The effects of pubertal timing on body image,school behavior,and deviance

Variation in the timing of pubertal maturation may result in behavioral differences among early, mid-, and late maturers. Using data from the National Health Examination Survey, a national probability sample of children and youth aged 12–17, we investigated the relationships between maturational timing and body image, school behavior, and deviance. In terms of body image, the early maturing boys were the group most satisfied with height and weight. The early maturing girls were most dissatisfied with weight, with 69% wishing to be thinner. This great dissatisfaction with weight reported by early maturing girls is probably not an affect of early maturation, but a concomitant of maturation in general. The majority of girls became dissatisfied with their weight as they matured, and females from the higher social groups were more likely to want to be thinner than females from lower groups. Thus, a normal developmental process is being viewed negatively by females and positively by males. Male early maturers more often had deviant behavior, but there were no consistent findings for girls. There was no effect of maturational timing on teacher reports of school absence, adjustment, popularity, need for discipline, or grade repetition.This research was supported by the Stanford Center for the Study of Youth Development and by the Robert Wood Johnson Foundation.Received M.D. from The Medical College of Pennsylvania.Received Ph.D. in anthropology from Stanford University.Received Ph.D. in sociology from the University of Chicago.Received M.D. from Columbia College of Physicians and Surgeons.Received Ph.D. in psychology from Harvard University.     

Physical attractiveness as a mediator of the impact of early pubertal changes for girls

The present study examined the effects of early pubertal development and physical attractiveness on the popularity, body image, and self-esteem of over 200 sixth-grade girls. Two rival hypotheses were explored. The first suggests that physically attractive girls, because of their more favorable social environment, will exhibit fewer psychosocial difficulties than unattractive girls during pubertal development. The second hypothesis argues that attractive girls will exhibit greater difficulty during pubertal transition because their self-image is more intimately connected with their physical appearance. Although there were no significant interactions between attractiveness and pubertal development for either popularity or body image, the second hypothesis was supported with respect to self-esteem. Specifically, developing attractive girls exhibited lower self-esteem than their unattractive counterparts. The results are discussed in terms of the relative vulnerability to bodily changes of girls differing in physical attractiveness.Support for this work was provided by NIMH Grant 2 R01 MH-30739 and the William T. Grant Foundation.Received his M.A. in clinical child psychology from the Ohio State University. Main interests are the psychological concomitants of physical appearance and developmental aspects of the body image.Received his Ph.D. in sociology from the University of Minnesota. Current interests include social relationships in adolescence, consequences and correlates of physical maturation, and effects of transitions in early adolescence.Received her Ph.D. in sociology from Columbia University. Current interests include gender differences in adolescence, the impact of transitions throughout the life span, and the psychosocial impact of organ donation.     

Transcultural comparisons of self-image in German and United States adolescents

Based on the Offer Selfimage Questionnaire (OSIQ), the selfimage of German and United States adolescents was compared. The German study was based on OSIQ protocols from 365 adolescents in West Berlin while the American sample comprised adolescents drawn from seven cities in the United States. With respect to three scales, United States adolescents report better adjustment than do the German adolescents. These scales were Mastery of the External World, Vocational and Educational Goals, and Superior Adjustment. In general, these two Western societies share more similarities than differences in the selfimages of their adolescents.Received M.D. and Ph.D in Psychology from the University of Hamburg, Germany. Research interests are high-risk studies, child psychiatric epidemiology, and adolescence.Received M.D. from the University of Chicago. Major interests are concepts of mental health and the developmental psychology of adolescence.Received J. D. from the University of Chicago School of Law; received Ph.D. in Human Development from the University of Chicago. Research interests are adolescence and delinquency.Received Ph.D. in Psychology from the University of Chicago. Major interests are psychotherapy research and adolescence.     

