N at the right. In the example the mother decided T F that statement 25 was true as applied to her child 25 I ; g and statement 26 was false as applied to her child. 26 : : ' It a statement is TRUE or MOSTLY TRUE, as applied to your child, use a pencil to blacken between the lines of the column headed YT (Yes or True column. See 25 in the example). It a statement is FALSE or NOT USUALLY TRUE. as applied to your child. blacken between the lines of the column headed NF (No or False column. See 26 in the example). In marking your answers on the answer sheet. be sure that the number of the statement agrees with the number on the answer sheet. Make your marks heavy and black. Erase completely any answer you wish to change. Do not make any marks on this booklet. Not to be reproduced in whole or in part Without written permisson of Western Psychological Serwces All rights reserved 2 3 4 S 6 7 8 9 Printed in USA 159 160 DO NOT MAKE ANY MARKS ON THIS BOOKLET 22. 23. 24. . My child learned to walk before he (she) was six years old. . My child seems average or above average in intel- ligence. . My child is small for his age. . Sometimes I think I‘m too easy with the child. . My child never talks to strangers. . My child tends to pity him (her) self. . My child often plays with a group of children. . My child usually kisses me before going to school or to play. . My child hardly ever smiles. . Others always listen when my child speaks. . My child has hit a school official (teacher etc.). . Several times my child had complaints, but the doctor could find nothing wrong. . Other children often get mad at my child. . Usually my child kisses his (her) parents before going to bed. . My child hard‘y ever needs punishment. . My child thinks others are against him or her for racial or religious reasons. . My child worries about things that usually only adults worry about. . My child was a blue baby. . I often wonder if my child is lonely. . Usually my child takes things in stride. 21. My child has many friends. My child is troubled by constant coughing. My child is likely to take remarks the wrong way. Little things upset my child. 25 26. 27. 28. 29. 30. 3|. 32. 33. 34. 35. 36. 37. 38. 39. 40. 4|. 42. 43. 45. 47. My child keeps thoughts to him (her) self. My child sometimes thinks he or she is someone else. Often my child has to go to bed with a cold. As a younger child, it was impossible to get my child to take a nap. It has been a long time since our family has gone out together. At one time my child was unconscious with an injury to his (her) head. My child‘s manners sometimes embarrass me. My child has never mentioned his (her) heart racing or pounding. My child seldom gets a restful sleep. My child often tries to show off. My child is always humming to him (her) self. My child has had to have drugs to relax. My child has usually been a quiet child. At times my child has seriously hurt others. My child has never had cramps in the legs. My child has had a severe case of one or more of the following: measles. mumps, encephalitis (sleep- ing sickness), chicken pox, scarlet fever, whooping cough, meningitis. My child has a good sense of humor. At times my child yells out for no reason. My child sometimes sees things that aren‘t there. . As a child, my child hit other children on the head with sharp toys. My child often complains of being hungry. . My child is worried about sin. Stuttering has been a problem for my child. GO ON TO THE NEXT PAGE 48. 49. 50. SI. 52. 53. 54. 55. 57. 58. 59. 6|. 62. 63. 65. 67. 69. 70. 7|. 72. 73. 74. 161 My child will beg until l give in. The child‘s father has been fired from his job several times. Other children don't seem to listen to or notice my child much. My child is fairly helpful in doing chores around the house. My child is rather unattractive. My child is liable to scream if disturbed. My child sometimes undresses outside. My child hardly ever kisses me. . My child has little self confidence. Certain foods make my child ill. My child has no special talents. Our family seems to enjoy each other more than most families. . My child usually undresses him (her) self for bed. I often wish my child would be more friendly. My child broods some. My child could do better in school if he (she) tried. . My child can comb his (her) own hair. My child never liked to be cuddled. . At times my child gets so excited you can‘t under- stand his (or her) talk. Often my child destroys other children‘s toys. . The child‘s father seems jealous of the child. My child is usually rejected by other children. My child seems to enjoy destroying things. At times my child pulls out his (her) hair. My child usually comes when called. Now and then my child writes letters to friends. I am afraid my child might be going insane. 75. 76. 77. 78. 79. 80. 8 I. 82. 83. 84. 85. 86. 87. 88. 89. 9|. 92. 93. 94. 95. 97. 98. l0|. l02. My child sweats very little. My child seems to delight in smashing things. My child is over-confident in most things. My child has trouble making decisions. My child has had convulsions. Thunder and lightning bother my child. The school says my child needs help in getting along with other children. Lately my child has shown interest in religion. My child loves to hug and kiss. My child often gets up at night. Most of my child‘s friends are younger than he (she) is. Eating is no problem for my child. Others think my child is “easygoing". Sometimes I think my child's memory has been lost. There is a lot of swearing at our house. . l have found out my child has had sex play with the opposite sex. My child never takes the lead in things. My child often asks if I love him (her). My child first sat up before he was one year old. My child would probably take blame rather than lie. My child changes moods quickly. . Other children look up to my child as a leader. My child could ride a tricycle by age five years. My child takes criticism easily. My child sometimes gets angry. . My child often jumps into things without thinking. My child sometimes hears things others don‘t hear. My child sometimes swears at me. GO ON TO THE NEXT PAGE |03. l05. 106. |07. l08. l09. IIO. ll l. 112. ”3. H4. ”5. H6. “7. ”8. “9. l20. l2l. |22. l23. l24. [25. l26. |27. l28. 162 My child is not worried about disease. . My child frequently complains of being hot even on cold days. My child‘s behavior often makes others angry. My child seems bored with school. The child‘s parents are now separated or divorced. My child gets exhausted so easily. My child belongs to a gang. My child plays a musical instrument. My child often expresses dislike for teachers. My child tends to talk faster than he (she) can think. 1 can‘t get my child to do his (her) school lessons. My child stays close to me when we go out. Often my child goes about wringing his (her) hands. My child is sometimes cruel to animals. Recently my child has complained of eye trouble. My child likes to build things from clay or sand. The child‘s parents have broken up their marriage several times. Sometimes my child runs errands for me. Others think my child is talented. My child is afraid of animals. My child frequently has gas on the stomach (sour stomach). My child is good at lying his (her) way out of trouble. My child often carries a cloth or doll for comfort. The child's parents sometimes forbid the child to play with certain other children. Sometimes my child gets so excited he (she) can‘t sleep at night. It is not too unlikely that my child will stay in the house for days at a time. I 29. 