Craig W. Thomas

Primary and Secondary Control Over Age‐Related Changes in Physical Appearance

Beliefs about appearance-related changes due to aging were used to test the effects of perceived control and secondary control (acceptance) in a sample of 412 young, early-middle-age, and late-middle-age college-educated adults. Mean difference in aging-related appearance control and hypotheses regarding the adaptiveness of primary and secondary control were examined. Primary control over aging-related appearance was lower in older adults and secondary control was higher. In addition, the results indicated support for the Primacy/Back-Up Model that primary perceived control is important at all levels of actual control. Those with stronger beliefs in their primary control were less distressed. Secondary control served a back-up function in that it was related to less distress only for those who had medium or lower beliefs in primary control. The implications of these findings, that primary control may be advantageous even in low-control circumstances, are discussed.

Caregivers of Stroke Patient Family Members: Behavioral and Attitudinal Indicators of Overprotective Care

Three models of the sources of overprotection in stroke patients were tested in a study of the behavioral and attitudinal concomitants of overprotective caregiving. Stroke patients and their family member caregivers were interviewed to assess feelings of overprotection, physical and mental functioning, and caregiving-related attitudes. The couples were then videotaped as they interacted on four tasks and the tapes were coded for specific behaviors and affect directed toward the patient. Strong support was found for the Resentment model that overprotection will be associated with an overcontrolling caregiving style, negative affect, and resentment toward the stroke patient. No support was found for the model that overprotection is related to overhelping. In addition, feeling overprotected was associated with patient dependency. Two conclusions are discussed: the variety of potential sources of overprotection found in the study and the importance of the emotional tone with which caregivers provide assistance.

Synthesis and Implications

We conclude this volume by synthesizing the body of findings resulting from the Fullerton Longitudinal Study with respect to numerous issues. Findings germane to the understanding of the developmental course of temperament and the significance of individual differences in youngsters temperament for their development in the domains of personality, behavioral adjustment, intelligence, and school functioning are first addressed. Individual differences in children’s temperament interface with their home and family contexts of development, yielding additional clues as to the mechanisms by which temperament and developmental outcomes may come to be related. In support of the parental report methodology, results from the FLS yielded strong evidence regarding the predictive and construct validity of parent reports of children’s temperament. The interaction of individual differences in temperament with contexts of development, subject variables, and developmental differences in other domains have only begun to be examined. However, findings from the FLS demonstrate that even aspects of temperament assessed as early as 1.5 years have long-term sequelae, both through direct main effects and also through interactions with developmental contexts, and that early differences in temperament are associated not only with developmental risk, but also can serve as assets that may facilitate or enhance not only the children’s context of development but also their developmental trajectory.

Temperament and Behavioral/Emotional Adjustment

The linkage between early temperament and behavior problems during middle childhood found by the NYLS group served as a catalyst for the study of temperament. Data based on the FLS and other longitudinal studies have not only replicated the findings of the NYLS researchers regarding the predictive power of early temperament with respect to behavioral adjustment, but also have extended downward the age at which temperament has been shown to foretell childhood behavior problems. This chapter details results from the FLS data regarding concurrent and predictive relations between temperament and behavior problems using multiple informants (parents, teachers, adolescent self-reports) and state-of-the-art behavior problem measures.

Temperament in the School Context

The Fullerton Longitudinal Study is somewhat unique in terms of methodology in comparison to much of the extant research on the relation of temperament to various aspects of children’s experiences in the school context. Much, if not most, prior research on temperament in the school context uses intact classrooms of students, often utilizing teacher reports of temperament based on their observations of the children’s behavior in the classroom environment. This arrangement provides numerous research advantages, including, for example, temperament ratings of children based in somewhat standard settings with somewhat consistent demands/stressors and expectations, temperament ratings by teachers who are experienced and knowledgeable observers of behavior who have had a wealth of opportunity to observe the range of behavioral styles present in most classrooms, and access to regularly-collected indicators of academic performance (grade reports, standardized tests, etc.). In contrast, the FLS is not restricted to a single, isolated classroom, school district, or region. The FLS sample is comprised of children attending a variety of types of schools (e.g., private non-sectarian, private with religious affiliation, public) across a broad geographic region spanning not only Southern California, where all study children were born, but, as the study progressed, increasingly diverse geographic locations. Indeed, as the study participants advanced through their formal education, they were living in states as varied as California, Oregon, Washington, Arizona, Texas, Ohio, and Minnesota, for example.

Methodology of the Fullerton Longitudinal Study and Data Analytic Strategy

The Fullerton Longitudinal Study (FLS) is a contemporary investigation of 130 infants and their families. Children were 1 year old at the initiation of the project; they were tested every 6 months from 1 year through 3.5 years and at yearly intervals from 5 through 17 years thus far, as displayed in Table 2.01. Thus, across the span of investigation covered in this book, participants visited our lab for 19 assessments and FLS researchers made 3 direct in-home assessments. Subject retention has been excellent, with 109 (84%) returning for assessment at age 17.