A self-image questionnaire for young adolescents (SIQYA): Reliability and validity studies

This article describes a questionnaire measure of self-image designed for young adolescents. It represents a downward extension of the Offer Self-Image Questionnaire and utilizes nine scales from that instrument: Emotional Tone, Impulse Control, Body Image, Peer Relationships, Family Relationships, Mastery and Coping, Vocational/Educational Goals, Psychopathology, and Superior Adjustment. This 98-item questionnaire elicits responses on a 6-point Likert-type scale. The alpha coefficients for each scale are high, indicating a high degree of internal consistency among the items. The validity of this instrument is examined through factor analyses and through the association of these scales with other measures of self-image. The results suggest that this questionnaire provides a useful way to assess self-image among young adolescents.This research was supported by Grant MH 30252/38142 to Anne Petersen.Received Ph.D. from the University of Chicago in 1973. Research interest is biopsychosocial development in adolescence, with a focus on sex differences.Research interests are vocational development and contextual influences on development in early adolescence.Research interests are the influences of adolescent and parent development on family relationships.Received M.D. from the University of Chicago. Research interests are the psychology and psychopathology of adolescents.Current research interests are social cognition and peer relationships during early adolescence.     

Introduction: The changing life space of early adolescence

The socialization of children and adolescents is shaped by how they spend their time. This special issue maps the daily experience of white American 5th–9th graders, describing both the quantities of time they spend in different contexts and the affective states associated with these contexts. Each paper examines a separate segment of daily activity (e.g., schoolwork, talking, sports), employing data from a common study in which 401 participants provided experience sampling reports at random times during their daily lives when cued by a pager. The findings show substantial age and sex differences, indicating significant changes in the daily life space of girls and boys between preadolescence and early adolescence. Implication for socialization to healthy adult adjustment in love, work, and play are discussed in each paper.Received Ph.D. from Committee on Human Development, University of Chicago, Research focuses on daily experience associated with health and psychopathology in adolescence and across the life span.Received Ph.D. from Committee on Human Development, University of Chicago. Current research interests: pubertal development, precursors of eating disorders, and the effects of maternal employment on young adolescents.     

Measurement of the psychological well-being of adolescents: The psychometric properties and assessment procedures of the how I feel

The assessment procedures and psychometric properties of the How I Feel (HIF), an instrument used to assess psychological well-being in a population of Black adolescents are described. The audiovisual mode of presentation obviates problems related to reading skill; in addition, it standardizes the administration of the instrument. The How I Feel appears to measure reliably and validly several multi-item constructs representing psychological well-being. These constructs relate to other instruments and constructs in meaningful and interesting ways. A major result of our validity studies is that there appear to be two major components of psychological well-being, psychopathology and self-esteem.Research presented in this paper is from the Social Psychiatry Study Center, Department of Psychiatry, University of Chicago and was supported by a grant from the National Institute on Drug Abuse (DA-00787).Received her Ph.D. from the University of Chicago. Main research interests are biopsychosocial development in adolescence, especially for girls, and applications of statistical and psychometric methods to problems in longitudinal research.Received his M.D. from University of Maryland. His psychiatric residency and research training was at Yale University and National Institute of Mental Health. Main research interests are long-term studies of social adaptation, psychological well-being (including psychopathology), and social structure and processes of the family and other social fields such as the psychiatric ward.     

Cognitive Vulnerabilities Amplify the Effect of Early Pubertal Timing on Interpersonal Stress Generation During Adolescence

Early pubertal timing has been found to confer risk for the occurrence of interpersonal stressful events during adolescence. However, pre-existing vulnerabilities may exacerbate the effects of early pubertal timing on the occurrence of stressors. Thus, the current study prospectively examined whether cognitive vulnerabilities amplified the effects of early pubertal timing on interpersonal stress generation. In a diverse sample of 310 adolescents (M age = 12.83 years, 55 % female; 53 % African American), early pubertal timing predicted higher levels of interpersonal dependent events among adolescents with more negative cognitive style and rumination, but not among adolescents with lower levels of these cognitive vulnerabilities. These findings suggest that cognitive vulnerabilities may heighten the risk of generating interpersonal stress for adolescents who undergo early pubertal maturation, which may subsequently place adolescents at greater risk for the development of psychopathology.     