130. |3l. I32. I33. I34. I35. I36. I37. I38. I39. MI. I 42. 143. I44. [51. 152. 153. 154. l55. My child shows a lot of affection for a pet. My child usually gets up without being called. My child has had brief periods of time when he (she) seems unaware of everything that is going on. My child often cheats other children in deals. The child‘s parents have to keep after him (her) to do his (her) chores. My child is good at leading games and things. My child is more nervous than most children. My child‘s feelings are hurt easily. My child usually runs rather than walks. My child sometimes irritates others with practical jokes. My child never played peek-a-boo. . My child never worries about what others think. Sometimes my child earns extra money by doing small jobs around the neighborhood. The child's parents try to be as permissive as pos- sible. My child likes to dress like older children. Usually my child eats all the food on his (her) plate. . My child is different than most children. . A child has a right to disagree with his (her) parents. . Others have remarked how polite my child is. . My child has original ideas. . At one time my child had speech difficulties. . My child usually completes something once it is started. My child is afraid of dying. My child carries a weapon (knife, club, etc.). Pestering others is a problem with my child. My child believes in God. My child can cut things with scissors as well as can others of his (her) age. GO ON TO THE NEXT PAGE l56. [57. I58. l59. I70. l7|. I72. I73. I 74. I75. I76. I77. I78. I79. I80. 163 I feel I am very close to my child. My child has never been elected to an office in a club or school. My child doesn‘t seem to care for fun. My child often talks about how strong he (or she) is. . At times my child has hit and kicked me. . My child sometimes feels things that aren't there. . Mistakes are often made by my child just because of hurrying. . My child worries about hurting others. . My child doesn‘t seem to care to be with others. . My child seems to enjoy talking about nightmares. . Others have told me 1 baby my child. . My child has difficulty doing things with his (her) hands. . Several times my child has performed in front of a group. . Several times my child has asked if he (she) were adopted. Often my child will sleep most of the day on a hohday. Others think my child is mean. My child often stays in his (her) room for hours. My child seems to know everyone in the neigh- borhood. My child can cry one minute and laugh the next. At times my child scratches his (her) face until it bleeds. Voices sometimes tell my child to do things. Often my child talks back to me. My child has never had any paralysis. My child would never take advantage of others. My child will take the blame for others. [81. I82. I83. I84. I85. I86. I87. I88. I89. I9I. I92. I93. I94. I95. I96. I97. I98. 200. 201. 202. 203. 204. 205. My child has to be coaxed or threatened before he (She) will eat. My child has had an operation on his (her) head. My child's allowance is his (her) own to spend. My child usually blames others for any trouble. My child has more than three bowel movements a day. My child can be left home alone without danger. Starting school was very difficult for my child. My child jumps from one thing to another. My child is always talking about the future. . My child has been in trouble for attacking others. My child seldom breaks rules. How to raise the child has never been a problem at our house. My child belongs to a club. Several times my child has threatened to kill him (her) self. My child usually doesn't trust others. My child seems too serious minded. My child has more friends than most children. My child cries if left home alone. . Often my child goes to the toilet outside the house. Strength impresses my child. My child often hits younger children. My child has many friends of the opposite sex. Often my child does things before thinking. My child seems unhappy about our home life. When my child gets mad, watch out. . My child seems shy with the opposite sex. . My child never really forgives anyone. . My child really has no real friend. GO ON TO THE NEXT PAGE 2I0. III. 2I2. 2I3. 2l4. 2I5. 2I6. 2I7. 2l8. 2I9. 220. 22I . 222. 223. 224. 225. 226. 227. 228. 229. 230. 23l. 232. 233. 164 . My child often tells jokes. My child often tattles (tells) on others. My child has never been away from home at night. My child is as happy as ever. Others often remark how moody my child is. We often argue about who is the boss at our house. My child could walk downstairs alone by age five years. Sometimes my child will go into a rage. My child often complains that others don't under- stand him (her). My child has to be prevented from eating and drink- ing too much. The trouble with my child is a “chip on the shoul- der.‘ My child has very few friends. My child loves to make fun of others. My child likes to play active pines and sports. Others often remark how relaxed my child is. Sometimes I worry about my child‘s lack of concern for other’s feelings. Blushing is a problem for my child. Nothing seems to scare my child. My child can wash him (her) self as well as other children his (her) age. Often my child is afraid of little things. Often my child smashes things when angry. My child doesn't seem to be interested in practical things. I have often been embarrassed by my child‘s sassi- ness. My child tends to see how much he (she) can get away with. Others think my child is a “cry baby”. 234. 235. 236. 237. 238. 239. 240. 24 I . 242. 243. 244. 245. 246. 247. 248. 249. 250. 25 I. 252. 253. 254. 255. 256. 257. 258. 259. 260. My child can‘t seem to keep attention on anything. My child has never been in trouble because of sex behavior. My child almost never argues. My child gives in too easily. Playing with matches is a problem with my child. My child often disobeys me. The child's mother frequently has crying spells. My child cries when scolded. My child is better than average at sports. Falling down is a problem for my child. The child‘s parents are not active in community affairs. My child likes to show off. My child sometimes chews on his (her) lips until they are sore. My child has never been spanked. My child loves to rock back and forth when sitting down. My child is a good loser. My child loves to stay over night at a friend‘s house My child usually plays with older children. The child's father changes jobs frequently. My child has a weight problem. School has been easy for my child. Others have said my child has a lot of “personality". Sometimes my child wets the bed. My child goes to bed on time without complaining. My child belongs to Boy Scouts, Girl Scouts or some younger branch of these organizations. “Spare the rod. spoil the child” is a true saying. My child can‘t sit still in school because of ner- vousness. GO ON TO THE NEXT PAGE 26 I. 262. 263. 264. 265. 266. 267. 268. 269. 270. 27 l . 272. 273. 274. 275. 276. 277. 278.. 285. 286. 