Temperament in the Family Context

In this chapter, we examine the extensive FLS longitudinal data related to the participants’ home and family environment, including family relationships. As we have previously reported that temperament in the FLS was not pervasively related to family socioeconomic status (A. W. Gottfried et al., 2003), we focus in this chapter on proximal variables. By proximal variables, we mean those that assess processes or specific aspects of the environment that may influence children’s development. These would include, for example, language stimulation, access to toys, games, and reading materials, opportunities for variety in stimulation, etc. Keeping in mind the bidirectional and transactional traditions within developmental science perspectives, we examined child temperament as a concomitant, as a predictor, as an outcome, and as a moderator of development. Five major issues will be addressed herein: 1 Are children’s temperament characteristics related to proximal variables in their home? Are there consistent or changing patterns of relation between child temperament and proximal home environment across development from infancy through childhood? In the FLS, homes of participants were visited during the infancy, preschool, and elementary years, and it was with these detailed and extensive data collected via the well-known HOME scales, based on both direct observation and interview methodology, that we first examined temperament-environment relations. 2 To what extent are specific dimensions of child temperament related to concurrent and cross-time assessments of the quality of family climate? Are the strengths and patterns of relation between child temperament and family climate constant across development? Assessments of the family environment were regularly collected using the Family Environment Scale. These data provide another measure of proximal environment variables that may influence children’s development, and the relation of child temperament to these variables across the childhood and adolescent years is assessed herein. 3 How do individual differences in children’s temperament relate to parents’ and children’s assessments of their relationship with each other? Parent-child relationships, which also may be considered proximal environmental variables, are considered as a separate category. During the middle childhood and/or adolescent period, study participants and parents rated the quality of their relationship with each other, and these assessments are examined in relation to parent and adolescent assessments of the adolescents’ temperament. 4 How might variations in children’s contexts influence the expression of their temperament? Which aspects of the proximal environment, if any, are predictive of children’s subsequent temperament? In this section, we examine the relation between earlier measures of home environment and subsequent measures of child/adolescent temperament. 5 How do temperament and environment interact in influencing children’s developmental outcomes? In some cases, rather than direct main effects, individual differences in children’s temperament may modify the impact of the environment on their development. In the final section, we present evidence of such temperament-environment interactions based on the FLS data. Specifically, we examine interactions between infant temperament and family conflict as predictors of behavior problems.

Consistency and Change in Temperament

In this chapter, we address two central issues in the study of development: change and consistency. The longitudinal research methodology is best suited to address the nature of temperament with advancement in age during childhood, as it allows examination of both the consistency of the individual’s rank position compared to others (i.e., stability) as well as change/consistency in the dimensions of temperament across individuals as a function of age (i.e., continuity). As noted in chapter 1, a particularly unique aspect of the Fullerton Longitudinal Study (FLS) is the frequency and continuity of temperament assessments; 10 assessment waves were conducted at designated and regular intervals from infancy (1.5 years) through adolescence (16 years). The temperament measures used during each developmental era and dimensions assessed on each measure are summarized in Table 3.01.

Temperament and Intelligence

In this chapter, we address one of the central themes of development charted in the FLS, intelligence. The interface and overlap between temperament and cognitive ability from infancy though adolescence are examined. As noted in chapter 1, five issues will be addressed: 1 1. Which dimensions of temperament correlate with concurrently-administered standardized tests of intelligence? The temperamental concomitants of intelligence are examined across the infant, preschool, middle childhood, and adolescent eras. 2 2. Can the temperament-intelligence relationship be explained by test-taking behaviors during the administration of the intellectual tests? At three points in time during the FLS, test-taking behavior as a possible mediating variable between temperament and intelligence test scores was examined. 3 3. Are the relations evident between temperament and cognitive skills as measured by standardized tests also evident when parent reports of developmental progress are used? As another approach to explore the possible mediating effect of test taking behavior on the relation between temperament and intelligence, we examined the relation between temperament and parent reports of developmental progress on standardized inventories of child development. 4 4. To what extent do aspects of earlier temperament forecast or predict intellectual development throughout childhood and adolescence? A necessary but not sufficient observation for temperament to influence children’s cognitive/intellectual growth is that earlier temperament relate to subsequent intelligence. In this section, the degree of relation between earlier temperament and intelligence across shorter and longer intervals of time are examined. 5 5. To what extent do earlier measures of children’s intelligence predict temperament and personality? In the final section, FLS measures of intelligence are examined as potential predictors of children’s behavioral style and personality. Concurrent temperament-intelligence relations do not elucidate whether individual differences in children’s ability to process and coordinate information and communicate their needs effectively may impact their behavioral styles. Hence, earlier measures of intelligence as predictors of subsequent temperament are examined to ascertain whether empirical evidence exists to support this possible direction of relation. Furthermore, discussion regarding the intersection between personality and intelligence has been revisited recently (Collis & Messick, 2001). In chapter 3, data from the FLS failed to show direct links between temperament during infancy, preschool years, or middle childhood with personality at age 17. To conclude chapter 5, we examine the extent to which early measures of intelligence foretell personality.

Temperament in the Extremes-Developmental Trajectories of Infants in Three Temperament Groups

At this point in the course of the FLS, comprehensive data detailing the development of over 100 children and their families have been collected from infancy through high school completion. In this chapter, we examine the FLS data to determine whether unique and differentiated long-term sequelae of the earliest measure of temperament employed in the FLS are evident. As noted in chapters 1 and 2, many researchers have suggested the potential benefit of employing an extreme groups approach. Hence, for each factor of challenging infant temperament assessed via the Infant Characteristics Questionnaire at 1.5 years, three groups were formed: low, intermediate, and high. The “low” group comprised infants scoring in the bottom quartile and viewed by parents as exhibiting the least challenging temperament features, the “intermediate” group comprised infants scoring in the center two quartiles and served as a baseline for comparison, and the “high” group comprised infants in the top quartile representing infants rated as displaying the most challenging temperament characteristics.