Biological and ecological factors in the expression of adolescent self-esteem

The relationship between self-esteem and pubertal maturation is examined based on results of the first two years of a six-year naturalistic longitudinal study of the ecology of self-esteem during adolescence. A random sample of 40 seventh-grade adolescents of both sexes from all socioeconomic classes was selected from a local junior high school. Four measures of pubertal maturation level were obtained for participants during each year of the study. Self-esteem was assessed by means of a repeated measures self-report technique. Each adolescent carried a beeper paging device and responded in writing to beeps that occurred on a random schedule eight times daily. Males were significantly higher in self-esteem than females. The developmental pattern for the relationship between self-esteem and pubertal maturation differed between boys and girls. Three ecological contextual variables were explored: setting, activity, and participants. Physical setting was significantly associated with level of self-esteem. Activities and participants present were not significantly predictive of self-esteem. Results are discussed in terms of the onset and timing of pubertal maturation and its influence on self-esteem.This paper is a revised version of a paper presented at the annual meeting of the American Psychological Association, Montreal, September 1–5, 1980. The research project is supported by a grant from the Spencer Foundation, Chicago, to the second author.University of Notre Dame. Received her Ph.D. from Cornell University. Major interest are adolescent and human lifespan development, and divergent thinking.Received his Ph.D. from the University of Chicago. Major interests are adolescent development, self-esteem, and ethology.     

Hungarian adolescents' self-concept

Hungarian and United States adolescents' self-image was studied using the Offer Self-Image Questionnaire (OSIQ). In Hungary, 1,163 younger and older male and female adolescents were studied using a Hungarian translation of the OSIQ. Analyses of endorsement patterns of OSIQ items showed that Hungarian and American adolescents endorsed many items in the same way. Similarities in endorsement patterns were much more common between the two countries than were differences. Analyses of OSIQ scales showed that for most scales younger Hungarian adolescents reported better adjustment than younger American adolescents. Differences were not as great or reversed in the older age groups. Implications for cross-cultural studies of adolescent self-image were drawn based on these results.Received M. D. from the Semmelweis Medical university in Budapest. Research interest is complex somato-mental health care of adolescents.Received M. D. from the University of Chicago. Major interests are concepts of mental health and the developmental psychology of adolescence.Director, Forensic Psychology, Michael Reese Hospital and Medical Center. Received J. D. from the University of Chicago School of Law; received Ph. D. in human development from the University of Chicago. Research interests are adolescence and delinquency.Professor of Psychology, Northwestern University. Received Ph. D. in psychology from the University of Chicago. Major interests are psychotherapy research and adolescence.     

Family contexts of pubertal timing

This study examined the influence of pubertal timing upon family interactions in normal and psychiatric adolescent samples. An important feature of our approach is its emphasis upon micro-analysis of family behaviors (individual speeches) and family processes (theoretically specified speech pairings). Rather than assume that global family patterns (e.g., power) shift in response to pubertal changes, we follow how types of speeches and speech sequences are associated with different pubertal timing. Using the previously constructed family coding system, the Constrainig and Enabling Coding System, we found that on-time adolescents and their parents differed from both off-time groups (early or late). These results are discussed in terms of current implications and suggestions for future research.An earlier version of this paper was presented at the SRCD Study Group on Timing of Maturation, October, 1983, at the Education Testing Service, Princeton, NJ. This research was supported by NICHD Grant 1 R01 HD18684-01, and an NIMH Research Scientist Award (Dr. Hauser).Received his Ph.D. from Harvard University. Research interest is adolescent development within the family and impact of chronic illness on adolescent development and family interaction.Received her B.A. from Wellesley College. Research interests are in humor and attractiveness.Received his M.A. from Boston University. Research interests are in methodology and statistics.Received Ed.D. from Harvard University. Research interests are in adolescent development within the family, and family coping with stress.Received M.D. from University of Chicago. Research interests are in psychosocial aspects of diabetes.Received Ed.D. from Harvard University. Research interests are in developmental psychopathology, and moral and ego development.Received Ph.D. from Ohio State University. Research interests are in assessment of ego development and family systems.Received Ph.D. from University of Miami. Research interests are in family studies and adolescent development.     

Longitudinal Association Between Parenting Practices and Early Sexual Risk Behaviors among Urban African American Adolescents: The Moderating Role of Gender

