CORPORAL PUNISHMENT AND BEHAVIOR PROBLEMS ACROSS TWO GENERATIONS: ARE THERE COHORT EFFECTS?. Miriam Ehrensaft1, Stephanie Kasen1, Cohen Patricia1, 1Columbia University, New York, NY United States
Current popular child rearing literature strongly advises against the use of corporal punishment (CP). Yet, previous cohorts of parents were historically advised to use CP as an effective punishment. Some research on very young children suggests that recent generations of parents do use less punitive disciplinary practices than prior generations. Historical shifts in child rearing mores gives rise to questions about cohort effects on CP and child behavioral outcomes. Is the current generation of parents indeed using less CP than previous generations of parents? Do their attitudes towards CP as a legitimate disciplinary tool in fact differ from those of the previous generation? Are the behavioral effects of CP more pronounced in today´s cohort of parents and children, compared to previous cohorts, and if so, how may such differences be explained? We previously found a decrease in the use of aversive discipline, including spanking, of children under age 3, in mothers from one generation (G1) to the subsequent generation (G2). We have now collected subsequent assessments of both disciplinary practices and behavior problems in their offspring who are now in middle childhood and adolescence. We investigate generational differences in the prevalence of CP and its association with child behavior in a longitudinal study of a community sample of parents and children (N = 821) followed for over 25 years. Parents were first interviewed in 1975 about parenting practices and child psychosocial characteristics when their children were 1 to 10 years old, and re-interviewed multiple times until the offspring were in their early 30s. During data collection waves in both the early and late 1990s, we assessed parenting practices of offspring who now have their own children. Our earlier analyses of mothers of toddlers showed a decline in slapping or spanking from 87% to 72% of mothers in g1 versus 2. Aversive discipline predicted toddlers´ angry temperament with similar magnitude in both generations. We extend our prior work by examining the prevalence of CP by mothers of offspring in middle childhood, and compare the magnitude of association of such discipline with child behavior problems in both generations, during both early childhood and middle childhood. We compare parental attitudes to CP in both generations, and test for interactions of these effects with gender of offspring, social class, marital status, maternal and offspring age, and religiosity of each generation of parents. Analyses test explanatory mechanisms for any cohort differences in the strength of association of CP with behavioral outcomes. Societal changes in parenting practices and implications for current parent programs are discussed.
RACIAL AND ETHNIC IDENTITY AS PREDICTORS OF ACADEMIC OUTCOMES, READING SKILLS LINKED TO DEPRESSIVE SYMPTOMS, GENDER AND DEVELOPMENTAL DIFFERENCES
Chair: Dorothy Browne
RACIAL AND ETHNIC IDENTITY: PREDICTOR OF ACADEMIC OUTCOMES AND FOCUS FOR INTERVENTION. Inna Altschul1, Daphna Oyserman1, Deborah Bybee2, 1University of Michigan, Ann Arbor, MI United States; 2Michigan State University, East Lansing, MI United States
The gap in academic achievement between racial-ethnic minority and majority youths has been well documented. Although socio-economic factors contribute significantly to this gap, they do not fully explain the generally lowered achievement of racial-ethnic minority youths. Moreover, socio-economic status is a factor that does not lend itself easily to manipulation via intervention. Consequently, other factors, such as, the relationship between academic outcomes and racial-ethnic identity (REI), have been targeted for empirical investigation. REI is a particularly useful factor for intervention with middle-school youth as identity exploration is a major developmental task of early adolescence, a time when school failure becomes a more realistic prospect. With intervention in mind, Oyserman and colleagues (1995) developed a model of REI focused explicitly on the link with academic achievement, which consists of three discreet elements – in-group connectedness, achievement as an in-group value, and awareness of racism – proposing that these elements of REI are necessary to maintain academic engagement.