287. I65 My child has older brothers or sisters. I do not approve of most of my child's friends. My child vomits frequently after meals. Constipation has never been a problem for my child. My child tells of having the same dream over and over. My child likes to “boss” others around. Reading has been a problem for my child. I sometimes “blow up” at the child. My child doesn't seem to have any fear. Parents should be strict with their children. My child is very jealous of others. Five minutes or less is about all my child will ever sit at one time. My child is often restless. \Ve seldom argue about religion at our house. A scolding is enough to make my child behave. My child seldom misses school because of illness. Frequently my child looks under the bed before going to bed. We frequently argue about money matters at our house. . My child often talks about the Devil. . Often my child sings around the house. . My child sometimes disobeys his (her) parents. . My child tends to doubt everything others say. . Usually my childs legs or arms are swinging. . Several times my child has been in trouble for stealing. My child seldom complains of stomach aches. Neither parent has ever been mentally ill. My child takes sleeping pills to get to sleep. 288. 289. 290. 291. 292. 293. 294. 295. 296. 297. 298. 299. 300. 30 I . 302. 303. 304. 305. 306. 307. 308. 309. 3I0. 311. 312. 3I3. My child has never failed a grade in school. If my child can't run things. he (she) won‘t play. The child‘s parents can‘t seem to live within their income. Others have remarked about my child's unusual imagination. I have heard my child swear at others. The child's parents are often out socially. My child is in a special class in school (for slow learners). At times my child has to be held down because of excitement. Others think my child has a “know it all” attitude. My child usually plays alone. My child won‘t go into the bedroom without some- one else there. Several times my child took money from home without permission. Our family attends Church together. My child often talks to him (her) self. \ffection is frequently shown in our home. My child loves to work with numbers. Usually my child sees good in everybody. My child often talks about religion. My child sometimes eats too many sweets. My child has never been in trouble with the police. My child often brings friends home. My child could feed him (her) self fairly well by age five years. My child seldom visits a doctor. My child’s favorite stories are fairy tales or nursery rhymes. The child‘s father doesn‘t understand the child. Nakedness embarrasses my child. GO ON TO THE NEXT PAGE 3l4. 3l5. 3I6. 3I7. 3I8. 319. 320. 32 I . 322. 323. 324. 325. 326. 327. 328. 329. 330. 33 I. 332. 333. 334. 335. 336. 337. 338. 339. 166 Dizzy spells are no problem with my child. My child usually falls right to sleep once in bed. My child learned to count things by age six years. The child‘s father drinks too much. I have several times found my child masturbating (playing with self sexually.). My child could print his (her) first name by age six years. My child tends to brag My child doesn‘t seem to learn from mistakes. My child would rather be with adults than with children his (her) own age. My child can‘t seem to wait for things like other children do. My child tends to be pretty stubborn. My child rarely gets excited. My child often asks questions about sex. My child gets spanked about once a day. My child seldom talks. My child is constantly moving about. My child is very critical of others. My child seldom gets into mischief. My child always does his (her) homework on time. Sometimes during the night my child will crawl in bed with me. My child often vomits when getting a headache. My child is usually a leader in groups. Sometimes my child lies to avoid embarrassment or punishment. l have a terrible time getting my child to take a bath. Car sickness is a problem with my child. I always worry about my child having an accident when he (she) is out. 340. 341. 342. 343. 345. 347. 348. 349. 350. 35I. 352. 353. 354. 355. 356. 357. 358. 359. 360. 36l. 362. 363. 364. 365. Other children make fun of my child‘s different ideas. Our whole family seldom gets to eat together. My child usually stays neat and clean. Reading is my child‘s favorite pasttime. . My child loves excitement. My child is often ashamed of the family. . Often my child plays to hard. The child's father usually makes the important decisions at our house. “Bad days” are frequent with my child. My child often visits art museums or attends con- certs. My child insists on keeping the light on while sleeping. My child could be trusted to walk upstairs alone before he (she) was four years old. My child seems to prefer adults to children. Sometimes my child‘s muscles twitch. Much of my child‘s time is taken up with art or music. My child sometimes smears self and walls after going to the toilet. Punishment is usually given by the child‘s father. My child never stays out too late at night. My child seldom if ever has dizzy spells. Chewing fingernails is a problem for my child. My child is dependent on others. An interruption is likely to get my child angry. A lot of my child’s suggestions as well as actions are very impractical. During the past few years we have moved often. My child worries about talking to others. My child never sleep walks. GO ON TO THE NEXT PAGE 366. 367. 368. 369. 370. 371. 372. 373. 374. 375. 376. 377. 378. 379. 380. 381. 382. 383. 384. 385. 386. 387. 388. 389. 390. 391. 167 My child first talked before he (she) was two years old. My child gets common colds more often than most children. My child will usually admit being wrong. The child‘s parents disagree a lot about rearing the child. School teachers complain that my child can‘t sit still. Often my child locks himself (herself) in the bed- room. My child has some bad habits. Several times my child has spoken of a lump in his (her) throat. “Head in the clouds” describes my child. We often have friends in for a social evening. My child often wakes up screaming. My child drools when eating. My child has been with me since he(shc) was born. Often my child will laugh for no apparent reason. My child frequently has nightmares. My child is often the center of attention. My child almost never acts selfishly. My child sometimes skips school. My child is usually in good spirits. The child‘s parents are active in church. My child seems fearful of blood. My child is not as strong as most children. My child seems more clumsy than other children his (her) age. Others have remarked how self confident my child is in a group. Others often remark how sensible my child is. The child‘s father seldom helps around the house. 392. 393. 394. 395. 396. 397. 398. 399. 100. 101. 402. 403. 404. 405. 407. 408. 410. 4”. 412. 4I3. 4l4. 415. 416. 417. My child loves to play in water. Arguing is my childs biggest downfall. My child seems to understand everything that is said. My child will do anything on a dare. My child always seems to have a cold. At times my child just keeps on spinning around. Sometimes the child‘s father will go away for days after an argument. Sometimes my child gets so nervous his (her) hands shake. Skin rash has been a problem with my child. I have often found my child playing in the toilet. The child‘s father sometimes gets drunk and mean. My child often plays sports. My child sometimes becomes envious of the posses- sions or good fortune of others. Shyness is my child‘s biggest trouble. . My child often talks in rhymes. The child‘s mother makes most of the important decisions in the home. My child will do anything for a laugh. . My child is a healthy child. My child thinks others are ploting against him (or her.) My child has difficulty holding his (her) head up. Usually my child gets along well with others. The child‘s parents do not get along with the neighbors. My child seems eager to please others. My child seems to have no shame. Usually my child plays inside. The child‘s father seldom misses work. GO ON TO THE NEXT PAGE 418. 419. 420. 421. 422. 