A sample of 274 African American families, living in impoverished neighborhoods with high HIV rates, participated in a longitudinal study of adolescent sexual development when children were in the 4th or 5th grade. Self-report and observational measures of parental warmth and parental behavioral control were collected from adolescents and parents at Time 1, and youth reported if they had initiated intercourse at Times 1 and 2. Regression analyses suggested that gender moderated associations between parental behavioral control and engagement in adolescent sexual behaviors. More generally, findings suggested that boys reared in low control/high warmth (i.e., permissive) homes and girls reared in high control/low warmth (i.e., authoritarian) homes were particularly at risk for early sexual behaviors. Clinical implications and directions for the future research are discussed.Doctoral Candidate in Clinical Psychology at Loyola University Chicago. Received her B.S. in Psychology and African & African American Studies from Duke University and her M.A. in Clinical Psychology from Loyola University Chicago. Her major research interests include the role of family and mental health factors in HIV risk exposure among urban African American adolescents.Professor, Department of Psychology, Loyola University Chicago. Received his Ph.D. in 1987 from Virginia Commonwealth University. His major research interests are family relations during adolescence, physical disabilities, pediatric psychology, developmental psychopathology, and statistical applications in psychologyAssociate Professor of Psychology in the Department of Psychiatry, Institute for Juvenile Research, University of Illinois, Chicago. Received her PhD in Child Psychology from the Institute of Child Development at the University of Minnesota in 1987. Her current research interests include developmental transitions during adolescence, as well as from pre-school to middle childhood, among typically developing children as well as children with special needs     

Adolescent pubertal status and affective family relationships: A multivariate assessment

This research attempted to more adequately dimensionalize the study of affective family relations during adolescent pubertal maturation by employing a multivariate approach. Fifty-one families responded to an assessment battery consisting of a series of questionnaires designed to measure affective relations between parents and their adolescent children. Families were classified into prepubertal, transpubertal, or postpubertal groups according to their modal response on pubertal status criteria. A stepwise discriminant analysis was performed to identify measures that maximize between groups differences. Based on this analysis, a correct classification rate of 70% was achieved for all families. The two significant discriminant functions that emerged indicate that differences in family relationships are due to the transformation of mother-adolescent and father-adolescent relations. The pattern of results replicate those of other researchers and provide elaborations concerning the affective nature of family relations during adolescence.Received his Ph.D. in Life-span Developmental Psychology from West Virginia University in 1984. Research interests in dialectical models of adolescent development.Received his Ph.D. from the University of Arkansas in 1979. Research interests in decision-making in adolescence and adulthood.     

A self-report measure of pubertal status: Reliability,validity, and initial norms

Puberty is a central process in the complex set of changes that constitutes the transition from childhood to adolescence. Research on the role of pubertal change in this transition has been impeded by the difficulty of assessing puberty in ways acceptable to young adolescents and others involved. Addressing this problem, this paper describes and presents norms for a selfreport measure of pubertal status. The measure was used twice annually over a period of three years in a longitudinal study of 335 young adolescent boys and girls. Data on a longitudinal subsample of 253 subjects are reported. The scale shows good reliability, as indicated by coefficient alpha. In addition, several sources of data suggest that these reports are valid. The availability of such a measure is important for studies, such as those based in schools, in which more direct measures of puberty may not be possible.This research was supported by a grant to A.C.P. from the National Institutes of Mental Health (MH 38052/38142).She earned the Ph.D. in Measurement, Evaluation, and Statistical Analysis at the University of Chicago. Her research interests focus on biopsychosocial development in adolescence.She earned the Ph.D. in Human Development at the University of Chicago. Her research interests focus on social development and gender roles.She earned the Ph.D. in Human Development at the University of Chicago. Her research interests focus on adolescent development and gender issues.His research interests focus on parent-adolescent relationships.     

The sporting life: Athletic activities during early adolescence

The decline in sports participation that begins in early adolescence has been well documented, and there has been considerable controversy regarding the reasons for this attrition. The present study addressed the attrition process by focusing on the subjective experience of sports as a function of grade, gender, and sport context. Following the procedures of the Experience Sampling Method, 401 5th–9th-grade boys and girls carried electronic pagers, similar to those worn by doctors, for one week, and filled out self-report forms on their activities and subjective states in response to signals received at random times. Older respondents spent less time in sports than their younger peers. This age difference was due primarily to a decline in informal sports participation, with less pronounced attrition from organized sports. Our findings suggest that the reasons for attrition from sport may be context specific. While informal sports were experienced more positively than gym class or organized sports, perceptions of skill were lowest during informal sports and declined with age. It seems youngsters stop participating in organized sports because these activities are less enjoyable to them, while attrition from informal sports is more performance based. Boys spent more time in sports than girls, and this difference was based primarily upon significant gender differences in informal sports participation. In spite of their differential rates of participation, boys and girls reported similar levels of affect, arousal, and skill during sports.This research was supported by NIMH grant number MH38324, Stress in Daily Life During Early Adolescence, awarded to Reed Larson, and was carried out while the first author was a Dissertation Fellow in the Department of Psychology at Loyola University of Chicago.Received Ph.D. in psychology from Loyola University of Chicago. Current research interests include athletic, involvement during early adolescence, body image, and adolescent mental health.Current research interests are stress and coping during early adolescence, and the adaptive use of leisure time.Received Ph.D. from the Committee on Human Development at the University of Chicago. Current research interests include pubertal development, the precursors of eating disorders, and the effects of maternal employment of young adolescents.     

