“Via - '-.'_..' .'-....‘. :mm.whiztétm-m .IMM .I n:.;.‘ 53‘ 4]“. . I‘.‘ CONGRUENCE 0F PARENTAL PERCEPTION, MARITAL SATISFACTION AND ’ CHILD ADJUSTMENT Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY DEBORAH RUDI‘SILL ALLEN 1977 -~ Venn-.2535? K‘PI LI“ 7.7 f 75 17 (o‘It‘fi— S filial-"gen State University This is to certify that the thesis entitled CONGRUENCE OF PARENTAL PERCEPTION, MARITAL SATISFACTION AND CHILD ADJUSTMENT presented by Deborah Rudisill Allen has been accepted towards fulfillment of the requirements for _Eh..D.._ degree in Jmhnlggy Mdjor professor DateE’in-t‘ QR) INN \( 0-7639 1 W7 “9%0212003 '3‘" ".V’ .. b‘fl/‘j 14“";25 ABSTRACT CONGRUENCE OF PARENTAL PERCEPTION. MARITAL SATISFACTION AND CHILD ADJUSTMENT By Deborah Rudisill Allen This study investigated the interrelations among: congruence of marital partners' self-perceptions and perceptions by mate; mar- ital adjustment and satisfaction; congruence of parents' perceptions of their child; and child adjustment. 0n the basis of previous theory and research, it was hypothesized that all of these variables would be fbund to be positively intercorrelated. The Locke-Wallace scale, the Interpersonal Checklist, and the Children's Behavior Checklist, Form Q, were used to assess marital adjustment and satisfaction, self- and mate-perceptions, and parents' perceptions of their child, respectively. Three measures of child adjustment were also derived from parents' ratings of their child on the Children's Behavior Checklist. Both fathers and mothers of llO children aged five to seven (63 boys, 47 girls) completed all three questionnaires. An independent assessment of the children's adjust- ment was obtained from their teachers' ratings of them on the Behav- ior Rating of Pupils scale and the Children's Behavior Checklist, Form Q. A correlational analysis yielded somewhat different results for girls and boys. For the sample of boys, most of the hypothesized intercorrelations among congruence of parents' self-perceptions and Deborah Rudisill Allen perceptions by mate, congruence of parents' perceptions of their child, and child adjustment as rated by teachers and parents proved signifi- cant. In addition, wives' marital adjustment was significantly cor- related with congruence of parents' self-perceptions and perceptions by mate, and with two measures of child adjustment as rated by the wives. Among the families of girls, parents' marital adjustment was significantly correlated with congruence between husbands' self- perceptions and their wives' perceptions of them. Congruence between wives' self-perceptions and their husbands' perceptions of them was significantly associated with congruence of parents' perceptions of their child and with two measures of child adjustment as rated by the teacher. In addition, congruence of parents' perceptions of their child was significantly correlated with parents' ratings of their child's adjustment. Several explanations fbr these sex differences were proposed, including possibly greater validity of teacher ratings of adjustment for boys than for girls. Post hoc analyses revealed mostly signifi- cant interrelations between congruence of parents' self-perceptions and marital adjustment, child adjustment, and congruence of parents' perceptions of their child. Results of the study were interpreted as evidence fbr a general dimension of family harmony (vs. conflict) related to children's social adjustment. CONGRUENCE OF PARENTAL PERCEPTION, MARITAL SATISFACTION AND CHILD ADJUSTMENT 'By Deborah Rudisill Allen A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1977 ACKNOWLEDGMENTS Now that I am finally finished, I would like to express my appreciation to all of those who helped make it possible. First and fbremost, I would like to thank Dr. Lucy Ferguson, my chair- person, for her support throughout the past five years. Her inval- uable guidance, thoughtful suggestions, sense of humor, and enthusiastic dedication to my research helped to keep me going through all of the ups and downs. I would like to thank Lucy es- pecially fbr allowing me to work at my own pace, fbr her unflagging belief in my competence, fer her overwhelming generosity with her time during these last hectic months and most of all, fbr her under- standing of my feelings throughout my work on this project. Her support has truly been above and beyond the call of duty, which makes it feel very special. I would also like to thank Dr. Ellen Strommen and Dr. Helen Benedict, two members of my committee, fbr the wealth of knowledge they brought to this project as well as for their thought provoking questions. To Dr. Larry Messe', the fourth member of my committee, I would like to say thanks fbr not showing up fbr my orals. I would also like to acknowledge the contributions of Jerry Michaels, Linda Giacomo, and Richard Ince, who ran all over town helping me to collect data; of Gary Stollak and Larry Hesse', who donated grant money to help me pay subjects; and of Don Grumman, who found me some NIMH money to help with statistical analyses. ii Special thanks go to all of the parents, teachers, and school principals who took part in this research; without their curiosity and support this project would not have been possible. Finally, I would like to acknowledge my friends, who have kept me sane throughout this past year: Bob Dave', who taught me how to play,'made sure that I did, and who lived with all my ups and downs; Lori LaFerriere and Peg Geggie, who cheered me up when I needed to laugh, provided a shoulder when I needed to cry, and who, most of all, were always there; and Mom and Dad, who came through once again and paid fbr this. Many, many thanks to all: I couldn't have done it without you. iii TABLE OF CONTENTS LIST OF TABLES ....................... LSIT OF FIGURES ...................... CHAPTER INTRODUCTION . ................... Child Maladjustment and Parental Conflict ..... Teachers' Ratings of Child Adjustment and Maladjustment ................. The Relationship Between School Adjustment and Overall Adjustment ............... Teachers' Ratings Compared to Clinicians' Judgments. ................... Validity of Teacher Ratings for Individual Children .................... Objectives of the Study .............. Hypotheses .................... METHOD ....................... Subjects ..................... Instruments and Procedures ............ Hypotheses .................... RESULTS ....................... Characteristics of the Sample ........... Distribution Characteristics ........... Relationships among Adjustment Measures ...... Hypothesis l ................... Hypothesis 2 ................... Hypothesis 3 ................... Hypothesis 4 ................... Hypothesis 5 ................... Hypothesis 6 ................... Hypothesis 7, 8, 9, l0, and ll .......... Post Hoc Findings ................. DISCUSSION ..................... Validity of Ratings of Child Adjustment ...... iv Page vi ll 20 20 Results Relating Congruence of Perceptions of the Child to Child Adjustment ........... Marital Adjustment and Child Adjustment . . . ; . Congruence of Parent Perception Measures and Child Adjustment ............... . Interrelations Among other Variables ....... Sumnary of Interrelations ............ Implications of Findings for Future Research . . . REFERENCES ........................ APPENDICES ........................ A Letter to East Lansing Parents ......... . Letter to Okemos Parents ............. C Instructions to Parents ............. D Background Information Sheet ........... E Locke-Wallace Scale ............... F The Interpersonal Checklist ........... G Children's Behavior Checklist, Form Q ...... H Behavior Rating of Pupils Scale ......... 1 Distribution Infbrmation for Variables of this Study ..................... Page 70 74 94 96 116 Table TO LIST OF TABLES Page Means and standard deviations of sample and popula- tion distributions for each of the BRP items ..... 36 Correlations and corresponding probabilities of total BRP scores with other teacher and parent adjustment neasures ....................... 38 Correlations and corresponding probabilities among parent and teacher adjustment measures derived from the CBC-Q ...................... 40 Correlations and corresponding probabilities for relationships among parent adjustment measures derived from the CBC-Q ................ 42 Correlations and corresponding probabilities for hypotheses relating congruence in parents' percep- tions of their child to the child's adjustment as rated by the teacher (Hypotheses la, lb, TC, and ld) . 44 Correlations and corresponding probabilities for hypotheses relating congruence in parents' percep- tions of their child to parents' ratings of their child's adjustment (Hypotheses 2a, 2b, and 2c) . . . . 46 Correlations and corresponding probabilities for hypotheses relating parents' marital adjustment to their child's adjustment as rated by the teacher (Hypotheses 3a, 3b, 3c, and 3d) ........... 48 Correlations and corresponding probabilities for hypotheses relating parents' marital adjustment to their ratings of their child's adjustment (Hypothe- ses 4a, 4b, and 4c) ................. 50 Correlations and corresponding probabilities fbr hypotheses relating congruence of parents' self- perceptions and perceptions by mate to their child's adjustment as rated by the teacher (Hypotheses 5a, 5b, 5c, and 5d) ............... . . . . 52 Correlations and corresponding probabilities for hypotheses relating congruence of parents' self- perceptions and perceptions by mate to their ratings vi Table ll 12 13 Page of their child's adjustment (Hypotheses 6a, 6b, and 6c) ..................... . . 55 Correlations and corresponding probabilities for Hypotheses 7, 8, 9, IO, and II ........... 58 Correlations of congruence of parents' self- perceptions with child adjustment as rated by teachers and parents ................ 61 Correlations of parents' and teacher's CBC-Q ratings of the child with total BRP scores as rated by the teacher ....................... 62 vii LIST OF FIGURES Figure Page I Distribution of total BRP scores for children of participating and non-participating parents (l5 classes 33 2 Distribution of Competence scores for children of participating and non-participating parents (total sample) ....................... 35 viii CHAPTER I INTRODUCTION A great deal of clinical literature in the areas of child psychopathology and family dynamics has suggested that congruence in parental perceptions of basic aspects of family life is important for the favorable development of children. In an investigation of this hypothesis, Ferguson and Allen (in press) examined the interre- lations among the following variables: congruence of marital partners' self-perceptions and perceptions by mate, marital adjustment and satis- faction, congruence of parents' perceptions of their child and child adjustment. As predicted, all of these variables were found to be positively intercorrelated, with most of the correlations reaching significance at the .05 level or better. The strongest set of asso- ciations found was between congruence of parents' perceptions of the child and child adjustment. Both similarity in partners' self-concepts and similarity between partners' descriptions of themselves and des- criptions of them by their spouses were significantly associated with marital satisfaction, which in turn was related to child adjustment. However, the dynamics of the relationship between parental harmony and child adjustment appeared to operate differently for boys and girls. The girls seemed to be more vulnerable to disharmony in the parental relationship: when either parent was happy in the marriage, he or she tended to rate the daughter more favorably. For boys, adjustment 1 was more closely associated with congruence in parental perceptions, especially of the father; when there was congruence between the wife's perceptions of her husband and his perceptions of himself, both par- ents rated the son more favorably. An overview of the research ante- cedents of Ferguson and Allen's study puts these findings in perspective. Child Maladjustment and Parental Conflict Both role theory and social learning theory argue that child adjustment problems result in part from inconsistencies and incompat- abilities in the parents' demands and expectations of their child. Hhen parents disagree about the kinds of behaviors which are rewarded, the child becomes frustrated and uncertain about what behavior is appropriate. Empirical attempts to relate interparent agreement about a particular child to that child's behavior have been inconclusive. Leton (l958), using Shoben's Parent Attitude Survey and a revision of the Minnesota Teacher Attitude Inventory for parents, found wider dis- agreement in attitudes between mothers and fathers of poorly adjusted children than between mothers and fathers of well adjusted children. His findings were not confirmed, however, by Medinnus (I963). Com- paring mothers' and fathers' responses to the Parent Attitude Research Instrument (PARI), the Attitude Toward Education Scale, and a Q-sort procedure, Medinnus fbund no significant correlations with adjustment fbr first grade children. It seems likely that this lack of confir- mation of Leton's results was due in part to Medinnus' use of a correla- tional analysis with a very restricted sample (33-38 pairs of parents). In addition, parent attitude inventories such as those used by Leton 3 and Medinnus tap parents' attitudes about children in general rather than their feelings about a particular child. A number of researchers have concluded that parental attitudes about their general philosophy of childrearing are less potent in affecting their child's behavior than their attitudes about that particular child (e.g. Becker and Krug, l965). Support fbr this notion is found in a study by Medinnus and Johnson (1970). Mothers and fathers rated their child on a 50-item semantic differential scale made up of inferential behavior items (e.g., "dominant-submissive" or "friendly-unfriendly"). Medinnus and Johnson fbund that parents of well-adjusted kindergarten children showed significantly greater agreement in ratings than parents of poorly adjusted kindergarten children. A similar study by Ferguson, Partyka, and Lester (1974) looked at interparent agreement about directlygobservable behaviors of a particular child. Parents' re- sponses to the Children's Behavior Checklist revealed 66 items which proved to discriminate between clinic and nonclinic children. An overall factor analysis revealed three general dimensions which dis- criminated significantly between clinic and nonclinic children: Impulsivity and Moodiness (which were attributed to clinic children to a greater extent) and Competence (which was attributed to nonclinic children to a greater extent). It was hypothesized that parents of nonclinic children would describe them in more favorable and less pathological terms and would show closer agreement in their perceptions than would parents of the clinic children. However, significant results were obtained only for the younger males in the sample (5-7 years old). Ferguson and Allen (in press) attempted to look at a more representative 4 sample of families with children in the primary grades to see if these sex differences could be replicated. ' The rationale for Ferguson and Allen's inclusion of additional measures of marital satisfaction and self- and spouse-perception has an empirical basis. Children who exhibit deviant social behavior have often been fbund to come from families characterized by marital strife and dissatisfaction as well as parent-child conflict (e.g., Hinder and Rau, I962; Becker, et al., I959; Clark and van Sonlners, I96l; Vogel and Bell, I968). The notion that the disturbed child is the target or scapegoat for his parents' conflicts appears frequently in the clin- ical literature. It would thus seem likely that agreement in parents' perceptions of their child is associated with more favorable adjust- ment for the child and is also an expression of marital harmony. In addition, couples who express satisfaction with their marriage have been fbund to be more similar (Dymond, I953, I954; Corsini, I956) and more congruent in their mutual perceptions than dissatisfied couples (Dymond, I953, I954; Luckey, l960a, b; Hobart and Klausner, I959; Taylor, I967; Tharp, I963). This latter variable (psychological empathy) has been considered to be an index of good communication be- tween marital partners. Mangus (I957), for example, suggested that the integrative quality of a marriage is reflected in the degree of con- gruence between partners' perceptions of themselves and their percep- tions of each other. Thus it was hypothesized that parents who described their children as more favorably adjusted would show closer agreement in their perceptions of their children, would report greater marital satisfaction, and would show greater agreement between descrip- tions of themselves and descriptions of them by their spouses than 5 parents who described their children as less favorably adjusted. As previously mentioned, this hypothesis was supported by Ferguson and Allen (in press). A shortcoming in the design used by Ferguson and Allen makes it difficult to interpret the major results. A brief review of the meth- odology of that study illustrates the problem. Both fathers and mothers of 97 children, aged five to seven,completed three question- naires. The Locke—Wallace Scale, the Interpersonal Checklist, and the Children's Behavior Checklist were used to assess marital adjustment and satisfaction, self— and mate-perceptions, and parents' perceptions of their child, respectively. The Children's Behavior Checklist was also used to derive a meaSure of child adjustment. By using husbands and wives as the source of all data, including child adjustment, no information about the representativeness of the sample could be ob- tained; Without an independent measure of child adjustment, it could not be determined whether the children actually represented a continuum of adjustment or whether the ratings simply reflected the rating styles or biases of their parents. For example, parents who are unhappy in their marriage might tend to rate their child and their spouse nega- tively as a reflection of their negative feelings about the family in \I \ general. Thus a child could be rated poorly adjusted by his parents \ when an independent observer might see him or her as well adjusted. Teachers' Ratings of Child Adjustment and Maladjustment The present study is designed to determine whether the relation- ships found among the variables in Ferguson and Allen (in press) can be replicated when an independent measure of child adjustment is also 6 obtained. Teacher judgments have been considered to be an economical and efficient means of identifying maladjusted children. As such, they seem well suited to the task of independently assessing child adjustment. A review of the literature offers support for this pre- mise. As Bower (I974) points out, there is a myth that "someone, some- where, somehow can assess behavior and/or mental health as a charac- teristic or state of being independent of the social context and social institutions in which the individual is living and functioning“ (p. 229). In actuality, competent or incompetent behavior can only be evaluated in relation to the primary institutions in which the behavior takes place. For the young child, the family and the school are the most important institutions. Thus, it makes sense to compare the child's adjustment in the school with his adjustment at home. Teachers have regular and continuous contact with a child and _sometimes his or her family over the course of a year. Most teachers are well educated about age-appropriate behavior and personality de- velopment of children and have observed a variety of children in a wide range of situations. When children are overly aggressive, apathetic, have learning problems, or stand out in any way, they attract the attention of their teachers. The teachers' day-to-day experience with a wide variety of normal behavior "gives them an un- equaled perspective for appraising inappropriate or deviant behavior" in children (Bower, I969, p. I4). Thus, when a child attracts their attention, they are able to use their backlog of experiences to assess whether a child needs additional support or assistance. As a result, the school is often the first social institution to observe the 7 inadequate intellectual or emotional development of the child (Bower, l96l). For the present research, the crucial question is not whether the teachers are the first to bbserve inadequate development in child- ren but whether their judgments of inadequate development are valid. Three aspects of teachers' ratings of child adjustment explored in the literature are relevant: I) how teachers' ratings of the child's school adjustment relate to the child's overall adjustment; 2) how teachers and clinicians compare as to .what constitutes maladjusted be- havior in children; and 3) the validity of teachers' ratings of the adjustment of individual children. A brief review of the literature in each of these three areas provides evidence for the validity of teachers' judgments as an independent assessment of child adjustment. The Relationship Between School Adjustment and Overall Adjustment There is some evidence to indicate that teachers' ratings of child adjustment in school are indicative of the more general adjust- ment of the child. Bower (I969) points out that failure in school restricts the occupational and social freedom of the individual and comes perilously close to constituting failure in life as well. Some support for this position is provided by Ginzberg (I959). In a monu- mental study of the ineffective soldier during world War II, Ginzberg fbund that although a higher level of educational achievement was no safeguard against emotional disturbance, lower educational achievement was associated with a higher incidence of emotional disturbance. Ginz- berg and his associates concluded that "inadequate education and emo- tional instability may have been reinforcing each other for a long time. 8 A disturbed childhood is likely to be reflected in learning difficul- ties" (p. ll8). Similar resutls were fbund by Robins (I966) in a 30-year follow- up of 524 child guidance clinic patients and a group of 100 nonpatients matched fbr age, sex, race, IQ, and neighborhood. The control group consisted of children who had not been seen at a psychiatric clinic, had not repeated a grade, and had an IQ of 80 or more. Robins con- cluded that "a high proportion of those who as adults have psychiatric problems and social maladjustment must show very gross signs of dif- ficulty while still in elementary school. Hhile having repeated grades 'hielementaryschool certainly does not efficiently predict serious adult problems, having ggt_had serious school difficulties may be a rather efficient predictor of absence of gross maladjustment as adults" (p. 70). A similar study by Fitzsimons (I958) involved a follow-up of I58 persons who had been referred I5 to l8 years earlier by their teachers to a school agency because of poor adjustment. She found that the teachers were able to select accurately those children in need of psychological or psychiatric treatment. Fitzsimons concluded that "teachers are quite competent in identifying young children who are prone to serious maladjustment later" (p. I49). Thus, the evidence strongly suggests that the child's adjustment in the classroom reflects his/her adjustment in the world outside the classroom as well. The present study is based on that assumption, in that teachers' ratings provide the independent index of the child's adjustment. It must be noted, however, that some children may behave very differently at home and at school. For example, a child may find 9 in the classroom the stability s/he lacks at home; as a result, the child might be a model pupil at school and a hellion at home. In the present study the teacher's adjustment rating of such a pupil could not be distinguished from that of a child who was well adjusted both at home and at school. However, as the research evidence suggests, these cases are probably few. TeacherS' Ratings Compared to Clinicians' Judgments Another way in which the validity of teachers' judgments of their pupils' adjustment have been assessed is by comparing the teachers' judgments with those of clinicians. The assumption in such comparisons is that clinicians are the experts on adjusted and maladjusted behavior; the more the teachers' judgments agree with those of the clinicians, the more valid the teachers' judgments are presumed to be. Although the present study is ultimately concerned with the validity of teacher's judgments about individual children, a brief review of the literature with regard to how teachers make these judgments provides relevant back- ground information. Comparisons of teachers' and Clinicians' views of what constitutes maladjustment behavior have been a subject of controversy for many years. A classic study by Hickman (I928) revealed that teachers and mental hygienists differed markedly when asked to rate the seriousness of 50 behavior traits of children (correlation of about zero). This study has been widely quoted as indicating that teachers are poor judges of the mental health of their pupils. In fact, the ratings were made on the basis of distinctly different instructions and thus are not directly comparable. 