423. 424. 425. 426. 427. 428. 429. 430. 431. 432. 433. 434. 435. 4.16. 437. 438. 439. 44!. 442. 443. 168 My child gets lost easily. My child has the habit of picking. his (her) nose until it bleeds. My child has had asthma attacks. My child is put to bed early if he (she) disturbs the rest of the family. Often my child takes walks alone. My child often has headaches. The child‘s parents have set firm rules that must be obeyed. Often my child will wander about aimlessly. My child seems to get along with everyone. My child is easily embarrassed. My child is very popular with other children. My child gets confused easily. The child‘s father dislikes his present job. My child is almost always smiling. My child has more accidents resulting in cuts, bruises. and broken bones than other children. Several times my child has threatened to run away. At time? my child has difficulty breathing. There is always a lot of argument at our dinner table. Others don't understand my child. My child plays with friends who are often in trouble. My child seldom has nose bleeds. My child often talks of loving someone much older. . Parents should teach their children who is boss. My child has never been expelled from school. Sometimes my child acts like a clown. My child loses most friends because of his (or her) temper. . Our house is always in a mess. 445. 447. 448. 449. 450. 45 I . 452. 453. 454. 455. 456. 457. 458. 459. 461 . 462. 463. 464. 465. 466. 467. 468. 469. 470. 471. My child whines a lot. . My child is shy with children his (her) own age. My child doesn‘t seem to feel pain like others. My child was difficult to toilet train. My child wants a lot of attention when sick. My child saves most of his (her) spending money. The child’s mother or father have never been di- vorced. My child can count change when buying something. Winning a game seems more important than the fun of playing to my child. The child's mother strongly dislikes housework. My child has never run away from home. My child needs laxitives. My child shows unusual talent. A mother’s place is in the home. Speaking up is no problem for my child. . I had an especially difficult time with temper tan- trums in my child at an early age. My child worries a lot about physical health. My child can tell the time fairly well. Sometimes my child comes home with torn clothes. Sharing things has been no problem for my child. Many times my child has become violent. The child‘s parents always discuss important mat- ters before making a decision. I have a problem stopping my child from eating everything. The child‘s mother can‘t stand to stay home all day. Murder and crime stories seem to be my child‘s favorites. My child' insists on polished shoes. My child can take a bath by him (her) self. GO ON TO THE NEXT PAGE 472. 473. 474. 475. 476. 477. 478. 479. 480. 48 I . 482. 483. 484. 485. 487. 488. 489. 490. 491. 492. 493. 494. 495. 496. 497. 169 My child smokes at home. Recently my child has complained of chest pains. The child's father f requently“blows up” at the child. My child sees strange things. My child is shy with adults. Before going to sleep my child needs a teddy bear or doll in bed. Frequently my child argues with others. I have heard that my child drinks alcohol. There is seldom a need to correct or criticize my child. My child is rather absent-minded. Others have remarked how pale my child looks. My child bites his (her) fingernails or toenails. The child‘s father is home almost every evening. My child repeats numbers and letters over and over. . My child is always telling lies. Receme the child‘s parents have argued with the school officials. When talking my child often jumps from one topic to another. By the age of five years, my child could dress him (her) self except for tying things. ' My child most always tells me where he (she) is going to play. The child's parents seldom visit the school. My child boasts about being sent to the principal in school. My child never has fainting spells. My child is crabby most of the time. My child spends over fifteen minutes at a time combing his (her) hair. Music lessons have to be forced on my child. The child‘s father is too strict with the child. 498. 499. 500. 50 I . 502. 503. 505. 507. 508. 509. 510. 511. 5l2. 513. 514. 515. 516. 517. 518. 519. 520. 52l. 522. 523. 524. My child has as much pep and energy as most children. Recently the school has sent home notes about my child‘s bad behavior. A parent should try to treat a child as an equal. My child often has unusual ideas. My child will never clean his (or her) room. Sometimes my child will put off doing a chore. . My child is able to keep out of everyday dangers. My child often talks about death. . My child usually does just what you tell him (her) not to do. My child has frequently been hospitalized. My child likes parties. My child always shows affection to me. The child‘s father gets along fine with the child. Sex seems to concern my child more than others. My child is usually rested after a good sleep. My child has been difficult to manage. Children should be seen and not heard. Hardly a day goes by when my child doesn‘t get into a fight. My child often sits and reads the dictionary. Others say our family is close. Working puzzles is one of my child's favorite hobbies. Most of my child‘s time is taken up watching tele- vision. Frequently my child has a high fever. Sometimes my child's room is messy. l have seen my child laugh when others get hurt. My child often talks of flying off into space. Sometimes my child irritates me. GO ON TO THE NEXT PAGE 525. 526. 527. 528. 529. 530. 531. 532. 533. 534. 535. 536. 537. 538. 539. 540. 541. 542. 543. 544. 545. 547. 548. 549. 550. 551. 170 Often my child tells fantastic stories. The child‘s father is hardly ever home. My child is seldom short of breath. Sometimes I don’t understand what my child means. My child usually feels sorry when he (or she) has hurt others. My child is usually afraid to meet new people. My child almost never needs punishing or scolding. My child speaks of him (her) self as stupid or dumb. My child could eat with a fork before age four years. Often my child complains of blurring (blurred vision). There is a lot of tension in our home. My child needs protection from every day dangers. My child has a terrible temper. My child daydreams quite a bit. It is necessary for the child‘s mother to work outside the home. Several times my child has threatened to kill others. The child‘s father spends very little time with the child. My child refuses to do anything around the house. My child usually stays mad a long time. My child needs help when going to the toilet. My child is adopted. . My child runs around the house naked. My child always insists on wearing clean clothes. My child respects the property of others. My child seldom has back pains. Frequently my child will put his (her) hands over his (her) ears. The child‘s father has very little patience with the child. ' 552. 553. 554. 555. 556. 557. 558. 559. 560. 561. 562. 563. 564. 565. 566. 567. 568. 569. 570. 571. 572. 573. 574. 575. 576. 