Psychiatrists' concept of adolescent self-image

The objectives of the study were as follows: (1) to determine the accuracy of psychiatrists' concepts of the self-image of healthy and disturbed adolescents, and (2) to examine the effects of psychiatrists' gender, experience, and specialization on the accuracy of these concepts. Toward this end, eighty-four psychiatrists completed a modified Offer Self-image Questionnaire as they believed a mentally healthy 15-year-old of their gender would respond. These responses were compared to those of 140 nonpatient 15-year-olds. Sixty-one psychiatrists completed the questionnaire as they believed a disturbed 15-year-old of their gender would respond. These responses were compared to the responses of 70 psychiatric inpatient 15-year-olds of the same diagnosis. The results were that psychiatrists had a more accurate concept of normal adolescents' self-image than was found in 1981. However, their view of the self-image of disturbed adolescents was more psychopathological than the disturbed adolescents viewed themselves. Psychiatrists' gender, number of years postresidency, and type of training had no effect on the accuracy of the psychiatrists' concepts. The results demonstrate that while psychiatrists work with disturbed adolescents, they have a more accurate perception of the self-image of normal adolescents. This can effect the diagnosis and treatment of adolescent patients.Presented on May 24, 1994, at the annual meeting of the American Psychiatric Association, Philadelphia, Pennsylvania.Received M.D. from Rush Medical College. Research interest is in adolescent psychology.Received M.D. from University of Chicago. Research interests are in adolescent development psychology and psychopathology.Received Ph.D. from University of Chicago. Research interests are in psychotherapy research and adolescent psychology.Received Ph.D. in Clinical Psychology from Northwestern University. Research interests are in adolescence and child development.     

Cognitive mediators of ego functioning in adolescence

It has often been assumed that a relationship exists between higher levels of cognitive functioning, particularly formal operations, and mature ego functioning in adolescence. This research examined the relationships between ego functioning and two domains of operational thinking: social interpersonal reasoning and physical-mathematical reasoning in 139 high school seniors. Subjects were given two measures of physical-mathematical reasoning, two measures of interpersonal reasoning, and the Sentence Completion Test of ego functioning, as well as a measure of verbal intelligence. Results indicated significant differences between males and females in patterns of correlations as well as in patterns of relationships in a causal analysis. Ego functioning was predicted by interpersonal reasoning for females and by physical-mathematical reasoning and verbal intelligence for males.This research represents a portion of the doctoral dissertation completed by the senior author in 1981. The research was supported by a grant to A. Petersen from the Spencer Foundation.Received Ph.D. from University of Illinois at Chicago. Current interests are sex-related differences in adolescent ego development and psychosocial variables in adolescent chronic illness.Received Ph.D. from the University of Chicago. Current interest is biopsychosocial development in adolescence, primarily early adolescence.Research Affiliate, Laboratory for Study of Adolescence, Michael Reese Hospital. Received Ph.D. from University of Chicago. Current interest is sex-related differences in the psychological effects of puberty.Research Associate, Laboratory for the Study of Adolescence, Michael Reese Hospital; Program Associate, Health Program, MacArthur Foundation. Received Ph.D. from Syracuse University. Current interest is sex-related differences in socialization.     

Adolescent depression: Why more girls?