10 In spite of this procedural shortcoming, Schrupp and Gjerde (I953) replicated Hickman's study. They wanted to see if expanded and improved knowledge of factors influencing the growth and develop- ment of children, and the increased emphasis in teacher education pro- grams on understanding the psychological makeup of the child, had resulted in a changed teacher identification of, and attitude toward behavior problems of children. Although the clinician and teacher groups agreed much more closely in I953 than in 1928 (correlation of .56 as compared to about zero), definite disagreements were still evident. Again, it was difficult to determine the extent to which the disagreement was a result of the different instructions to teachers and clinicians. A later study by Mitchell (I942) attempted to correct this problem in experimental design; MitChell's study involved teachers from the same school systems used by Hickman. Mitchell used a mod- ification of the original Hickman scale, and had teachers as well as 63 mental hygienists rate the traits, fbllowing the directions given to the teachers in Hickman's original study. Mitchell reported a correlation of .70 between teachers and mental hygienists. This sug- gests that there can be good agreement between the ratings of these two groups of professionals about what constitutes adjusted and mal- adjusted behavior. As previously mentioned, what is more important for the purpose of this study is whether their judgments about adjustment agree for individual children. A review of the literature with regard to the validity of teachers' ratings of individual children is relevant and provides a clear justification for the use of teacher ratings in this investigation. 11 Validity of Teacher Ratjggs fbr Individual Children Teacher rating scales range from the simple to the complex. One of the most simple is that developed by Gildea, Glidewell, and ~Kantor (I96I). They asked teachers to rate children along a fOur-point scale: well adjusted; no significant problems; subclinically disturbed; clinically disturbed. Each of these categories was defined behaviorally. Ratings by psychiatric social workers who had followed the children in school or in treatment at the child guidance clinic were also obtained. On a sample of 9l third grade children, teachers and social workers agreed 86% of the time. On a later sample of 49 third grade children, they agreed 80% of the time. There were sex and social class effects in the teacher ratings, however. In general, the girls were seen by the teachers as better adjusted than the boys, and the middle-class children were seen as better adjusted than either the lower or upper- class children. A number of rating systems involve simple checklists of items. For example, Eisenberg, et al. (I962) designed a symptom checklist and a health inventory for use with nursery school children. Teacher ratings of disturbed and normal children (those referred to a program for emotionally disturbed children vs co-op nursery students) were made on both scales. Results indicated that both scales discriminated significantly between the well adjusted and the poorly adjusted child- ren. In addition, among the well adjusted children but not the poorly adjusted children, the boys showed significantly more symptoms on the symptom checklist than the girls; no sifnificant differences were found between the girls and boys on the health inventory for either the well adjusted or the poorly adjusted children. The authors of this study 12 note that the validity of this instrument is limited to differentia- tion between groups, since variation in scores within both the well adjusted and the poorly adjusted populations were such that there was considerable overlap. A slightly more complex checklist is described by Rutter (I967). It consists of 26 statements concerning the child's behavior. Rather than simply checking if the behavior is present, however, teachers check whether the statement "certainly applies,“ "applies somewhat,“ or “doesn't apply" to the child in question. Rutter found test-retest reliability of .89 and inter-rater reliability of .72. The validity of the scale was tested by comparing the scores of children in the general population with scores of children attending psychiatric clinics for emotional or behavioral disorders. Results confirmed that the scale was reasonably efficient in differentiating children attending psychiatric clinics from those in the general p0pulation. Four similar screening devices were compared in a study by Cowen, Dorr, and Orgel (I97l). The Teachers' Behavior Rating Scale, the Teachers' Adjective Check Test, the Ottawa School Behavior Survey, and the AML Behavior Rating Scale all had been shown previously to be reliable and to have predictive validity for identifying school mal- adaptation in primary grade students. The Teachers' Behavior Rating Scale (TBRS) consists of a 25-item list of behaviors that could be con- sidered maladaptive in class. The teacher rates each child on a 4-point scale for each characteristic: "does not apply," "shows mildly,“ "shows moderately," "shows very strongly." The teacher also rates the child on a subscale (TOR) which is a single 5-point scale of overall adjustment that ranges from well-adjusted (l) to poorly 13 adjusted (5). The TBRS requires approximately five to ten minutes per child to complete. Cowen, Zax, 1220, and Trost (I966) found that both the TBRS and the TOR discriminate significantly between third grade children identified by clinicians as Red-Tag (manifesting mod- erate to severe maladjustment) and Non Red-Tag children (p_< .OOI). The second screening device for primary grade students examined by Cowen, Dorr, and Orgel (I97I) is the Teacher Adjective Check List (TACL) (Cowen, et al., I966; Zax and Cowen, I969). This measure con- sists of 34 descriptive adjectives, l7 judged positive and I7 negative. The teacher rates each child on a three-point scale: "does not apply," "applies somewhat," and "describes very well." The TACL requires about five to ten minutes per child to complete. Like the TBRS, it has been shown to discriminate between adjusted and maladjusted child- ren (Cowen, et al., I969; Liem, et al., I969). The third screening device examined was the Ottawa School Be- havior Survey (0585) which lists 20 behaviors known to occur with some frequency in primary grades. The teacher simply checks whether each item does or does not apply. This scale can be completed in about two minutes per child. Pimm and McClure (I967) found that this sur- vey discriminated significantly between children diagnosed as emotion- ally disturbed and a control group matched on the basis of age, sex, and IQ score. The fourth screening device examined was the AML Behavior Rating Scale (Brownbridge and Van Vleet, I969) which consists of II items designed to screen rapidly for school adaptation problems. It includes five aggression items, five moodiness items, and one item that re- flects learning difficulty. Children are rated on a five-point scale 14 from one ("seldom or never") to five ("all of the time"). This scale requires about one minute per child to complete. Brownbridge and Van Vleet (I969) have shown that children with high AML scores do signif- icantly more poorly than peers on a variety of educational and person- ality measures. As previously mentioned, Cowen, Dorr, and Orgel (I97I) investi- gated the interrelations among these four screening devices. Data were obtained for a sample of 266 kindergarten children and lOl first graders. Correlations relating total scale scores to each other ranged from -.6l (0585 vs. TACL, first grade) to .90 (AML vs. TBRS, first grade), with a median of .775. The negative correlation (-.6I) re- flects the different scoring systems of the two scales rather than an inverse association; high scores on the OSBS and lbw scores on the TACL indicate maladjustment. Thus, in spite of differences in content and item specificity, the four measures were found to be similar in tapping maladjustment. One limitation all of these measures is that they do not take into account the problems of rater bias and halo effect. There are no data to indicate whether all children judged to be poorly adjusted by one teacher are similar, or whether those judged to be poorly ad- justed by one teacher are similar to those judged to be poorly adjusted by another teacher. Nadine Lambert and Eli Bower, in their research for the California State Department of Education, developed a teacher rating procedure that seems to minimize rater bias and halo effect. The Behavior Rating of Pupils scale-3 (Lambert and Bower, l96l; Lam- bert, I967) is a simplified Q-sort procedure in which a teacher places every child in his/her class on a grid arranged with squares in an 15 approximately normal distribution. One grid is used for each of eight items referring to children's behavior in school. Every child's name goes in one of the squares on the grid, depending on whether the be- havioral item being rated is seen by the teacher as being more or less typical of that child. The sum of the teacher's ratings of each child on the eight items provides an overall rating of how the teacher views the child's coping ability and school functioning. Each of the items has been validated by comparing the distribution of ratings for a given item with clinical appraisal of school adjustment status as well as with other criteria of school effectiveness such as grades, achievement test results, and peer ratings of behavior. For second and fifth grade pupils, the individual item ratings as well as the total ratings have been shown to be correlated significantly with indepen- dent judgments of the presence and absence of school problems made by a clinical team (Lambert, 1967). This rating procedure appears to minimize the problems of halo effect and rater bias, since children judged to be poorly adjusted on a particular item by one teacher were similar to those judged to be poorly adjusted by another teacher (Lanbert, l967). Several earlier studies provide further evidence for the val- idity of the items in the Behavior Rating of Pupils scale. Bower (I960), using a successive category procedure, obtained teachers' ratings of 200 fourth, fifth, and sixth grade children. Teachers com- pleted a rating form for every pupil in their classes. The initial validity of this procedure was determined by taking the teachers' ratings of children known to a child guidance clinic and comparing them to the ratings of the remainder of the pupils in the class. 