577. My child wants to sit in the bath tub for hours. The child‘s father has held the same job for the last five years (or since marriage). I have no trouble getting my child to bed at night. My child often speaks of being smarter than others. My child loves to read about murder and other crimes. My child didn't have colic as an infant. My child learned to drink from a cup by age three years. The child's parents frequently quarrel. Often my child sets goals that are too high. My child's headaches usually start with a pain in the back of the neck. Everything has to be perfect or my child isn‘t sat- isl'ied. The child‘s parents belong to several clubs or com- munity groups. My child gets pneumonia almost every year. Spanking doesn‘t seem to affect my child. Lately my child has had diarrhea a lot. My child was a “planned” child. My child talks a lot about his (her) size or weight. My child tends to repeat everything (parroting). My child has never had face twitchings. My child was completely toilet trained by three years of age. My child often will cry for no apparent reason. Both parents enjoy children. My child seldom talks about sickness. My child tends to swallow food without chewing it. My child will worry a lot before starting some- thing new. My child is afraid of strangers. GO ON TO THE NEXT PAGE 578. 579. 580. 581. 582. 583. 584. 585. 586. 587. 588. 589. 171 My child has trouble swallowing. My child had difficulty breathing at birth. My child shows a lot of interest in fire. My child usually looks at the bright side of things. My child is afraid of the dark. Our marriage has been very unstable (shaky). My child usually keeps his (her) mouth open. My child often has crying spells. My child often talks about the future. My child never seems to have a goal. Sometimes my child gets hot all over without reason. Nothing seems to get my child upset. 590. 591. 592. 593. 594. 595. 596. 597. 598. 599. END Delivery of my child was with instruments. Often my child will lick his (her)lips. My child seems tired most of the time. My child refused or couldn‘t suck as an infant. My child is exceptionally neat and clean. Others have remarked how smart my child is. My child takes illness harder than most children. My child was a premature or over-due baby. Money seems to be my child's biggest interest. My child goes on dates with the opposite sex. . Usually my child will sleep all night without awakening. APPENDIX D TEACHER RATING AND SCHOOL INFORMATION LAFAYETTE CLINIC DIVISION OF CHILD AND ADOLESCENT PSYCHIATRY 951 E Lafayette Detrort. Michigan 48207 SEQUENCE NO. cflo 1 l I - TEACHER RATING AND SCHOOL , . .I . . INFORMATION I FORM MAILED TO: PRINCIPAL or /19 (NAME or scaoou nouaan AND s'rnss'r NAME ' eirv ‘ srars ' ZIP cooe A CHILD'S sums Fem Middle Last sex summons 0 M C] r /19 II a 14 as The above named child or adolescent has been referred to the Division of Child and Adolescent Psychiatry at Lafayette Clinic. To help us in our evaluation of this student. would you please see that this questionnaire is completed by the staff members who are most familiar with this child. In addition. it is eatremely important that all psychological. psychiatric and social work information be forwarded to our department. Release Of information forms have already been signed and are enclosed. Please return this complete questionnaire in the envelope provided. Your cooperation is greatly appreciated. Thank you. Intake Secretary Telephone Please respond to each item: (To be completed by Principal or Social Worker) U D 1. The school has information regarding a psychological assessment of this child. These results will be forwarded to us No Lafayette Clinic. 9s 9 2. The school has information regarding social work contacts and a summary will be forwarded to Lafayette Clinic. 0 D 3. Psychiatric or other evaluation material are available and will be forwarded to Lafayette Clinic. ‘ res no cars mmcwacs SIGNATURE TELEPHONE NO. EXTENTION OHM-2471! I72» EVALUATION OF BEHAVIOR PROBLEMS. (To be completed by the teacherisi and Other school staff who have had the most contact with this student). Please read each statement carefully and decide if you have observed that behawor items in the Child s response pattern during the last two month period (or during the last two months of school. if this form is completed during summer vacation). ll y0u have observed the behaVIor described in the statement during this period. place an X' in the YES' eclumn. if not. place an 'x 173 In the “NO" column. g COL COL Complains about others‘ unfairness and/or discrimination towards him. it Comments that nobody likes him. I35 ls little. and continually tired. ‘2 Repeats one idea. thOught, or aetwity over and Over. 3 Does not conform to Iil'l'lltl on his own without control from others. ‘3 Has tanner tantrums. 37 “memwnmmmnem. 14' Reremo hlrnieireedunib.tnioid.orhoepeb~. h Comments that no one understands him. 16 Does not .99... m mm, ”mm“ 3 Perfect-emetic: meticulous about having eeerything exactly right. 15 mmmrrubxx'oiwx' "k“ °‘" "" Imm‘mm” so mamvflnzzy? ”mm" m m m "u" "M m " °' 17 Has rapid rnood I'llfts: dence-so one moment. menu: the next. “ Other children act as it he were taboo or tainted. is Does not obey until threatened with punishment. ‘9 Hes difficulty concentrating for any length of time. 19 Complains of nluitrnares. bad dreams ‘3 Is overactive. restleu. and/or continually drifting body positions 20 Expresses concern bout being lonely, unhappy. ‘4 Apotogises mpaatedly for him-If and/or his behavior 21 Openly strikes back with angry behavior to taaslng of other dilldren. ‘5 Distorts m. mm w M.” moonm ‘0 feet. 122 finer.- concern about something terrible at horrible hopaning to a. Underachieving: performs below his demonstrated ability leeeI. 23 Has no friends. 4" Oimibeother children: mmmlm mitigation. 24 Muphweapproyelreneauemiedoroomolepd. 4| Trleetowoldoellineettentiontehimeelr. . 12s Displayeohyecelaureaaiontowaideobiecporpenons ‘ rig“ distrustful or mapleiOus remarks about actions of others toward 2. u "wanna of “mm". 150 :wxm:::::m:r' in mm. with MW my 21 Does not complete tsdts attempted. 5' Argues and must have the last word in verbal exchange; 27 Ooem't protest when others hurt. tease. or criticise him. 53 Approaches new tadrs and situations with an "I can't do it" 29 ISM“ or mu. heterosexual “mu“ 5; :l'asmus ties: muscle-twitching, eye-blinking, nail-biting, hand- x ISteels "m” from “M children. 5‘ Habitually rejects the school experience through actions or comments 31 Does not initiate relationships with other children. 5' Has enuresis. (Wars bed.) 32 Reacts with defiance to instructions or commands. 5‘ Utters nonsense syllables andlor babbles to himself 33 Weeps or cries withOut provocation. 57 Continually seeks attention. 34 IStuttfls. stammers. or blocks on mine words 5' Easily distracted away from the task at hand by ordinary clauroom 50 stimuli, i.e., minor movements of others. house. etc. Frequently stares blankly into mace and is unaware of his surroondmgs when deing so. Reproduced by permission walker Problem Behavior Identification Checklist COpyrigl-it @ I970 by Western Psychorogical Serialces, Los Angeles. All rights reserved. F—-———_—-——-n , f ' P00"- -" NW“ Platte an X to lidicale THIS student's Current grade placement 1 l 11 - 12 _J 1135! completed grade If during Summer) I REGULARCLASSES «T172713 14151617 [819 {10111 I121 OR SPECIAL EDUCATION: E mentally .mpaireo Learning disabled E Emor-onally mos-red Please compare this child's current achievement with other children of the same ag and place an ‘X‘ to indicate the appropriate achievement level In each area: 1-2 years 2 or more below "m bum Do nor know COLUMNS ACHIEVEMENT LEVEL Gulls 4 17 9 II Mathematics If any area of this child's achievement is below age expectation, please indicate the reasons you feel form the basis for this difficulty. (mark all Items that apply): 'X' ICOL. Has with vision Hos instructions Poor motor coerdination class attendance Poor fine motor coordination ‘t homework Bored/not interested 't eta. Lacks basic skills from Gives up failure Bilingual limrted nutrition Emptional uninm in child's direction Poor drills Talkative reacts (behaved without - Spawn problems WI"! At this point, what would be your recommendation for the most appropriate (ideal) class placement? 