This paper presents the results of a longitudinal study examining academic outcomes and REI among low-income African American and Latino youths in Detroit using Oyserman´s REI model. Three hypotheses were tested: (1) REI predicts GPA over time such that youths high in all three REI components have higher GPAs; (2) the effects of REI are consistent across racial-ethnic groups; (3) the effects of REI are gendered with Connectedness promoting academic achievement for boys, and Embedded Achievement promoting academic achievement for girls. A random sample of African American (n=98) and Latino (n=41) students from three urban middle schools were surveyed at the beginning and end of eighth and ninth grades (4 data points). Grades were obtained from school records at each time point. Using hierarchical linear modeling (HLM), we show that REI can both enhance and hamper academic achievement depending on the particular components of REI endorsed. Partially supporting hypothesis 1, higher endorsement of two components of REI – Connectedness and Embedded Achievement – predicts higher GPA. The third component, Awareness of racism, has a negligible impact on GPA. Although, on average, girls have higher GPA´s than boys, no support was found for a gendered effect of REI (hypothesis 2); effects were equally strong for both genders. The effects of REI on GPA were the same for African American and Latino youths. Implications of these effects are discussed in the context of an intervention with low-income minority youths that aims to utilize REI to enhance connection with school.
LINKING READING SKILLS TO THE DEVELOPMENT OF DEPRESSIVE SYMPTOMS AND SHY BEHAVIOR: GENDER AND DEVELOPMENTAL DIFFERENCES ACROSS GRADES 1 TO 4. Scott Gest1, Frank Lawrence1, Sheppard Kellam2, Nicholas Ialongo3, 1Pennsylvania State University, University Park, PA United States; 2American Institutes for Research, Washington, DC United States; 3Johns Hopkins University, Baltimore, MD United States
The transactional relations between reading skills and two domains of internalizing behavior problems (depressive symptoms, shy behavior) were explored in two cohorts of students in a poor, urban school district. One question concerned directions of effect: do poor reading skills contribute to internalizing problems, do internalizing problems undermine reading skills, or do both processes occur? A second question focused on potential gender and developmental differences: does the strength or direction of links differ for girls and boys or for younger and older students? Prior results from Cohort I of the Baltimore Prevention Project revealed bi-directional links between reading skills and depressive symptoms during 1st Grade and in predictions from 1st Grade to the end of 5th Grade. This report builds on those findings in three ways: by examining year-to-year transactional effects between reading and depressive symptoms, by analyzing links between reading and shy behavior, and by analyzing parallel data from Cohort 2. Data were analyzed for students in the Control (non-intervention) conditions in Cohort 1 (N = 674) and Cohort 2 (N = 655) of the Baltimore Prevention Project. Reading skills were measured with the California Achievement Test (CAT), depressive symptoms were measured with the Baltimore How I Feel scale (BHIF), and shy behavior was measured with the Teacher Observation of Classroom Adaptation (TOCA-R). All measures were obtained in the Spring of Grades 1 through 4. Parallel sets of models were fit for depressive symptoms and shy behavior. In one model, intercept and growth parameters for the internalizing dimension were fit and scores on reading skills at each assessment were used to predict deviations from the trajectory-predicted internalizing score at the subsequent assessment (e.g., Read1->Dep2, Read2->Dep3, Read3->Dep4). In a second model, growth in reading skills was modeled and the reverse cross-lag coefficients were estimated (Dep1->Read2, etc.). Preliminary results from Cohort 1 indicate that reading skills consistently predicted decreased depressive symptoms, with somewhat stronger effects for boys and for older students. Effects of depressive symptoms on subsequent reading skills were weaker and were reliable only among older students. Reading skills were also associated with decreased teacher-rated shy behavior, but these effects were weaker than those for depressive symptoms and were somewhat stronger for girls. Shy behavior predicted lower reading skills only weakly and only among older students.