0f 16 . those pupils known to a child guidance clinic as emotionally handi- capped, 87% were rated by their teachers as among the most poorly adjusted in the class. Teacher judgments of emotional disturbance were very similar to those of clinicians. In addition, teachers sel- ected about the same number of children who were overly aggressive or defiant most of the time as those who were overly withdrawn or timid nest of the time. A fellow-up study by Lambert (described in Lambert, I967) com- pared those children who had been identified by the school guidance office as having problems in school with a group of children rated by their teachers as having problems in school, but who had not been referred to the guidance office. The study revealed that the two groups had a comparable nunber of delinquents, dropouts, and school failures. This confirmed the predictive validity of teachers' percep- tions of problems and absence of problems when entire classes were rated. In a later Study by Lambert and Bower (l96l), teachers rated 650 primary grade pupils on the same eight statements of school be- havior used in the Behavior Rating of Pupils scale—3. A successive category procedure was used (all of the time, some of the time, rarely, never.) which assigned a. numerical value to each rating. In order to control for differences between teachers in the degree to which they perceived the presence of any of the behaviors in their classrooms, the five pupils with the most negative ratings for each teacher's class were studied. Clinical psychologists subsequently evaluated each of these children. In 9l% of the cases, the psychol- ogist confirmed that the children who were rated by their teachers as 17 having the most trouble in school were the children with real psycho- logical problems . Lambert and Bower (I96l) also attempted to devel0p and validate measures of peer ratings and self-ratings of behavior. A later study (Lambert, I964) investigated the unique contribution of each of these sources of information for the criterion measure (clinicians' combined judgments of school adjustment). IQ scores and social status measures reflecting father's occupation, income, family housing, and neighbor- hood were found to increase predictive efficiency slightly. The teacher rating was found to have the highest correlation with the independent clinical measures, although peer and self-ratings did contribute pos- itively to the assessment of pupil mental health status. Several other studies have provided additional information about the Behavior Rating (rf Pupils scale. Maes (I966) studied 40 emotion- ally disturbed children in grades four, five, and six and their 548 normal classmates using the same variables as in Bower (l960). Using a multiple regression method, Maes fbund that the variables most pre- dictive of emotional disturbance were teacher ratings (behavior), arithmetic achievement, group intelligence score, peer ratings, teacher rating (physical status), and reading achievement. However, he found that the prediction achieved with teacher ratings of behavior and group IQ scores alone was as effective as the use of six variables. ‘A slightly modified version of Bower's (I960) screening process was given by Stennett (l97l) to children in grades four, five, and six. This included teacher, peer, and self-ratings as well as measures of each child's achievement. A year later, ratings of 46 children who had been screened out as either moderately or seriously emotionally 18 handicapped were compared with ratings made by their current teachers. Results indicated that 83% of the children were classified in the same way each year (g_< .00l). In addition, the screening system was found to be equally reliable in identifying emotionally handicapped children as nonemotionally handicapped children. Three years later, a number of the original sample who were screened while in grades four, five, and six were screened again. A simplified pupil behavior rating scale and a ten-item sociometric type test were used. Seventy-two percent of the children were classified in the same way after a three-year interval. Stennett concluded that the screening devices were adequate for identifying emotionally handicapped children and that a significant number of children identified as emotionally handicapped did not re- solve their adjustment problems without help. An alternate explana- tion for these findings is possible, however. It is not clear from Stennett's study whether the teachers had knowledge of the classifica- 'tion given to the children (i.e., emotionally handicapped or non- 'emotionally handicapped). If that infbrmation was available to the teachers, it is possible that the labeling process and the teachers' responses to the children as a result of the labels accounted for the persistence of the children's problems. A study by Harth and Glavin (l97I) attempted further to validate teacher judgments about personality adjustment as a screening tech- nique. Using an abbreviated Bower-type screening instrument (Bower, I960), teachers completed a rating sheet noting the children whom they considered to be the five best adjusted and the five poorest adjusted .students. No definitions of emotional disturbance were given to the teachers. The California Test of Personality (CTP) was also 19 administered. From the teacher ratings, three groups of l09 children each were selected: best adjusted (BA), poorest adjusted (PA), and average adjusted (AA). Results indicated that the BA group had a significantly higher CTP score than both the AA and PA groups. In addition, the AA mean score was significantly higher than the PA mean score. Thus, when the criterion is CTP scores, teacher ratings on a Bower-type scale seem to be a valid technique for screening emo- tionally disturbed children. To summarize, the validity criteria for the Behavior Ratings of Pupils have been CTP scores, referral to a child guidance clinic, con- current clinical team judgment of effectiveness of behavior, and follow-up information indicating the persistence of earlier school problems. The results of this work suggest that when using the Bower- Lambert Behavior Rating of Pupils scale, "teachers can provide a good index of the child's ability to cope in the classroom situation .and . . . they are able to make valid judgments of those children who are more or less effective in managing the school setting" (Lambert, I967, p. 439). It is clear from the literature that a wide variety of valid techniques are available fbr determining teacher ratings of child adjustement. However, the Behavior Rating of Pupils developed by Bower and Lambert, which was designed to minimize rater bias and halo effect, seems to be the one best suited for the purposes of the pre- sent investigation. 20 Objectives of the Study It has been assumed that if "high risk" children can be ident- ified early in their lives they can most easily be helped and their problems most effectively removed. Very few studies have fecused on early identification of parents and prospective parents, whose child- rearing values and attitudes, and actual behavior with children, might be highly predictive of their children's development of psychological problems. The overall purpose of this study is to discover the rela- tionships between differences in marital partners' perceptions of each other; marital dissatisfaction; differences in marital partners' perceptions of their children; and child maladjustment in school. If these factors are found to be highly correlated, there will be impli- cations for prevention and treatment of psychopathology. These may involve fbcusing interventions more on the family as a whole or just the marital partners rather than fecusing mainly on the child. Family-oriented therapy might include specific procedures to reduce discrepancies in marital partners' perceptions of each other as well as work on specific conflict areas. If high correlations among these factors are found in the expected directions there will also be implications for using the marital perceptions or dissatis- faction measures as a screening device to identify prospective par- ents whose children may have a high likelihood of becoming psycho- logically disturbed. Hypotheses In light of the theoretical and research evidence suggesting relationships among marital satisfaction, congruence of self-perception 21 and perception by mate, congruence of parents' perceptions of their child, and child adjuStment, and the evidence presented for the val- idity of teacher ratings of child adjustment, the following hypotheses are proposed: I) There is a significant positive correlation between congru- ence in parents' perceptions of their child and their child's adjust- ment as rated by the teacher. 2) There is a significant positive correlation between congru- ence in parents' perceptions of their child and their ratings of their child's adjustment. 3) There is a significant positive correlation between parents' marital adjustment and their child's adjustment as rated by the teacher. 4) There is a significant positive correlation between parents' marital adjustment and their ratings of their child's adjustment. 5) There is a significant positive correlation between congru- ence of parents' self-perceptions and perceptions by mate and their child's adjustment as rated by the teacher. 6) There is a significant positive correlation between congru- ence of parents' self-perceptions and perceptions by mate and their ratings of their child's adjustment. 7) There is a significant positive correlation between congru- ence of parents' self-perceptions and perceptions by mate and congru- ence of their perceptions of their child. 8) There is a significant positive correlation between congru- ence of parents' self-perceptions and congruence of their perceptions of their child. 9) There is a significant positive correlation between parents' III 22 marital adjustment and congruence of their perceptions of their child. IO) There is a significant positive Correlation between parents' marital adjustment and congruence of their self-perceptions and percep- tions by mate. II) There is a significant positive correlation between congru- ence of parents' self-perceptions and their marital adjustment. CHAPTER 11 METHOD Subjects The total sample consisted of ll0 families, each of which in- cluded a father, a mother, and a child in the five to seven year age range--47 female children and 63 male children. Mean education level for the mothers was one to three years of college; mean educa- tion level for the fathers was a bachelor's degree plus some graduate school, short of a Master's degree. Subjects were recruited in the following manner: letters to parents requesting their cooperation in a research study were given to teachers of all of the kindergarten, first. and second grade children in the East Lansing. Michigan, Public School System, and to first and second grade teachers in eight classes in the Okemos, Michigan, Public School System. Copies of the letters are fbund in Appendices A and B, respectively. The teachers distrib- uted the letters to the children in their classes, and the children were requested to take the letters home to their parents. Parents who were willing to participate were asked to fill out an enclosed, stamped, addressed postcard and return it to the experimenter. They were offered $5 for their participation. ' The returned postcards were sorted according to school classes. Teachers from those classes in which parents of five or more children agreed to participate were asked to participate themselves. Seventeen 23 24 teachers (l6 female, I male) from the East Lansing School System and eight teachers (all female) from the Okemos School System agreed to participate. Parents of the children in these classes who had agreed to participate (lOl from East Lansing, 75 from Okemos) were then mailed a packet containing the fbllowing: a letter thanking them for their participation and Containing instructions for completing the instru- ments; a background infbrmation sheet; two copies of the Locke-Hallace ‘ Marital Adjustment scale; two copies of the Interpersonal Checklist; two copies of the Children's Behavior Checklist, Form Q; and a pre- addressed stamped envelope in which to return the completed instru- ments. Parents were asked to fill out the instruments independently of each other. Copies of the letter, background information sheet, and each of the instruments are found in Appendices C, D, E, F, and C, respectively. I Of these I76 families who had agreed to participate, Ill re- turned completed packets (63 from East Lansing, 48 from Okemos). Each family was paid $5 for their participation. For each mother- father pair, the instruments were inspected to insure that the items were not checked identically. 0n the Interpersonal Checklists and Children's Behavior Checklists, if at least five disagreements were not fbund between a given couple, that couple was dropped from the sample on the suspicion that the parents may have discussed the items. 0n the Locke-Hallace scale, at least three disagreements between each husband-wife pair were required or the couple was dropped from the sample. Only one set of checklists was dropped for this reason. .‘I 25 Instruments and Procedures The Locke-Hallace Marital Adjustment scale (L-H) was used as a measure of marital adjustment and satisfaction. This scale consists of fifteen items relating to different aspects of marital life. Locke and Hallace (I959), using 236 subjects, found that this scale had a split-half reliability of .90 (Spearman-Brown correction). In addi- tion, using two groups of 48 subjects matched for age and sex, Locke and Hallace found that this scale clearly differentiated between per- sons who were welI-adjusted and persons who were maladjusted in marriage. A score of ICU or above is considered to be an indicator of marital happiness, whereas a score below l00 reflects marital discord and dis- harmony (Locke and Hallace, I959). A copy of the Locke-Hallace scale and its scoring key is found in Appendix E. A Locke-Hallace scale was completed by each parent. Separate scores were determined for the husband and wife. A third marital adjustment score, a congruence score, was determined by calculating the correlation (product moment correlation coefficient) between each husband's and wife's responses to the Locke-Hallace scale. The Interpersonal Checklist (ICL) (LaForge and Suczek, I955) was the instrument used to assess parents‘ perceptions of self and spouse. This scale consists of 128 descriptive items selected to represent interpersonal variables. Instructions to subjects marking the ICL solicit direct, conscious, verbal responses which tap the sub- ject's perception and his willingness to express this perception. The self- and spouse-concepts were operationally defined for the purpose of this study as all of the statements made by the individual on the ICL when describing him/herself and his/her spouse, respectively. 