'X' E; Remain at present grade placement 52 Remain at present grade placement. but needs special attentionlrernedial tutoring 53 Remain at present grade placement. but repeat grade next year 54 Transfer to special education: emOtIOnsIIy impaired 55 Transfer to special education learning disabled 56 T-ran-sfer to special educatIOn: mentally impaired 57 Transfer to vocational traonmrogrom 58 Camera to earlier grade 9 Promote to a higher grade 60 Fla-evaluate for possible return to regular clmom (if new in pecial education) at Has this child repeated a If "YES," indicate the number of times K i 2 3 4 5 6 7 B 9 to it 12 regular grade? tool that gradelsl has been repeated. 0 Yes D No “656768697071727374757677 None Mild :v.rv;'a-w Severe Column Considering your total teaching experience With children Of this ago, how much of a problem is the child at this time? 1 2 3 4 ,9 Y Y Y 3;."5 j): Dims DA'E 3: ‘”H|-0‘ - co: -\ D n ‘ ‘fl‘fl P} q _“S \ :‘q -E‘ ‘\ °“ "‘3" ""5 -‘-u".‘=SSiO-\J ~oac\T cc-lvl‘lG E C k” "W G l (Please feel free to continue yOur comments on additional paper.) Speufic Problem Behaviors: (Briefly summarize this child's specific behavior problems as you see them in the classroom) Health or Physical Problems if any: (i.e., wears glasses. falls asleep in class. poor grooming, etc.) Student's Reaction to Success and Failure: Family Attitude Toward School and the Student What are your personal reactions to this student? Signature lPrimnry person filling cut report) Title Date APPENDIX E CONSENT TO PARTICIPATE OFFICE or PUPIL PERSONNEL smevrces AND SPECIAL EDUCATION Dear Parents: A study is being conducted in that will give additional information on special education programs. Miss Elaine Clark, formerly a psychologist in the district and now a graduate student at Michigan State University, will be asking teachers and parents to fill out a behavior checklist for a number of special education students. In addition she would like to check the students' records to look at individual and group test scores. There will be no questioning or other contact with the stu- dents. If you are willing to participate in this study by filling out a checklist and allowing‘Miss Clark to review records, please sign and return this letter in the enclosed envelope within five days. All information will be confidential and coded so that names will not be used. If there are questions, please call the Special Education Office at Extension 307. Thank you for your cooperation. Sincerely, / . ./ (‘1?! II' ,1- fi (if 1' (1 {(1-1 Cri'jf'ém. l/ Katherine J. Elliott Ph.D. Executive Director Pupil Personnel Services and Special Education I agree to participate in the above study. Signature of Parent or Guardian Name of Student 176 APPENDIX F LETTER TO PARENTS MICHIGAN STATE UNIVERSITY COLLEGE OF EDUCATION EAST LANSING ' MICHIGAN ° 48824 DEPARTMENT OF COUNSELING AND EDUCATIONAL PSYCHOLOGY Dear Parents, Thank you for agreeing to participate in this study. Enclosed you will find the inventory and scoring sheets which parents are requested to complete. It generally requires 45-75 minutes to fill out. If possible, each parent should complete a separate inventory without discussing the responses. Please indicate your relationship to the student (mother, father, stepmother, etc.) at the top of the scoring sheet. Once the scores are coded, your child's name will be removed so as to keep all responses confidential. It is important that you try to respond to as many items as you can so that we can determine the usefulness of this inventory for public school children. Read and follow the instructions on the front of the test booklet. Please return the test booklet and scoring sheets in the return envelope within tgg_ ()0) days. If you have any questions, I may be contacted by calling 435-8400, extension 307 or 308. Sincerely, ..:,“ - +7 / “a M 2.- (p: 6.” Elaine Clark l7? MS U is are Affirmative Action/Equal Opportunity Institution REFERENCES REFERENCES Achenbach, T. M. The classification of children's psychiatric symptoms: A factor-analytic study. Psychological Monographs, 1966, 89, 1- 37 (7, Whole No. 615). Achenbach, T. M. Psychopathology of childhood: Research problems and issues. Journal of Consulting and Clinical Psychology, 1978, 56 (no. 4), 759-776a APA. Standards for educational and psychological tests and manuals. Washington: APA, 1966. Anastasi, A. Psychological testing (4th ed.). New York: Macmillan, 1976. Becker, N. C. The relationship of factors in parental ratings of self and each other to the behavior of kindergarten children as rated by mothers, fathers and teachers. Journal of Consulting Psychologm 1960,.g1. 507-527. Becker, L. Learning characteristics of educationally handicapped and retarded children. Exceptional Children, April 1978, 502-511. Bell, C. Parent attitudes and involvements in the placement process of handicapped students in special education programs in Aurora School District. Dissertation Abstracts International, 1976,.81, 7067. Bergan, J. R., & Tombari, M. L. Consultant skill efficiency and the implementation and outcome of consultation. Journal of School Psychology, 1976, lfi_(no. 1), 3-14. Berman, N. B. Behavioral correlates and validity of the PIC--Spanish translation. Unpublished doctoral dissertation, Indiana Univer- sity, 1979. Bernard, R. B. A description of a population of school verified learn- ing disabled children in the State of Michigan across certain se- lect variables. Unpublished doctoral dissertation, Michigan State UniVersity, 1978. Bryan, T., & Bryan, J. H. Understanding learning disabilities. Port Washington, NY: Alfred Pub. Co., 1976. 178 179 Cichon, M. J. Screening for personality disturbances in young children using the PIC. Unpublished doctoral dissertation, University of Minnesota, 1979. Clarizio, H. F., & McCoy, G. F. Behavior disorders in children (2nd ed) New York: Thomas Y. Crowell Co., 1976. Connolly, C. Emotional problems. In H. R. Myklebust (Ed.), Progress in leagning disabilities (Vol. 2). New York: Grune and Stratton, 97 . Cromwell, R. L., & Foshee, J. 6. Studies in activity level: IV. Ef- fects of visual stimulation during task performance in mental de- fectives. American Journal of Mental Deficiency, 1960, gg, 248-25l Cruickshank, N. M. Some issues facing the field of learning disabili- ties. Journal of Learning Disabilities, 1972, g, 380-388. Culbert, J., & Gdowski, C. L. Reading retardation: An initial investi- gation with the Personality Inventory for Children. Paper pre- sented at the Tenth Annual Meeting of International Neuropsychology Society, Pittsburg, February 1982. Cytryn, L., & McKnew, D. J., Jr. Proposed classification of childhood depression. American Journal of Psychiatry, 1972, 1g2, 149. Darlington, R. 8., & Bishop, C. H. Increasing test validity by consid- ering inter-item correlations. Journal of Applied Psychology, 1966, §Q, 322-330. Dawdy, S. C. An exploration of the relationship between Wechsler Verbal- Performance 10 discrepancies and disordered behavior in children. Dissertation Abstracts International, 1979, 59, 5805. Dean, R. 5. Patterns of emotional disturbance on the WISC-R. Journal of Clinical Psychology, 1977, §§_(no. 2), 486-495. DeBell, 5., & Vance, H. Concurrent validity of three measures of arith- metic achievement. Perceptual and Motor Skills, 1977, 35, 848. DeHorn, A. 8., Lachar, 0., & Gdowski, C. L. Profile classification strategies for the PIC. Journal of Consulting and Clinical Psy- chology, 1979, AZ (no. 5), 874-881. Dewitte, L. L. The effects of client race and social class on the at- titudes and decisions of school psychologists. Unpublished doc- toral dissertation, University of Cincinnati, 1976. Downey, C. Differential diagnosis of mildly handicapped. Dissertation Abstracts International, 1979, 59, 161. Draguns, J., & Phillips, L. Psychiatric classification and diagnosis: An overview and critique. New York: General Learning Press, 1971. 