SOCIO-EMOTIONAL COMPETENCE AS A SALIENT FACTOR IN THE EARLY SCHOOL SUCCESS OF CHILDREN FROM LOW-INCOME FAMILIES. Jason Downer1, 1University of Virginia, Charlottesville, VA United States
Data from the 2002 U.S. Census Bureau indicate that approximately 17% of all children under the age of 18 are living at or below the federal poverty level, while rates are even higher for families with children under the age of six (Proctor & Dalaker, 2003). Research consistently indicates that low family income is detrimental to the development of young children and in particular cognitive and achievement outcomes (Duncan & Brooks-Gunn, 2000). However, some children demonstrate remarkable resilience, overcoming the adversity related to family poverty and reaching positive developmental outcomes (Masten & Coatsworth, 1998). Given that academic trajectories are relatively stable by third grade, it is important to establish ways to reduce risk and encourage resilience during the preschool and early elementary school years when children´s growth is malleable. A recent educational practice and policy shift emphasizes the need for young children in low-income families to receive more explicit instruction in the area of emergent literacy to bolster later achievement. However, evidence also indicates that relational experiences and socio-emotional development play key roles in promotion of academic performance over time (Pianta, 1999). This presentation will describe findings from a secondary data analysis of the National Institute of Child Health and Human Development Study of Early Child Care (NICHD SECC) sample that indicates a strong link between socio-emotional competence and academic achievement, particularly for children in families with lower income-to-needs ratios.
The NICHD SECC is a longitudinal data set that was designed to measure the impact of child care on child functioning, but it also offers an opportunity to identify targets for prevention of academic failure during early schooling. A cross-sectional design and multi-group structural equation modeling were utilized to test the path from socio-emotional competence to academic achievement at 54-months, first grade, and third grade. Initially, measurement models were fitted with the whole sample, which established the latent variables of socio-emotional competence (maternal ratings) and academic achievement (standardized tests). Then, children were dichotomized based on their family´s income-to-needs ratio, and multi-group path models were fitted. This presentation will report findings that indicate the path between children´s socio-emotional competence and their academic functioning during the early school years is more salient and strong for children from poor families than children from non-poor families. Implications for prevention efforts will be discussed, recognizing a need to address the promotion of socio-emotional competence during early years in order to ensure academic success.
FRIENDS DON´T LET FRIENDS: PSYCHOSOCIAL FACTORS RELATED TO ADOLESCENT´S INTERVENING IN THEIR FRIEND´S SUBSTANCE USE
Chair: Constance Flannagan
FRIENDS DON´T LET FRIENDS: PSYCHOSOCIAL FACTORS RELATED TO ADOLESCENT´S INTERVENING IN THEIR FRIEND´S SUBSTANCE USE. Elvira Elek1, Constance Flannagan1, Les Gallay1, Mike Stout1, Amy Bertelsen1, 1Pennsylvania State University, University Park, PA United States
Responsible choices form a cornerstone of prevention programs for young people, but the focus typically lies on individuals making informed decisions about their own health. The Social Responsibility Project instead focuses on the potential of peers to act as allies in prevention. Do adolescents act on the adage `friends don´t let friends ´? Do an adolescent´s beliefs about privacy and compassion for others influence their inclination to intervene in their friends substance use? What influence does their trust in society have on their feelings of responsibility towards their friend? Does their loyalty towards their friend influence their choice of intervention strategy (i.e., make them less willing to talk to an adult)? What role do developmental factors play in these relationships?
This line of research addresses the gap in knowledge regarding the above questions. The three papers in this symposium utilize the data of a longitudinal (3-wave) survey study of 5th-12th graders. Four thousand and two hundred students from thirty schools in Michigan and Pennsylvania participated over the course of the study. Active parental consent, obtained for all participants, resulted in a 54% participation rate. The sample included urban, rural, and suburban school districts and represented a broad range of socioeconomic groups. The racial/ethnic backgrounds of the participants reflected the populations of the school districts: 10% were African American, 5% were Hispanic, 1% were Asian American, and 84% were European American.
The three papers in this symposium focus on various psychosocial factors related to adolescents´ own substance use as well as their willingness to intervene in the substance use of their friends. The first paper examines student´s changes over a three year period in the health beliefs related to substance use in three distinct developmental groups – early, middle and late adolescents. The second paper will look both at how adolescents´ generalized social trust changes as they age along with how these changes in social trust influence the likelihood that an adolescent will intervene when their friend engages in substance use. The final paper focuses on the developmental differences in an adolescent´s willingness to intervene in a friends substance use along with the influence of loyalty to their friends in this difference.