26 Each parent was asked to go through the checklist twice, the first time checking any of the l28 items which applied to self, and the second time checking those that applied to spouse. Thus, for each family unit, there were four sets of ICL scores: The husbands' perceptions of himself (H); the husband's perceptions of his wife (H-H); the wife's perceptions of herself (H); and the wife's percep- tions of her husband (H-H). Items that were checked were given a score of one; items not checked were scored zero. Inter-item phi coefficients expressed the agreement between those ICL items the hus- band checked for himself (H) and those his wife used to describe him (H-H); between the wife's description of herself (H) and her husband's description of her (H-H); and between the husband's description of himself (H) and his wife's description of herself (H). The Children's Behavior Checklist, Form Q(CBC-Q was the instru- ment used to assess parents' perceptions of thier children. It consists of 64 readily observable, interpersonal and symptomatic items referring to the behavior of children. Form Q of the CBC is a short form of the l54-item Children's Behavior Checklist compiled by Ferguson, MacKenzie, and Does at Michigan State University. It includes 50 items of the 66 which were found by Ferguson, Partyka, and Lester (I974) to discimin— ate significantly between clinic-referred and non-clinic children, 'twenty-five which were found to be significantly more characteristic of clinic-referred children and twenty-five which were found to be signifi- cantly more characteristic of non-clinic children. The fifty items were chosen in the following manner: of the 66 differentiating items, 53 were chosen which could be easily observed in a playroom situation (28 non-clinic items and 25 clinic items). Two hundred undergraduate students at Michigan State University were then asked to rate these 27 53 items as to whether they thought they were positive or negative behaviors for children to possess. An analysis of their ratings re- vealed that three of the non-clinic items were considered to be neg- ative traits for children. These items were then excluded from the group of non-clinic items, but were left in the checklist as filler items; eleven additional filler items were selected from the original checklist. Appendix 0 includes a copy of the checklist with each item labeled as a clinic item (C), non-clinic item (NC), or filler item (F). The checklist consists of two columns:, one asks about the child, "Does this apply at all?" and the other asks, "Is it characteristic?" Each parent is asked to go through the checklist twice, the first time checking those items which sometimes apply to the child, and the second time checking those items which are characteristic of the child's behavior (which describe the child most of the time). Agreement between father's and mother's ratings of their child on the CBC-Q was determined for each family as fellows: each parent's response to each of the 64 items was coded zero (not checked), one (first column checked), or two (second column checked). Inter-item correlations (product moment correlation coefficients) were then cal- culated between the mother's coded ratings of her child on the CBC-Q and those of the father.) The CBC-Q was also used as a measure of child adjustment. Each parent's score on the 25 clinic items (Clinic score) and on the 25 non-clinic items (Nonclinic score) was calculated. An overall adjust- ment score was also obtained by subtracting the Clinic score from the Nonclinic score. 28 The Behavior Ratingof Pupils 3 (BRP) (Bower, I969) was used to determine teachers' judgments of their pupils' adjustment. The} Bower scale consists of eight statements about children's behavior; a ninth statement, relating to the child's competence, was added to the scale. For each statement, the teacher placed every child in the class on a numbered grid with squares arranged so as to approx- imate a normal distribution. For each statement, the teacher placed the girls on one grid and the boys on a separate grid. Specific in- structions to teachers and copies of the rating scales are found in Appendix I. A total maladjustment score was determined for each child by obtaining the sum of his/her scores on each of the eight behavior ratings. Teachers rated all of the children in their classes on the BRP. In addition, teachers completed a CBC-Q for each child whose parents had agreed to participate in the study. Teacher's ratings of the childfis adjustment on the CBC-Q (Clinic score, Nonclinic score, overall adjustment score) were determined in the same way as the par- ents' ratings on that instrument. Teachers with ten or fewer pupils included in the study were paid $I5 for completing the rating forms; teachers with more than ten children included were paid $20. Hypotheses Based On the operational definitions given for each of the var- iables, the hypotheses are restated as fellows: la. There is a significant negative correlation between inter-item correlations of husbands' and wives' ratings of their child on the CBC-Q and the sum of the teacher's ratings of the child on the BRP. lb. There is a significant negative correlation between inter-item correlations of husbands' and ines' ratings of the child on the CBC-Q and the sum of the teacher's ratings of the child on the Clinic items of the CBC-Q. lc. 1d. 2a. 2b. 2c 3a. 3b. 3c. 3d. 4a. 4b. 4c. 29 There is a significant positive correlation between inter-item correlations of husbands' and wives' ratings of their child on the CBC-Q and the sum of the teacher's ratings of the child on the Nonclinic items of the CBC-Q. There is a significant positive correlation between inter-item correlations of husbands' and wives' ratings of their child on the CBC-Q and the teacher's ratings of the child's overall adjustment on the CBC-Q. There is a significant negative correlation between inter-item correlations of husbands' and wives' ratings of their child on the CBC-Q and the sum of the parents' ratings of the child on the Clinic items of the CBC-Q. There is a significant positive correlation between inter-item correlations of husbands' and wives' ratings of their child on the CBC-Q and the sum of the parents' ratings of the child on the Nonclinic items of the CBC-Q. There is a significant positive correlation between inter-item correlations of husbands' and wives' ratings of their child on the CBC-Q and the parents' ratings of their child's overall adjustment on the CBC-Q. There is a significant negative correlation between parents' Locke-Hallace scores and the sum of the teacher's ratings of the child on the BRP. There is a significant negative correlation between parents' Locke-Hallace scores and the sum of the teachers' ratings of the child on the Clinic items of the CBC-Q. There is a significant positive correlation between parents' Locke-Hallace scores and the sum of the teacher's ratings of the child on the Nonclinic items of the CBC-Q. There is a significant positive correlation between parents' Locke-Hallace scores and the teacher's ratings of the child's overall adjustment on the CBC-Q. There is a significant negative correlation between parents' Locke-Hallace scores and the sum of their ratings of their child on the Clinic items of the CBC-Q. There is a significant positive correlation between parents' Locke-Hallace scores and the sum of their ratings of their child on the Nonclinic items of the CBC-Q. There is a significant positive correlation between parents' Locke-Hallace scores and their ratings of their child's over- all adjustment on the CBC-Q. 5a. 5b. 5c. 5d. 6a. 6b. 6c. 10. 30 There is a significant negative correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL and the sum of the teacher's ratings of the child on the BRP. There is a significant negative correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL and the sum of the teacher's ratings of the child on the Clinic items of the CBC-Q. There is a significant positive correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL and the sum of the teacher's ratings of the child on the Nonclinic items of the CBC-Q. There is a significant positive correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL-and the teacher's ratings of the child's over- all adjustment on the CBC-Q. There is a significant negative correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL and the sum of the parents' ratings of their child on the Clinic items of the CBC-Q. There is a significant positive correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL and the sum of the parents' ratings of their child on the Nonclinic items of the CBC-Q. There is a significant positive correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL and the parents' ratings of their child's over- . all adjustment on the CBC-Q. There is a significant positive correlation between inter-item correlations of parents' self-perceptions and perceptions by mate on the ICL and inter-item correlations of their ratings of their child on the CBC-Q. There is a significant positive correlation between parents' Locke-Hallace scores and inter-item correlations of their ratings.of their child on the CBC-Q. There is a significant positive correlation between parents' Locke-Hallace-scores and inter-item correlations of their self-perceptions and perceptions by mate on the ICL. There is a significant positive correlation between inter-item correlations of parents' self-perceptions on the ICL and inter- item correlations of their ratings of their child on the CBC-Q. 31 II. There is a significant positive correlation between inter-item correlations of parents' self-perceptions on the ICL and their Locke-Hallace scores. The level of probability required for rejection of the null 'hypothesis in this study was equal to or less than .05. CHAPTER III RESULTS Characteristics of the Sample Since sensitive and concerned parents are more likely to volun- teer to participate in child development research, it was important to determine whether the current sample included children from the en- tire range of the adjustment continuum. Although the BRP was designed so that total adjustment scores could be obtained for all children in the participating classes, only 15 of the 25 teachers filled out the BRP's in such a way that this information could be obtained. Figure l illustrates the distribution of total BRP adjustment scores for all the children in those l5 classes, and for the children from those classes whose parents participated in the study. 0n the BRP, ‘high scores indicate poor adjustment whereas low scores indicate good adjustment. From Figure I, it appears that the current sample drawn from these I5 classes did include children from all along the adjust- ment continuum. A measure of the representativeness of the entire sample can be obtained by looking at the distributions of scores on the Competence item (Item 9) of the BRP. The Competence item was added to the BRP as part of another research study being conducted at the same time as the present study. Scores on the Competence item are not included in the total BRP adjustment scores. Figure 2 illustrates the 32' 33 55-56 = Total BRP scores of children in l5 - classes whose parents participated (N = 6I) 5l-54 = Total BRP scores of all children in IS classes 43—46 39-42 U) 0) S. 8 35-38 U) Q. 25. 31-34 '3 '3 27-30 '— 23-26 l9-22 l5-l8 II-l4 8-I0 l l l _l 4 I L L _l l J l L _F’ F F F I F I I I I l I—I’ I6 20 24 28 32 36 40 44 48 52 56 60 Number of Children Figure l. Distribution of total BRP scores for children of partici- pating and non-participating parents (I5 classes). 