180 Dunn, L. M. Children with moderate and severe general learning disabil- ities. In L. M. Dunn (Ed.), Exceptional children in the schools (2nd ed.). New York: Holt, Rinehart and Winston, 1973. Education for All Handicapped Children Act of 1975. Public Law 94-142. November 1975. 20 USC 1401. Elders, D. D. Biased teacher referrals and their effects upon school psychologists' diagnostic judgments and expectations. Unpublished doctoral dissertation, Michigan State University, 1977. Forness, S. R. Behavioristic orientation to categorical labels. Jour- nal of School Psychology, 1976, 15 (no. 2), 90-96. Frame, R. E. Interclinician agreement and bias effects of race, socio- economic status, and school achievement on school psychologists' diggnostic and treatment recommendations. Unpublished doctoral dissertation, Michigan State University, 1979. Froman, P. K. The development of a depression scale for the Personality Inventory for Children. Unpublished manuscript, University of Minnesota, 1971. Froman, P. K. The development of a mental retardation scale using the Personality Inventory for Children. Dissertation Abstracts Inter- national, 1973, 33, 5013-5014. Gaddes, W. H. Learninngisabilities and brain function. New York: Springer-Verlag, 1980. Gajar, A. H. Characteristics and classification of EMR, LD and ED stu- dents. Dissertation Abstracts International, 1979, 38, 4090. Gdowski, C. L. Typologies of childhood psychopathology and the rela- tionship with the PIC, Unpublished doctoral dissertation, Wayne State University, 1977 (a). Gdowski, C. L. Factor analyses of PAI and TRSI problem behavior check- lists. Unpublished data, 1977 (b). Gdowski, C. L., Lachar, 0., & Butkus. A methodological consideration in the construction of actuarial interpretation system. Journal of Personality Assessment, 1980, §3_(no. 4), 427-433. Gilbert, G. M. A survey of "referral problems" in metropolitan child guidance centers. Journal of Clinical Psychology, 1957, 1}, 37-44. Glidewell, J., & Swallow, C. The prevalence of maladjgstment in ele- mentary schools. Chicago: University of Chicago Press, 1968. Goldberg, L. R., & Werts, C. E. The reliability of clinicians' judg- ments: A multitrait-multimethod approach. Journal of Consulting Psychology, 1966, §Q_(no. 3), 199-206. 181 Gutkin, T. B. WISC-R scatter indices: Useful information for differ- ential diagnosis. Journal of School Psychology, 1979, 17 (no. 4), 368-371 . ‘- Hallahan, D. P., & Kauffman, J. M. Introdugtion to learning disabili- ties: A psycho-behavioral approach. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1976. Hallahan, D. P., & Kauffman, J. M. Labels, categories, behaviors: ED, LD and EMR reconsidered. Journal of Special Education, 1977, jl_(no. 2). 139-149. Hammill, D. 0. Defining "LD" for programming purposes. Academic Ther- gpy, 1976, lg_(no. 1), 29-37. Harris, L. M. R. A Spanish translation of the Personality Inventory for Children (PIC): A preliminary validation. UnpubTishedeoctoral dissertation, University of Minnesota, 1976. Hegeman, G. A PIC scale for hyperactive childggg. Unpublished doctoral dissertatibn, University of Minnesota, 1976. Heinicke, C. M. Learning disturbance in childhood. In B. Wolman (Ed.), Manual of child psychopathology. New York: McGraw-Hill, 1972. Hobbs, N. Issues in the classificatign of children (Vols. 1 and 2). San Francisco: Jossey-Bass, 1975. Hollingshead, A. B. Two-factor index of social position. Unpublished manuscript, Yale University, 1957. Huelsman, C. D. The WISC subtest syndrome and poor readers. Perceptual Motor Skills, 1970, 39, 535-550. Hughes, R. WISC-R factor loadings and students referred for special education services. Dissertation Abstracts International, 1977, _3__8_, 1384. Iano, R. P. Social class and parental evaluation of educable retarded children. Education and Training of the Mentallngetarded, 1970, 5. 62-67. Iscoe, I. The functional classification of exceptional children. In E. P. Trapp and P. Himmelstein (Eds.), Readings on the exceptional child. New York: Appleton-Century-Crofts, 1962. Johnson, D. M. Personality differences between low and high achieving boys using the Personality_Inventory for Children. UnpUblishedTT' doctoral dissertation, University of Minnesota, 1971. Kauffman, J. M. Characteristics of behavior disorders in children. Columbus, OH: Charles E. Merrill, 1977. 182 Kaufman, A. 5. Factor structure of the WISC-R at eleven age levels be- tween 6% and 16% years. Journal of Consulting and Clinical Psy- chology, 1975, 43, 135-147. Kaufman, A. 5. Intelligent tgsting with the WISC-R. New York: John Wiley and Sons, Inc., 1979. Keogh, B. K., & Hall, R. J. WISC subtest patterns of educationally handicapped. Psychology in the Schools, 1974, ll_(no. 3), 296-300. Keogh, B. K., Wetter, J., McGinity, A., & Donlon, G. Functional analy- sis of WISC performance of learning disorder, hyperactive and men- tally retarded boys. Psychology in the Schools, 1973, (no. 10), 178-181. Kirk, S. A., & Elkins, J. Characteristics of children enrolled in the child service demonstration centers. Journal of Learning Disabili- ties, 1975, 8 (no. 10), 630-637. Klinedinst, J. K. Relationships between MMPI and PIC data from mothers of disturbed children. Dissertation Abstracts International, 1972, 32, 4860. ’ Klinedinst, J. K. Multiphasic measurement of child personality: Con- struction of content scales using the PIC. Journal of Consulting and Clinical Psychology, 1975, 33, 708-715. Lachar, D. The development of a childhood psychosis scale using the PIC. Dissertation Abstracts International, 1971, 31, 4340. Lachar, 0. Construction of an academic achievement scale for the PIC. Unpublished manuscript, 1974. Lachar, 0., Abato, R., & Wirt, R. D. The construction of a delinquency scale for the PIC. Unpublished manuscript, 1975. Lachar, 0., Butkus, M., and Hryhorczuk, L. Objective personality assess- ment of children: An exploratory study of the Personality Inven- tory for Children (PIC) in a child psychiatric setting. Journal of PersonalityiAssessment, 1978, 12, 529-537. Lachar, 0., & Gdowski, C. L. Problem-behavior correlates of the PIC profile scales. Journal of Consulting and Clinical Psychology, 1979, 51, 36-45 (57: Lachar, 0., & Gdowski, C. L. Actuarial assessment of child and adoles- cent personality: An interpretive guide for_the PIC profile. Los Angeles: Western Psychological Services, 1979 (5). Lachar, 0., 8 Sharp, J. R. Use of parents' MMPIs in the research and evaluation of children: A review of the literature and some new data. In J. N. Butcher (Ed.), New developments in the use of the MMPI. Minneapolis: University of Minnesota Press, 1979. 