HEALTH BELIEFS, CHANGE, AND PREVENTION. Les Gallay1, Constance Flanagan1, 1Pennsylvania State University, University Park, PA United States
Health knowledge and beliefs are fundamental components of risk prevention programs. Models such as the Health Belief Model posit that beliefs provide the motivation to avoid risk behaviors. There has been little research, however, on how young people´s health beliefs change over time or on the impact of developmental changes in beliefs on behavior. Adolescents who use drugs consider such use a personal rather than a moral matter (Berkowitz et al., 1991). American youth overwhelmingly consider self harm a matter of personal rights and hold that it is acceptable to ignore laws if the individual is the only one harmed (Nucci et al., 1991). The commitment to an individual´s `right´ to choose to harm him/herself appears to peak in middle adolescence and to decline thereafter (Killen et al., 1991).
The Social Responsibility and Prevention Project was designed specifically to test the hypothesis that health belief changes over time have an influence on the willingness of young people to intervene to dissuade peers from risky behavior We hypothesized that early adolescents would be motivated to intervene based on their health beliefs and that middle adolescents would be more committed to an individual´s right to privacy and self-determination and less cognizant of the public harm associated with an individual´s decisions than older adolescents.
This paper reports the results of a three-year longitudinal study of 770 5th through 10th graders in Michigan and Pennsylvania. Participants responded to 16 items detailing their health beliefs, beliefs in the rights of individuals to endanger themselves and the appropriateness of intervening in the behavior of others. Surveys were conducted between February of 2002 and June of 2004.
Findings support our hypotheses that risk beliefs change over time and that changes occur differentially among early, middle and late adolescents. Analyses were conducted using Repeated Measures Analysis of Variance for the three waves of the study. We found significant changes in risk beliefs over time differences in the developmental patterns of change between age groups. Belief in the danger associated with risky behavior declined over time, with the greatest changes occurring during the transition from primary to middle school. Middle adolescents did endorse a more individual rights orientation toward risk behavior than did early or middle adolescents. By late adolescence increased levels of socially responsible beliefs were evident and understanding of the dangers associated with certain behaviors had begun to recover.
DEVELOPMENTAL TRAJECTORIES OF SOCIAL TRUST AND ITS RELATIONSHIP TO PEER INTERVENTION. Mike Stout1, Constance Flanagan1, 1Pennsylvania State University, University Park, PA United States
Recent scholarly literature on social capital posits that communities that foster norms of trustworthiness and reciprocity among its citizens are more likely to have citizens who actively “look out” for one another. The increased vigilance that results from high levels of social trust has been associated with reduced crime rates, drug use, and overall dangerous behaviors. As people become more vigilant they tend to adopt an increased sense of responsibility for fellow community members. That is, they feel that it is in everybody´s best interest if their neighbors are not engaging in dangerous behaviors that could potentially disrupt life for others in the community. Therefore, individuals have a personal stake in monitoring the behavior of their neighbors and friends. Unfortunately, recent literature has also shown that at the national level social trust has been on the decline in recent decades (Uslaner, 2002). While many analyses have been conducted to analyze the effects of social trust on behaviors for adults, the literature for adolescents is sparse.
This paper poses two questions. First, what are the developmental trajectories of generalized social trust for adolescents? Second, how are adolescents perceptions of generalized social trust related to their likelihood of actively intervening when a friend engages in dangerous or deviant behaviors, particularly regarding the use of alcohol, tobacco, or other drugs?
Data from three waves of the Social Responsibility and Prevention Project were used to examine these questions. The analysis utilizes a hierarchical linear modeling framework to measure changes in adolescents´ social trust over time, with time points nested in individuals across the three waves of data. We find that social trust generally declines as individuals make the transition through early, middle, and late adolescence, and that the likelihood that they would actively intervene when a friend engages in dangerous behaviors also declines across these stages.