34 distributions of Competence scores for children of participating and non-participating parents. It is clear from Figure 2 that in terms of teachers' ratings of their pupils' Competence, the present sample is representative of the larger population from which it was drawn. Looking at the distribution of scores on the Competence item appears to be a valid method of estimating what the distribution of total BRP scores would have looked like if all of those scores had been available. For the children who participated in this study, there was a strong relationship between Competence scores and total BRP scores. Correlations between the two variables were -.8376 for the distribution of boys, -.696l for the distribution of girls, and -.7762 for the overall distribution; all of these correlations were significant (p_< .0005). A third way of looking at the representativeness of the present sample is to compare the distribution of children of participating parents with the population distribution from which they were drawn for each of the BRP items. Table l lists the means and standard de- viations for each item. Item numbers correspoond to the order of the BRP items in Appendix I. As Table l illustrates, the means and stan- dard deviations of the sample distribution are very similar to those of the population distribution. However, there appears to be slightly more variation between the two distributions for the sample of boys than fer the sample of girls. Distribution Characteristics For the distribution of girls, the distribution of boys and the total distribution, means, standard deviations, skewness, and kurtosis v.0..- Number c. 35 = Competence scores of children whose parents participated (N = 63 boys, 47 girls) 65-9 so - 55 q 50 'i I 45 — 40-4 35.. 30-« 25_m Boys Competence score = Competence scores of all children in 25 classes (N = 323 boys, 289 girls) Girls Figure 2. Distribution of Competence scores for children of partic- ipating and non-participating parents (total sample). 36 Table I. Means and standard deviations of sample and population distributions fer each of the BRP items. Mean of Mean of Standard Standard total dis- partici- deviation deviation tribution pating of total of participa- sample distrib. ting sample Item I Boys: 4.0 3.89 .83 .93 Girls: 3.87 3.79 .86 .76 Item 2 Boys: 4.0 4.4 .73 .68 Girls: 4.l2 4.l9 .76 .83 Item 3 Boys: 3.67 3.57 .78 .7 Girls: 3.77 3.75 .79 .7 Item 4 Boys: 3.8 3.l3 .79 .57 Girls: 3.78 3.64 .8 .89 Item 5 Boys: 3.84 3.35 .84 .95 Girls: 3.78 3.79 .82 .82 Item 6 Boys: 3.82 3.52 .79 .8 Girls: 3.8l 3.96 .79 .72 Item 7 Boys: 3.78 3.52 .74 .85 Girls: 3.86 3.81 .73 .78 Item 8 Boys: 3.78 3.48 .80 .84 Girls: 3.6 3.5l .69 .72 Item 9 Boys: 4.09 -4.48 .80 .69 Girls: 4.l9 4.28 .80 .83 37 for each variable are given in Appendix I. For the following varia- bles, the distributions (for both boys and girls) were found to be slightly negatively skewed: husbands' and wives' Locke-Hallace scores; H and H correlations on the ICL; H and H-H correlations on the ICL; correlations of parents' ratings of their child on the CBC-Q; Com- petence scores and Item 2 scores on the BRP; husbands', wives', and teachers' ratings of the children on the Nonclinic items of the CBC-Q; husbands', wives', and teachers' overall adjustment scores, derived ifrom the CBC-Q; and correlations of teachers' and wives' ratings of the child on the CBC-Q. Both distributions were found to be slightly positively skewed for the fbllowing variables: scores on BRP items I, 3, 4, 5, 6, 7, and 8 and husbands', wives', and teachers' ratings of the children on the Clinic items of the CBC-Q. The distributions of total BRP scores; H and H-H correlations on the ICL; and correla- tions of husbands' and teachers' ratings of the child on the CBC-Q were found to be slightly positively skewed for the distribution of boys, and slightly negatively skewed for the distribution of girls and the overall distribution. It should be noted that there was a great deal of missing data on Item I of the Locke-Hallace scale. Sixteen husbands and seventeen wives failed to complete Item I; for these Ss, Item I was scored l5, which is the score given when the midpoint of the scale for Item I is marked. Relationships among Adjustment Measures Table 2 illustrates the relationships between total BRP scores and each of the adjustment measures derived from the CBC-Q. Table 2. Correlations and corresponding probabilities of total BRP scores with other teacher and parent adjustment measures. Total BRP Scores Boys Girls Total * * 'A' ' Clinic score by .3083 .2907 .2682 husbands (p_< .007) (p_< .024) (p_< .002) Clinic score by .3425* .274l* .2989* wives (2_< .003) (E_< .03l) (2.< .00I) Clinic score by .6652* .4162* .5565* teachers (p_< .0005) (p_< .002) (p_< .0005) Nonclinic score -.l769 -.0803 -.ll47 by husbands (p_< .083) (p_< .296) (p_< .ll6) Nonclinic score -.2759* .0082 -.l464 by wives (p_< .0l4) (p_< .478) (p_< .063) Nonclinic score -.5394* -.2464* -.4302* by teachers (p_< .0005) (p_< .047) (p_< .0005) Overall adjustment -.2958* -.2209 -.2284* by husbands (p_< .009) (2.< .068) (2.< .008) Overall adjustment -.3866* -.1452 -.2725* by wives (p_< .001) (p_< .l65) (p_< .002) Overall adjustment -.6834* -.4066* -.5792* by teachers (p_< .0005) (p_< .002) (p_< .0005) * indicates significant correlation (g < .05) 39 Correlations between total BRP scores and teachers' Clinic scores, Non- clinic scores, and overall adjustment scores were all significant. -How- ever, the correlations for the distribution of girls were all smaller than the correlations for the distribution of boys; these differences were all significant (p_< .05). Although the intercorrelations among the teacher measures were all significant, it should be noted that they represented only 29% to 46% of the common variance between mea- sures for the distribution of boys and only 7% to l8% of the common variance between measures for the distribution of girls. Correlations between total BRP scores and husbands' and wives' Clinic scores were all significant; again, the correlations for the distribution of girls were smaller than the correlations for the dis- I tribution of boys. Parents' Nonclinic scores were found to be only weakly associated with total BRP scores. Only the correlation of wives' Nonclinic scores with total BRP scores fer the distribution of boys reached significance. For the distribution of boys and the { overall distribution, parents' overall adjustment scores were sig- / nificantly correlated with total BRP scores; however, the correlations for the distribution of girls did not reach significance. Table 3 illustrates the relationship among the various parent and teacher adjustment measures derived from the CBC-Q. All of the correlations among teacher scores were found to be significant. Par- ents' Clinic scores were found to be correlated significantly with teachers' Clinic scores and teachers' overall adjustment scores for all three distributions, and with teachers' Nonclinic scores for the distribution of boys and overall distribution. Parents' Nonclinic scores were found to be Correlated significantly with teachers' 40 Table 3. Correlations and corresponding probabilities among parent and teacher adjustment measures derived from the CBC-Q. . . Nonclinic Overall g;12;§c:ggge score by adjustment teachers by teachers Clinic Boys: .3423*** - 2769** — 3515*** score by Girls: .2807* - 2344 - 3l96** husbands Total: .3437***** - 2321*** - 3372***** Clinic Boys: .3879**** - 2253* - 3462*** score by Girls: .3972*** - 0726 - 2797* wives Total: .4036***** l556* 3248***** Clinic Boys: - 5469***** - 8744***** score by 61‘. HS: 2796* 7688***** teachers Total: - 4384***** - 8396***** Nonclinic Boys: .l06l .3420*** .2470* score by Girls: .0888 .0015 .0527 husbands Total: .1284 .l760* .l803* Nonclinic Boys: .1547 .2480* n.2300* score by Girls: .0489 .0676 .0736 wives Total: .1305 .l638* .l74l* Nonclinic Boys: .5469***** .8844***** score by Girls: .2796* .8290***** teachers Total: .4384***** .8588***** Overall Boys: .2674* .3772**** .3572*** adjustment Girls: .2223 .1277 .2145 by husbands Total: .2796*** .25l4*** .3l25***** Overall Boys: -.3304*** .2995** .3573*** adjustment Girls: .2535* .0892 .2070 by wives 'TotaI: .3196***** .2026* .3056**** * = p_< .05 ** = p_< .Ol *** = p_< .005 ***‘k = E. < .00] ***‘k'k = E < .0005 41 Nonclinic scores and teachers' overall adjustment scores for the dis- tribution of boys and the overall distribution only; correlations for the distribution of girls did not reach significance. There were no significant correlations between parents' Nonclinic scores and teachers' Clinic scores for any of the distributions. For the distribution of boys and the overall distribution only, correlations of parents' overall adjustment scores with teachers' Clinic scores, teachers' Nonclinic scores, and teachers'overall adjust- ment scores were all found to be significant. For the distribution of girls, the only significant correlation between parents' overall adjustment scores and the teacher adjustment measures was the correla— tion between wives' overall adjustment scores and teachers' Clinic scores. Table 4 illustrates the interrelationships among the various parent adjustment measures. For the distribution of boys and the over- all distribution, all of the parent adjustment measures were found to be significantly intercorrelated. For the distribution of girls, however, husbands' Clinic scores were not correlated significantly with either husbands' or wives' Nonclinic scores, and wives' Clinic scores were not correlated significantly with wives' Nonclinic scores. Husbands' Clinic scores were also not correlated significantly with wives' overall adjustment scores. All of the other correlations among parent adjustment measures were significant for the distribution of girls. To summarize the results of Tables 2, 3, and 4, there were sig- nificant intercorrelations among almost all of the parent and teacher adjustment measures for the distribution of boys. For the distribution 42 _oo. v.m n .e.e Po. v.m mooo. v.m n «area moo. v.m n are mo. v.m n a mocoum aaaaaomo. "PmuOh acmEumth< aaaamme. uwpgww Fpmgm>o «raaammo. "whom .mucmamzz «aakammw. «areammv. "Pouch mmgoum «aerawew. ampm. umpgwo 0wcwrucoz «kaaawmw. aaraammm. ”mxom .mm>w3 «aaakomm. «aaaenew. «areamme. "Punch meoum aararmme. «airrmem. kkeppv. umpgww uwzwpucoz «eaaaemm. aaaaamqm. aaatammc. "whom .mncmamaz raaeemmm.l aaaaawmm.l «kmmm.i aaaaaoom.i "pouch mmLoum «aaaanmo.l «ramo¢.i omp.l ampm.i "mFme um:WFu «araammm.l aaaeammm.i ammm.i aaromm.l umxdm mm>w3 reaeawm¢.i ashrammm.l earpom.l rakm©N.l aaaaamme. "Psych mwgoum mmp.l aaaaamom.l ewe. mmo.i «araapmm. umpsww uwcwpo ...e.hmm.- .....Fea.- s...~mm.- .Pm~.- ...e.eem. ”mesa .meeanasx museum mocoom mmgoum mmgoom mmcoom Sewfiwwwmm< bewflwwumw< aaee_aeoz. aaee_beoz 6F=a_o .mm>w3 .mvcanmzz .mm>w3 .mucmamsx .mw>w3 .oiumu we» gone vo>mcmu mmcammme newspmzwum ecosma ocean mawgmcowpopmg Low mumpw_wnanogq mcwccoammecou ace mcowpmpmssou .e mpnme he 43 of girls, parents' Clinic and Nonclinic scores were not significantly intercorrelated. In addition, Nonclinic scores for parents and tea- chers were independent, i.e. parents' Nonclinic scores were not sig- nificantly correlated with any of the teacher adjustment measures and teachers' Nonclinic scores were not significantly correlated with any of the parent adjustment measures. Most of the other intercorrelations among parent and teacher adjustment measures were significant for the distribution of girls. Hypothesis 1 Hypothesis I concerned the relations between congruence of par- ents' perceptions of their child and four measures of Child adjustment as rated by the teacher. The correlations found for Hypotheses la, lb, IC, and Id are given in Table 5, with their corresponding proba- bilities. Separate results are given for the distribution of boys, the distribution of girls, and the overall distribution. Significant correlations are marked with an asterisk (*). It was predicted in Hypothesis Ia that there would be a signif- icant negative correlation between correlations of parents' ratings of their child on the CBC-Q and the child's total BRP score as rated by the teacher. Results for this hypothesis are found in row one of Table 5. This hypothesis was strongly supported only for the distri- bution of boys and the overall distribution. Although the results for the distribution of girls were in the predicted direction, they did not reach significance. It was predicted in Hypothesis lb that there would be a signif- icant negative correlation between correlations of parents' ratings 44 Table 5. Correlations and corresponding probabilities for hypotheses . relating congruence in parents' perceptions of their child to the child's adjustment as rated by the teacher-(Hypotheses la, lb, IC, and Id). Inter-parent correlations of ratings of child on the CBC-Q Boys Girls Total (N = 63) (N = 47) (N = 110) Total BRP scores -.2868* -.0426 -.I701* (teacher) (p_< .01) (p_< .39) (p_< .04) Clinic score by -.2579* -.1592 -.2471* teacher (CBC-Q) (p_< .02) (p_< .14) (p_< .005) Nonclinic score by .3767* -.0404 .2009* teacher (CBC-Q) (p_< .001) (p_< .39) (p_< .02) Overall adjustment .3621* .0658 .2632* score by teacher (p_< .002) (p_< .33) (p_< .003) (CBC-Q) of their child on the CBC-Q and teacher's ratings of the child on the Clinic items of the CBC-Q. Results for this hypothesis are found in row two of Table 5. The hypothesis was strongly supported only for the distribution of boys and the overall distribution. Although the results for the distribution of girls were in the predicted direction, they did not reach significance. It was predicted in Hypothesis lc that there would be a signif- icant positive correlation between correlations of parents' ratings of their child on the CBC-Q and the teacher's ratings of the child on the Nonclinic items of the CBC-Q. Results fer this hypothesis are found in row three of Table 5. This hypothesis was strongly supported only for the distribution of boys and the overall distribution. The 45 results for the distribution of girls were in the opposite direction from that predicted, but the correlation was very small. It was predicted in Hypothesis Id that there would be a signif- icant positive correlation between correlations of parents' ratings of their child on the CBC-Q and the teacher's ratings of the child's overall adjustment derived from the CBC-Q. Results for this hypothesis are feund in row four of Table 5. The hypotheses was strongly sup- ported only for the distribution of boys and the overall distribution. Although the results for the distribution of girls were in the pre- dicted direction, they did not reach significance. To summarize, Hypothesis 1 was supported strongly for the dis- tribution of boys and the overall distribution, but was not supported for the distribution of girls. Hypothesis 2 Hypothesis 2 concerned the relationship between congruence of parents' perceptions of their child and three measures of their Child's adjustment derived from their CBC-Q ratings. The correlations for Hypothesis 2a, 2b, and 2c are found in Table 6, with their correspond- ing probabilities. Separate results are given for the distribution of boys, the distribution of girls, and the overall distribution. Sig- nificant correlations are marked with an asterisk (*). It was predicted in Hypothesis 2a that there would be a signif- icant negative correlation between correlations of parents' ratings of their child on the CBC-Q and their ratings of their child on the Clinic items of the CBC-Q. Results for this hypothesis are found in rows one and two of Table 6. This hypothesis was strongly supported for the 46 Table 6. Correlations and corresponding probabilities for hypotheses relating congruence in parents' perceptions of their child to parents' ratings of their child's adjustment (Hypotheses 2a, 2b, and 2c). Correlations of parents' ratings of their child on the CBC-Q Boys Girls Total (N = 63) (N = 47) (N = 110) Clinic score by -.5448* -.3079* -.5003* husbands (CBC-Q) (p_< .0005) (p_< .02) (p_< .0005) Clinic score by -.3523* -.5676* -.4483* wives (CBC-Q) (p_< .002) (p_< .0005) (E.< .0005) Nonclinic score by .6154* .5185* .6034* husbands (CBC-Q) (p_ < .0005) (p < .0005) (E < .0005) Nonclinic score by .4231* .5856* .4991* wives (CBC-Q) (p_< .0005) (p_< .0005) (p_< .0005) Overall adjustment score by husbands .7272* .5792* .6989* (CBC-Q) (p_< .0005) (p_< .0005) (p_< .0005) Overall adjustment score by wives .4925* .7528* .6046* (CBC-Q) (p_< .0005) (p_< .0005) (E.< .0005) 47 distribution of boys, the distribution of girls, and the overall dis- tribution. It was predicted in Hypothesis 2b that there would be a signifi- cant positive correlation between correlations of parents' ratings of their child on the CBC-Q and their ratings of their child on the Non- clinic items of the CBC-Q. Results for this hypothesis are found in rows three and four of Table 6. This hypothesis was strongly supported fer boys, girls, and the overall distribution. It was predicted in Hypthesis 2c that there would be a signifi- cant positive correlation between correlatibns of parents' ratings of their child on the CBC-Q and their ratings of their child's overall adjustment derived from the CBC-Q. Results for this hypothesis are feund in rows five and six of Table 3. The hypothesis was strongly supported fbr the distribution of boys, the distribution of girls, and the overall distribution. Hypothesis 3 Hypothesis 3 concerned the relationship between parents' marital adjustment and four measures of their child's adjustment as rated by the teacher. The correlations for Hypotheses 3a, 3b, 3c, and 3d are f0und in Table 7, with their corresponding probabilities. Separate results are given for the distribution of boys, the distribution of girls, and the overall distribution. Significant correlations are marked with an asterisk (f). It was predicted in Hypothesis 38 that there would be a signif- icant negative correlation between parents' Locke-Hallace scores and their child's total BRP score as rated by the teacher. Results for 48 Table 7. Correlations and corresponding probabilities for hypotheses relating parents' marital adjustment to their Child's adjustment as rated by the teacher (Hypotheses 3a, 3b, 3c, and 3d . Husbands' Hives' Locke-Hallace Locke-Hallace Scores Scores Total BRP scores Boys: .0815 (p_< .26) .0405 Sp_< .38 (teacher) Girls: .2844 (p_< .03)* .0770 p_< .30 Total: .0533 (p_< .29) .0020 (p_< .49) . Clinic score by Boys: -.0803 (e < .27) -.0728 E < .29) teacher (CBC—Q) Girls: .0317 (p_< .42) -.l759 p_< .12) Total: .0511 (p_< .30) -.1041 (p_< .14) Nonclinic score by Boys: .0618 (p_< .32) -.Ol64 (p_< .45) teacher (CBC-Q) Girls: -.0997 (p_< .25) -.0005 (p_< .50) Total: .0021 (p_< .49) -.0091 (p_< .46) Overall adjustment Boys: .0806 (p < .27) .0311 (p < .40) score by teacher Girls: -.0849 (p_< .29) .1021 (p_< .25) (CBC-Q) Total: .0304 (p_< .76) .0538 (p_< .29) this hypothesis are found in row one of Table 7. The hypothesis was not supported. Only the correlations for boys were in the predicted direction, and they did not reach significance. The one significant correlation found, between BRP scores for girls and husbands' Locke- Hallace scores, was in the opposite direciton from that predicted. It was predicted in Hypothesis 3b that there would be a signif— icant negative correlation between parents' Locke-Hallace scores and the teacher's ratings of the child on the Clinic items of the CBC-Q. Results for Hypothesis 3b are found in row two of Table 7. The hypothesis was not supported, although five of the six correlations were in the predicted direction. It was predicted in Hypothesis Be that there would be a 49 significant positive correlation between parents' Locke-Hallace scores and the teacher's ratings of their child on the Nonclinic items of the CBC-Q. Results for Hypothesis 3c are found in row three of Table 7. The hypothesis was not supported. Although four of the six correla- tions were in the opposite direction from that predicted, they were very low, which suggests that there was little relationship between the variables. It was predicted in Hypothesis 3d that there would be a signif- icant positive correlation between parents' Locke-Hallace scores and their child's overall adjustment score as derived from the teacher's CBC-Q ratings. Results of Hypothesis 3d are found in row four of Table 7. The hypothesis was not supported, although five of the six correlations were in the predicted direction. To summarize, Hypothesis 3 was not supported for the distribution of boys, the distribution of girls, or the overall distribution. Hypothesis 4 Hypothesis 4 concerned the relationship between parents' marital adjustment and three measures of their child's adjustment derived from their CBC-Q ratings. The correlations for Hypotheses 4a, 4b, and 4c are found in Table 8, with their corresponding probabilities. Separate results are given fer the distribution of boys, the distribution of girls, and the overall distribution. Significant correlations are marked with an asterisk (*). It was predicted in Hypothesis 4a that there would be a signif- icant negative correlation between parents' Locke-Hallace scores and their ratings of their child on the Clinic items of the CBC-Q. Results Table 8. 50 Correlations and corresponding probabilities for hypotheses relating parents' marital adjustment to their ratings of their child's adjustment (Hypotheses 4a, 4b, and 4c). Husbands' Hives' Locke-Hallace - Locke-Hallace Scores Scores Clinic score by Boys: .1150 Ep_< .19) -.0230 (p_< .43) husbands (CBC-Q) Girls: .0264 p_< .43) -.0519 p_< .37) Total: .0547 (p.< .29) -.0257 p_< .40) Clinic score by Boys: .0738 (p.< .28) -.2051 (p.< .05)* wives (CBC-Q) Girls: .1894 (p_<..lO) -.Ol69 p_< .46) Total: .0048 (p_< .48) -.1338 pfi< .08) Nonclinic score by Boys: .3180 (p.< .006)* .1906 (p_< .07) husbands (CBC-Q) Girls: .2259 (p_< .063) -.0885 (p.< .28) Total: .1352 (p_< .08) .0790 (p_< .21) Nonclinic score by Boys: .0150 (p_< .45) .1269 (p_< .16) wives (CBC-Q) Girls: .0711 (p_< .32) .1177 (p_< .22) Total: .0427 (e < .33) .1190 (e < .11) Overall adjustment Boys: .1491 (g < .12) .1419 (p_ < .13) score by husbands Girls: .1657 p_< .13) -.O425 (p.< .39) (CBC-Q) Total: .0645 p_< .25) '.0695 (p_< .24) Overall adjustment Boys: 0530 (p_< .34) .2054 (p_< .05)* score wives Girls: 0532 ( < .36) .0944 (p_< .26) (CBC-Q) Total: 0262 (p_< .39) .1581 (E.< .05)* 51 for this hypothesis are found in rows one and two of Table 8. The hypothesis was supported only for the correlation of wives' Locke- Hallace scores with Clinic scores by wife, for the distribution of boys. However, seven of the remaining eleven correlations were in the predicted direction. It was predicted in Hypothesis 4b that there would be a sig- nificant positive correlation between parents' Locke-Hallace scores and their ratings of their child on the Nonclinic items of the CBC-Q. Results fer this hypothesis are found in rows three and four of Table 8. The hypothesis was supported only for the distribution of boys, and only for the correlation of husbands' Locke-Hallace scores with their ratings of their child on the Nonclinic items of the CBC-Q. However, ten of the twelve correlations were in the predicted dir- ection. Only the correlations for the distribution of girls, of par- ents' Locke-Hallace scores with husbands' ratings of their children ' on the Nonclinic items of the CBC-Q, were in the opposite direction from that predicted. It was predicted in Hypothesis 4c that there would be a sig- nificant positive correlation between parents' Locke-Hallace scores and their ratings of their child's overall adjustment, derived from the CBC-Q. Results f0r this hypothesis are found in rows five and six of Table 8. The hypothesis was supported for the distribution of boys and the overall distribution, for the correlation of wives' ' Locke-Hallace scores with their ratings of their child's overall adjustment. For the distribution of girls, three of the four cor- relations were in the opposite direction from that predicted. 52 Hypothesis 5 Hypothesis 5 concerned the relationship between Congruence of parents' self-perceptions and perceptions by mate and four measures of their child's adjustment as rated by the teacher. The correlations fer Hypotheses 5a, Sb, 5c, and 5d are found in Table 9, with their corresponding probabilities. Separate results are given for the dis- tribution of boys, the distribution of girls, and the overall distri- bution. I Table 9. Correlations and corresponding probabilities for hypotheses relating congruence of parents' self-perceptions and per- ceptions by mate to their child's adjustment as rated by the teacher (Hypotheses 5a, 5b, 5c, and 5d). H and H-H H and H-H. Correlations Correlations on ICL on ICL Total BRP scores Boys: -.1242 Ep_< .17) -.2572 p_< .02)* (teacher) Girls: .0841 p_< .29) .0074 p_< .48) Total: -.O3l7 (p_< .37) -.l3l6 (p_< .09) Clinic score by Boys: -.Il39 Ep_< .19) -.IO35 (p_< .21) teacher (CBC-Q) Girls: -.0502 p_< .37) -.2590 (p_< .04)* Total: -.1016 (11‘ .15) -.l821 (P. < .03)* Nonclinic score by Boys: .1373 (p_< .l4) .2062 (p_< .05)* teacher (CBC-Q) Girls: .0093 (p_< .48) .1360 (p_< .18)* Total: .0791 (E_< .21) .1584 (E.