Although there has been evidence for some time of a sex difference in depression, relatively little research has examined the developmental process by which women come to be at greater risk than men for depression. In this paper, the developmental pattern of depressed affect is examined over early and middle adolescence, with a special focus on the patterns of boys as compared to girls. In addition, a developmental model for mental health in adolescence is tested for its power in explaining the emergence of gender differences in depression. Longitudinal data on 335 adolescents randomly selected from two school districts were used to test the hypotheses. Results revealed that girls are at risk for developing depressed affect by 12th grade because they experienced more challenges in early adolescence than did boys. The sex difference in depressed affect at 12th grade disappears once early adolescent challenges are considered.This research was supported in part by grants MH30252/38142 to A. Petersen. We gratefully acknowledge the contributions of study participants and staff. Portions of this material were presented in a symposium at the 1988 meeting of the Society for Research on Adolescence, and one at the 1989 meeting of the Society for Research in Child Development.Received Ph.D. from University of Chicago in 1973. Research interest in biopsychosocial development in adolescence, with a focus on sex differences in mental health.Research interests include adolescent mental health, and parent and peer relationships.Received Ph.D. in psychology from The Pennsylvania State University. Research interests include development of affective and conduct disorders, especially sex differences in these problems.     

The factor structure of the Reckless Behavior Questionnaire

In this study, the relative fits of three different factor-structure models of adolescent reckless behavior were examined using the Reckless Behavior Questionnaire (RBQ) with individual samples of college and high school students. Both one- and two-factor models were found to be satisfactory representations of the RBQ with both samples. In order to test the construct validity of the one- and two-factor models, relations between instruments generally associated with reckless behavior were examined by gender. Using the two-factor model, gender differences were found for both the college and high school samples; thus, it was determined to be the more parsimonious fit of the data given previous research supporting gender differences. Findings are discussed in terms of current conceptualizations of factor patterns of adolescent problem behavior and implications for future investigations.B.A. from Oberlin College, and M.A. and Ph.D. in Psychology from University of Virginia. Research interests are development of aggression and correlates of reckless behavior.B.A. from University of Akron and M.S. from University of Pittsburgh. Research interests are addictive behaviors and psychopathology in adolescence.Formerly Assistant Professor at Ogelthorpe University (1985–1988). Received B.S. from Michigan State University, and M.A. and Ph.D. from University of Virginia. Research interest is development of reckless behavior.Received B.A. from Yale University, and M.A. and Ph.D. from University of Virginia. Research interests are developmental psychopathology and adolescence.     

A differential survey of the problems of privileged and underprivileged adolescents

The problems of adjustment during the midadolescent years were examined in 1298 privileged and underprivileged adolescents from the Sydney Metropolitan Area, Australia. Girls reported more problems than boys in all areas of adjustment and underprivileged adolescents more problems than the privileged group; however, a significant interaction effect was also found. Whether high-income or low-income group, the overwhelming number of problems were associated with educational adjustment and were interpreted as the reaction to pressures to achieve imposed on adolescents of this age.Received her M.A. (Hons.) and Ph.D. from Macquarie University. Member of Australian Psychological Society. Principal research interests are child psychopathology and psychology of adolescence.Received his M.A.(Hons.) and Ph.D. from Sydney University. Fellow of Australian Psychological Society. Principal research interest is psychology of adolescence.     

Satisfaction with body image for early adolescent females: The impact of pubertal timing within different school environments

During adolescence dramatic physical changes take place which the individual must incorporate into his or her evolving body image. The impact of different school environments on this incorporation process is explored using data on 225 White females from a longitudinal study. Differences in physical characteristics between early, middle, and late developers were assessed each year. The effects of pubertal timing on satisfaction with body image dimensions and self-esteem were then explored for sixth- and seventh-graders within different school environments. Reference group theory was used to examine three alternative hypotheses. Early versus late onset of menarche had different effects on certain aspects of satisfaction with body image, depending on the school environment. Results support the strength of the cultural ideal of thinness for women, but no other hypothesis had consistent support. The findings indicated the need to consider a multiplicity of factors in relation to specific body image dimensions.This study has been funded by NIMH grant R01 MH-30739 and a grant from the William T. Grant Foundation. In addition, the work of the second author has been supported by a Research Development Award from the National Institute of Mental Health, #2 K02 MH-41688.Received his Ph.D. from the University of Minnesota. Current research interests include stress and coping in adolescence, the role of social relationships in problem behavior, and the effects of different school contexts.Received her Ph.D. from Columbia University. Research interests include the social and psychological effects of puberty and sex differences in adolescence.Received M. A. in clinical child psychology from Ohio State University. Main interests are developmental aspects of body-image and research on personality disorders.