183 Lapouse, R., & Monk, M. A. Behavior deviations in a representative sam- ple of children: Variations by sex, age, race, social class, and family size. American Journal of Orthopsychiatry, 1964, 35, 436- 446. Lloyd, 1., Sabatino, D. A., Miller, T. L., & Miller, S. R. Proposed federal guidelines: Some open questions. Journal of LearninggDis- Lucas, A., Rodin, E., & Simson, C. Neurological assessment of children with early school problems. Developmental Medicine and Child Neu- rology, 1965,_Z, 145-156. McCarthy, J. M., & Paraskevopoulos, J. Behavior patterns of LD, ED and average children. Expgptional Children, 1970, 33, 69-74. McCoy, S. A. Clinical judgment of normal childhood behaviors. Journal of Consulting and Clinical Psychology, 1976,44 (no.5), 710 -7i4. McDermott, P. A. Congruence and typology of diagnosis in school psy- ghology: An empirical study. Psychology in the Schools, 1980, 11 n0. 1 ’ 12-230 Michigan Special Education Code as Amended January 14, 1977 Under the Provision of Public Act 198 of 1971. Lansing: Michigan Department Bf'Education, 1977. Myers, 6. H. A defensiveness scale for the Personality Inventory for Children. Dissertation Abstracts International, 1974, 35, 3504- 3505. Myers, C. E., & Hammill, D. D. Methods-for learningpdisorders (2nd ed). New York: John Wiley and Sons, Inc., 1976. Nicholls, C. J. Parameters of WISC-R performance among underachieving children. Dissertation Abstracts International, 1979, 49, 4498. Nie, N., Hull, C. H., Jenkins, J. G., Steinbrenner, K., & Brent, 0. H. Statistical package for the social sciences. New York: McGraw- Hill, 1975. Novick, J., Rosenfeld, E., Bloch, 0., & Dawson, 0. Ascertaining deviant behavior in children. Journal of Consulting Psychology, 1966, 39, 230-238. Oskamp, S. Overconfidence in case study judgments. Journal of Consult- ing Psychology, 1965, 33_(no. 3), 261-265. Paal, N., Hesterly, 5., & Wepfer, J. Comparability of the WISC and WISC-R. Journal of Learning Disabilities, 1979, 13, 348-351. 184 Peterson, C. R., 8 Hart, 0. H. Use of multiple discriminant function analysis in evaluation of a state-wide system for identification of educationally handicapped children. Psychological Reports, 1978, 43, 743-755. Peterson, C., 8 Hart, 0. Factor structure of the WISC-R for a clinic- referred population and specific subgroups. Journal of Consulting and Clinical Psychology, 1979, 42, 643-645. Pipp. F. D. Actuarial analysis of adolescent personality: Self report correlates for the PIC profile scales. Unpublished masters thesis, Wayne State University, 1979. Quay, H. C. The facets of educational exceptionality: A conceptual framework for assessment, grouping, and instruction. Exceptional Children, 1968, 33, 25-32. Quay, H. C., 8 Werry, J. S. Psychopathological disorders of childhood (2nd ed.). New York: John Wiley 8 Sons, Inc., 1979. Robbins, L. 0. Accuracy of parent recall of aspects of child develop- ment. Journal of Abnormal and Social Psychology, 1963, 33, 261- 270. Roskos, D. L. An investigation of the differences in mother's and fa- ther's responses on the Personality Inventory for Children. Un- published doctoral dissertation, University of Minnesota, 1974. Rutter, M. A children's behavior inventory for completion by teachers' preliminary findings. Journal of Child Psychology and Psychiatry and Allied Disciplines, 1967, 3_(no. 1), 1-11. Saccuzzo, J. E., 8 Lewandowski, D. G. The WISC as a diagnostic tool. Journal of Clinical Psychology, 1976, 33, 115-124. Schinka, J. A., 8 Sines, J. 0. Correlates of accuracy in personality ggzessment. Journal of Clinical Psychology. 1974, 30 (no. 3), -377. -_' Schooler, 0., Beebe, M., 8 Koepke, R. Factor analysis of WISC-R scores for children identified as learning disabled, educable mentally im- paired and emotionally impaired. Psychology in the Schools, 1978, 15, 478-485. Seat, P. Development of a general maladjustment scale for the PIC. Dissertation Abstracts International, 1969, 3Q, 1970. Seat, P. Construction of the F scale for the PIC. Unpublished manu- script, 1971. ' Seat, P., 8 Wirt, R. 0. Construction of a lie scale for the PIC. Un- published manuscript, 1973. Segal, M. Differential diagnosis of LD and ED. Dissertation Abstracts International, 1979, 33, 4152. 185 Shove, G. R. A test battery for the assessment of school learning dif- ficulties, and its relationship to reflection-impulsivity in second and third grade boys. Dissertation Abstracts International, 1972, 33, 5049. Silver, L. B. R. Recognition and treatment of anxiety in children and adolescents. In W. E. Fann, I. Karacan, A. D. Porkorny, 8 R. L. Williams (Eds.), Phenomenology and treatment of anxiety. New York: Spectrum Pub., 1979. Sines, L. K. The relative contribution of four kinds of data to accu- racy in personality assessment. Journal of Consulting Psychology, 1959. 2;. 483-495. Smith, M., Coleman J., Dobecki, P., 8 Davis, E. Intellectual character- istics of school verified LD children. Exgpptional Children, 1977, 53 (no. 4), 352-359. Stevenson, L. P. WISC-R analysis: Implications for diagnosis and in- tervention. Journal of Learning Disabilities, 1980, 13_(no. 6), 346-349. Strag, G. A. Comparative behavioral ratings of parents with severe MR, specific LD and normal children. Journal of Learning Disabilities 1972. §, 631-635. Torgensen, J. K. The role of nonspecific factors in the task perfor- mance of LD children: A theoretical assessment. Journal of Learn- jpgDisabilities, 1977, 19_(no. 1), 33-40. Vance, H., 8 Gaynor, P. Analysis of cognitive abilities for L0 chil- dren. Psychology in the Schools, 1976,.13 (no. 4), 477-483. Voelker, S. L. Methylphenidate in the treatment of hyperactivitygin children. Unpublished doctoral dissertation, Wayne State univer- sity, 1979. Waits, C. H. Cognitive functioning and the identification of behavior disorders and specific learning disabilities. Dissertation Ab- stracts International, 1979, 59, 175. Waldman, L. F. Intellectual characteristics of LD and emotionally handicapped children. Dissertation Abstracts International, 1979, 40. 1371. Walker, H. M. Construction and validation of a behavior checklist for the identification of children with behavior problems. Disserta- tion Abstracts International, 1967, 33, 978. Walker, H. M. Walker Problem Behavior Identification Checklist manual. Los Angeles: Western Psychological Services, 1970. 186 Webb, M. L. Individual differences in personality and adjustment of ex- ceptional children. Unpublished doctoral dissertation, University of Utah, 1977. Webster, R. E. An individualized method for assessing the LD child. Paper presented to the annual meeting of the Northeastern Educa- tional Research Association, New York, October 1977. Wechsler, 0. Manual for the WISC-R. New York: Psychological Corp., 1974. Werry, J. S., 8 Quay, H. C. The prevalence of behavior symptoms in younger elementary school children. American Journal of Ortho- psychiatry, 1971, 53, 136-143. Wirt, R. 0., 8 Broen, W. E., Jr. Booklet for the Personality Inventory for Children. Minneapolis: authors, 1958. Wirt, R. 0., Lachar, D., Klinedinst, J. K., 8 Seat, R. Multidimensional description of child personality: A manual for the—Personality Inventory_for Children. Los Angeles: Western Psychological Ser- vices, 1977. Wirt, R. 0., Seat, P. 0., 8 Broen, W. E., Jr. The Personalit Inventor for Children. Los Angeles: Western Psychological Services, 1977. Yarrow, M. R. Problems of methods in parent-child research. Child De- velopment, 1963, 34, 215-226. Zigler, E. Cognitive-developmental and personality factors in behavior. In J. M. Kauffman 8 J. S. Payne (Eds.), Mental retardation: Intro- duction and personal perspectives. Columbus, OH: Charles Merrill, 1975.