< .05)* Overall adjustment Boys: .1431 (e < .13) .1772 (p_ < .08) score by teacher Girls: .0355 (p_< .41) ' .2414 (p_< .05)* Total: .1061 (11 < .15) .2002 (p< .02)* It was predicted in Hypothesis 5a that there would be a signif- icant negative correlation between correlations of parents' self- perceptions and perceptions by mate on the ICL and their child's 53 total BRP score as rated by the teacher. Results for this hypothesis are found in row one of Table 9. Although all of the correlations for the distribution of boys and the overall distribution were in the pre- dicted direction, only one reached significance. Correlations for the distribution of girls were in the opposite direction from that pre- dicted. It was predicted in Hypothesis 5b that there would be a signif- icant negative correlation between correlations of parents' self- perceptions and perceptions by mate on the ICL and the teacher's ratings of their child on the Clinic items of the CBC-Q. Results for this hypothesis are found in row two of Table 9. Although all of the correlations were in the predicted direction, only two reached signif- icance. The hypothesis was supported only for H and H-H correlations on the ICL with Clinic scores as rated by teachers, and only for the distribution of girls and the overall dirstribution. It was predicted in Hypothesis 5c that there would be a signif- icant positive correlation between correlation of parents' self- perceptions and perceptions by mate on the ICL and the teacher's ratings of their Child on the Nonclinic items of the CBC—Q. Results for this hypothesis are fOund in row three of Table 9. Although all of the correlations were in the predicted direction, only two reached significance. The hypothesis was supported only fer H and H-H cor- relations on the ICL with Nonclinic scores as rated by teachers, and only for the distribution of boys and the overall distribution. It was predicted in Hypothesis 5d that there would be a signif- icant positive correlation between correlations of parents' self- perceptions and perceptions by mate on the ICL and the teacher's 54 ratings of the child's overall adjustment as derived from the CBC-Q. Results fer this hypothesis are found in row four of Table 9. Al- though all of the correlations are in the predicted direction, only two reached significance. The hypothesis was supported for the cor- relations between H and H-H correlations on the ICL and overall adjust- ment scores on the CBC-Q as rated by teachers, for the distribution of girls and the overall distribution only. To summarize, Hypothesis 5 was not supported for the correla- tions between H and H-H correlations on the ICL and the four measures of child adjustment as rated by the teacher. However, there was mod- erate support for Hypothesis 5 for the correlations between H and H-H correlations on the ICL and the feur measures of child adjustment as rated by the teacher. Hypothesis 6 Hypothesis 6 concerned the relationship between congruence of parents' self-perceptions and perceptions by mate and three measures of their child's adjustment derived from their CBC-Q ratings. The correlations for Hypotheses 6a, 6b, and 6c are found in Table IO, with their corresponding probabilities. Separate results are given for the distribution of boys, the distribution of girls, and the overall distribution. Significant correlations are marked with an asterisk (*). It was predicted in Hypothesis 6a that there would be a signif- icant negative correlation between correlations of parents' self- perceptions and perceptions by mate on the ICL and their ratings of their child on the Clinic items of the CBC-Q. Results for this hypothe- sis are found in rows one and two f0 Table 10. For the distribution of boys and the overall distribution, all correlations were in the Table 10. ceptions by mate 55 Correlations and corresponding probabilities for hypotheses relating congruence of parents' self-perceptions and per- to their ratings of their child's adjust- ment (Hypotheses 6a, 6b, and 6c). H and H-H H and H-H Correlations Correlations on ICL 0n ICL Clinic score by Boys: - .2842 (p_<.01)* - I474 (p_< .12) husbands (CBC-Q) Girls: -.1002 (p_< .25) -.O436 (p_< .39) Total: -.2348 (p < 0.07)* - 1614 (p_ < .05)* Clinic score by Boys: -.3245 (p_< .005)* -.4403 (p.< .0005)* wives (CBC-Q) Girls: .2285 (p_< .06) .0693 (p_< .32) ’ Total: -.1271 (E_< .09) - 2765 (E_< .002)* Nonclinic score by Boys: .l230 (p.< .17) .2316 (p_< .03)* husbands (CBC-Q) Girls: .2497 (p.< .05)* .1483 (p_< .16) Total: .1916 (p_ < .02)* .2413 (p_ < .006)* Nonclinic score by Boys: -.0342 (p_< .40) .1728 (p_< .09) wives (CBC-Q) Girls: -.0581 (E.< .35) .2901 (p_< .02)* Total: -.0304 (p_ < .38) .2427 (p_ < .005)* Overall adjustment Boys: .2456 (p_< .03)* .2407 (p_< .03)* score by husbands Girls: .2530 (p.< .04)* .I417 (p_< .17) (CBC-Q) Total: .2638 (E.< .003)* .2581 (p_< .003)* Overall adjustment Boys: .1650 (p_< .10) .3732 (p_< .OOI)* score by wives Girls: -.l707 (p_< .13) .1796 (p_< .ll) (CBC-Q) Total: .0498 (E_< .30) .3265 (p.< .0005)* 56 predicted direction, with three of the f0ur reaching significance for each distribution. None of the correlations for the distribution of girls reached significance. In addition, both correlations of H and H-H correlations and H and H-H correlations on the ICL with Clinic scores by wives were in the opposite direction from that predicted, fer the distribution of girls. It was predicted in Hypothesis 6b that there would be a signif- icant positive correlation between correlations of parents' self- perceptions and perceptions by mate and their ratings of their child on the Nonclinic items of the CBC-Q. Results for this hypothesis are found in rows three and four of Table IO. All of the correlations of H and H-H correlations and H and H-H correlations on the ICL with Non- clinic scores by husbands were in the predicted direction, with four of the six reaching significance. For H and H-H correlations with Nonclinic scores by wives, the correlations for all three distribu- tions were in the opposite direction from that predicted. For H and H-H correlations with Nonclinic scores by wives, the correlations for all three distributions were in the predicted direction, with two of the three reaching significance. 4 It was predicted in Hypothesis 6C that there would be a signif- icant positive correlation between correlations of parents' self- perceptions and perceptions by mate on the ICL and their ratings of their child's overall adjustment as derived from the CBC-Q. Results for this hypothesis are fbund in rows five and six of Table 10. All but one of the correlations were in the predicted direction. The hypothesis was strongly supported for correlations of H and H—H cor- relations and H and H-H correlations on the ICL with overall adjustment 57 score by husbands, and for correlations of H and H-H correlations on the ICL with overall adjustment score by wives. The hypothesis was not supported for correlations of H and H-H correlations on the ICL with overall adjustment score by wives. To summarize, Hypothesis 6 was moderately supported for the dis- tribution of boys and the overall distribution. Eight of the twelve correlations for the distribution of boys and nine of the twelve cor- relations for the overall distribution reached significance. There was only slight support-for Hypothesis 6 for the distribution of girls; only three of the twelve correlations reached significance. The hypothesis tended to be supported most strongly for correlations between H and H-H correlations on the ICL and husbands' ratings of child adjustment, and fer correlations between H and H-H correlations on the ICL and wives' ratings of child adjustment. Hypotheses 7, 8, 9, 10, and II Hypothesis 7 concerned the relationship between congruence of parents' self-perceptions and perceptions by mate and congruence of parents' perceptions of their child. Specifically, it was predicted that there would be a significant positive correlation between correla- tions of parents' self-perceptions and perceptions by mate on the ICL and correlations of their ratings of their child on the CBC-Q. Re- sults fbr this hypothesis are found in Table II, column one, rows two and three. All of the correlations were in the predicted direction, and fiVe of the six were significant. The hypothesis was strongly supported fer the distribution of boys and the overall distribution, and moderately supported for the distribution of girls. 58 Table II. Correlations and corresponding probabilities for Hypotheses 7, 8, 9, IO, and II. Parents' CBC-Q Correlations Husbands' Locke-Hallace Locke-Hallace Hives' Scores Scores H and w Boys: .4492* .2124* .5238* Correlations (g < .0005) (g < .05) (p_ < .0005) on ICL Girls: .1116 .3056* .3088* (p_< .23) (p < .02) (p.< .017) ’ Total: .3368* .2511* .4268* (p_< .0005) (E_< .004) (p_< .0005) H and H-H Boys: .2273* .1563 .3620* Correlations (p,< .04) (p < .11) (p_< .002) ' on ICL Girls: .1376 .3157* .2175 (p_< .18) (p_< .02) (p,< .07) Total: .2091* .2175* .3025* (p_< .Ol) (p_< .Ol) (p_< .001) H and H-H Boys: .3542* .1574 .3612* Correlations (p_< .002) (p_< .ll) (p_< .002) on ICL ' Girls: .2775* .2319 .1682 m< M) m<.%) (2mgmn sumo new new: even mo «Pagan 9.: :2. 7 9.: ~32 _ .N a .m . .e .m .N .— lllll III-IIIIIL r llllllllll ._ _i IIIIIIIIII ._ 4|... 8 _ - .3: :2. fl 9.: :8. v .5 .o .m .e .m .~ ._ r IIIIIIIIII L r IIIIIIIIII L APPENDIX I Distribution Information for Variables of this Study Distribution Information for Variables of this Study Standard Mean Deviation Skewness Kurtosis Age (in Boys: 86.67 9.41 -.116 -.935 months) Girls: 84.98 10.47 -.01 -l.03 Total: 85.95 9.87 -.088 -.962 Total BRP Boys: 29.06 7.85 .293 -.380 Scores Girls: 30.28 7.41 -.499 -.085 Total: 29.58 7.65 -.024 -.377 Correlations Boys: .791 .206 -1.274 1.205 of parents Girls: .808 .211 -l.586 2.014 L-w scores Total: .798 .207 -l.406 1.532 Husbands' Boys: 106.91 21.43 -.895 .376 L-N scores Girls: 109.11 17 -.798 1.87 Total: 107.85 19.6 -.917 .957 Hives' L-N Boys: 113.08 21.93 -.944 .916 scores Girls: 112.23 18.86 -l.061 1.811 Total: 112.72 20.6 -.979 1.249 H and H Boys: .387 .46 -.279 -.553 Correlations Girls: .435 .154 -.032 .573 on ICL Total: .408 .151 -.137 .055 H and H-H Boys: .519 .168 -.869 .393 Correlations Girls: .552 .164 -.074 -.342 on ICL Total: .533 .166 -.546 .255 H and H¥N Boys: .492 .151 .045 .237 Correlations Girls: .562 .143 -.133 -.230 on ICL Total: .522 .151 -.054 -.017 Parents Boys: .628 .172 -.888 1.674 CBC-Q Cor- Girls: .707 .148 -.522 -.169 relations Total: .662 .167 -.807 1.366 Clinic score Boys: 11.08 7.67 1.006 1.032 by husbands Girls: 7.32 5.7 .645 -.385 Total: 9.47 7.12 1.05 1.36 Clinic score Boys: 10 7.26 .475 -.677 by wives Girls: 7.79 6.17 1.117 1.335 Total: 9.06 6.88 .729 -.127 Clinic score Boys: 10.02 10.1 1.332 1.083 by teacher Girls: 7.53 8.15 1.54 2.33 Total: 8.96 9.36 1.45 1.68 116 117 (Continued) Standard Mean Deviation Skewness Kurtosis Nonclinic Boys: 35.24 9.06 -.291 -1.08. score by Girls: 39.72 8.41 -1.203 2.245 husbands Total: 37.16 9.03 -.62 -.24 Nonclinic Boys: 36.76 8.52 -.249 -.566 score by Girls: 39.11 8.48 -.602 .794 wives Total: 37.76 8.55 -.389 -.086 Nonclinic Boys: 37.32 10.5 -.528 -.481 scores by Girls: 35.79 9.32 -.874 .515 teachers Total: 36.66 10 -.617 -.115 Overall' Boys: 24.16 13.41 -.284 -.782 adjustment Girls: 32.40 10.56 -.813 .665 by husbands Total: 27.68 12.89 -.554 -.401 Overall Boys: 26.76 12.52 -.025 -.837 adjustment Girls: 31.3 11.27 -l.77 4.65 by wives Total: 28.7 12.16 -.66 .426 Overall Boys: 27.3 18.12 -.867 .271 adjustment Girls: 28.26 14.0 -.958 .97 by teachers Total: 27.71 16.42 —.93 .65 BRP Boys: 3.89 1.93 .05 -1.279 Item 1 Girls: 3.79 1.76 .037 -1.111 Total: 3.85 1.85 .053 -1.204 BRP Boys: 4.4 1.68 -.272 -.879 Item 2 Girls: 4.19 1.83 -.048 -.792 Total: 4.31 1.74 -l.77 -.837 BRP Boys: 3.57 11.70 .174 -.962 Item 3 Girls: 3.75 1.70 .03 -1.227 Total: 3.65 1.70 .112 -1.084 BRP Boys: 3.13 1.57 .518 -.307 Item 4 Girsl: 3.64 1.89 .085 -l.059 Total: 3.35 1.73 .349 -.733 BRP Boys: 3.35 1.95 .602 -1.08 Item 5 Girls: 3.79 1.82 .1 -.962 Total: 3.54 1.9 .383 l-1.l31 BRP Boys: 3.52 1.8 .45 -.776 Item 6 Girls: 3.96 1.72 .118 -l.12 Total: 3.71 1.77 .297 -.965 118 Standard Mean Deviation Skewness Kurtosis BRP Boys: 3.52 1.85 .441 -.879 Item 7 Girls: 3.81 1.78 .034 -l.045 Total: 3.65 1.82 .268 -.995 BRP Boys: 3.48 1.84 .375 ‘ -.906 Item 8 Girls: 3.51 1.72 .476 -.684 Total: 3.49 1.78 .412 -.815 BRP Item 9 Boys: 4.48 1.69 -.512 -.73 (Compe- Girls: 4.19 1.92 -.331 -.934 tence) Total: 4.36 1.79 -.446 -.803 CBC-Q Correla- Boys: .451 .222 .096 -.72 tions: Husbands Girls: .53 .246 -.722 .422 and Teachers Total: .485 .235 -.253 -.374 CBC-Q Correla- Boys: .471 . .238 -.222 -.658 tions: Hives Girls: .545 .21 -.785 .741 and Teachers Total: .503 .229 -.455 -.293 TQTE UN V. HICHIGRN S 11111111(IIWIIWIWI 312931 I LIBRARIES 111111”11111111111111” 60714 (36322