CULTURE OF SECRETS: PEER LOYALTY AND SOCIAL RESPONSIBILITY. Amy Bertelsen1, Constance Flanagan1, 1Pennsylvania State University, University Park, PA United States
Why don´t young people confide in adults if a peer is planning or doing something that could be harmful to themselves or others? One possibility is that they think that it is disloyal to tell an adult about concerns about a peer. Despite the importance of this issue to health and safety, no basic developmental research has addressed this issue.
Our study examined this question in a longitudinal (3-wave) study of early, middle, and late adolescents from Pennsylvania and Michigan. This abstract reports cross-sectional results from the first wave of the study (n=2,967). Students had active parental consent to participate in this study, which looked at those factors that were positively related to the likelihood of youth intervening to help peers avoid harmful behaviors. Youth were presented with vignettes in which a hypothetical peer engaged in harmful behaviors (e.g., riding in a car with an intoxicated friend that is driving, a friend is using drugs, a friend smokes cigarettes, you and friends are attending a party where ATOD will be present) and then asked to rate the likelihood that they would employ several intervention strategies. One of the intervention options youth could choose was to confide in adults (parents, teachers) about their concerns about a peer. We hypothesized that adolescents with high levels of peer loyalty would be more inclined to tell an adult because of their genuine concern for the health of their friend. Our loyalty measure included adolescents´ belief that friends could be trusted with a secret, that friends would keep a secret, that they would stand up for each other, the belief that they can go to a friend for help, and willingness to do anything for each other.
Multiple regression analyses revealed that loyalty to friends was high among all age groups but the likelihood of confiding in adults declined with age. Older youth were more likely to say that they would not talk to an adult and would ignore a friend´s harmful behaviors as “none of their business´. Additional analyses that controlled for age found no significant relationship between adolescents´ loyalty to their friends and their willingness to intervene in a friend´s harmful behaviors by talking to an adult (across all vignettes – alcohol, tobacco, drugs, party).
Further analyses examine whether a high sense of social responsibility (which taps a youth´s belief that it is important for young people to assume responsibility for classmates and people in their communities) moderates this relationship. Additionally, we explore the relationship between an adolescent´s loyalty to their friends and their willingness to intervene by talking directly to their friend about their concerns.
CONCURRENT 8, EPIDEMIOLOGY, Organized symposia
PREVALENCES, PREDICTORS, AND CONSEQUENCES OF NON-NORMATIVE BEHAVIOR TRAJECTORY
Chair: Katherine Masyn
PREVALENCES, PREDICTORS, AND CONSEQUENCES OF NON-NORMATIVE BEHAVIOR TRAJECTORY. Katherine Masyn1, Karen Nylund1, 1University of California, Los Angeles, Los Angeles, CA United States
When studying antecedents and consequences as well as the overall prevalences of problem behaviors in childhood and adolescence, it can sometimes be more instructive to examine prevalence and population heterogeneity in such behavior within a longitudinally frame rather than looking at rates of occurrence in singular cross-sections of developmental periods. Substantive theories will often define problematic behavior profiles with an explicit time-dependence, using terms such as “long-term”, “persistence”, etc. A type of behavior, such as physical aggression towards classmates, could certainly be considered problematic for a child, even in an isolated incident, but it could be the persistence of such behavior in primary school that would distinguish a child from others in his cohort as having higher risk for negative behavior consequences later in his development, such as conduct disorder. The predictors and distal outcomes related to problematic behavior trajectory profiles will likely be different from those variables related to the mere presence or absence of a behavior in a single time period.
The three papers presented in this symposium are population-based longitudinal studies that each utilize different advanced extensions of general growth mixture modeling that allow for the extraction of substantively meaningful profiles of behavior trajectories and provides estimates of the corresponding trajectory class prevalences. The behavior outcomes examined in these papers are aggression, internalizing/externalizing behavior problems, and juvenile arrest. Each paper explores the influence of antecedents (gender, temperament, and family background, respectively) on the likelihood of membership in the given trajectory classes. Also, the inclusion of long-term or distal outcomes in the same model allows investigation into the impact of trajectory class membership on the risk of later negative outcomes.