Craig A. Mason

Family, peer, and neighborhood influences on academic achievement among African-American adolescents: One-year prospective effects

Using a 1-year prospective design, this study examined the influence of family status variables (family income, parental education, family structure), parenting variables (maternal support and restrictive control), peer support, and neighborhood risk on the school performance of 120 African American junior high school students. In addition to main effects of these variables, neighborhood risk was examined as a moderator of the effects of parenting and peer support. Family status variables were not predictive of adolescent school performance as indexed by self-reported grade point average. Maternal support at Time 1 was prospectively related to adolescent grades at Time 2. Neighborhood risk was related to lower grades, while peer support predicted better grades in the prospective analyses. Neighborhood risk also moderated the effects of maternal restrictive control and peer support on adolescent grades in prospective analyses. These findings highlight the importance of an ecological approach to the problem of academic underachievement within the African American community.

Popularity, Friendship Quantity, and Friendship Quality: Interactive Influences on Children's Loneliness and Depression

A mediational model positing that the effects of popularity on childrens loneliness and depression are passed through indexes of friendship experiences was tested using structural equation modeling. Children (193 3rd through 6th graders) completed a battery of sociometric and self-report questionnaires from which measures of popularity, multiple friendship dimensions (i.e., quantity and quality of best and good friendships), and loneliness and depression were derived. Confirmation of a slightly modified model supported the mediational hypothesis. Although popularity exerted no direct impact on the adjustment indexes, it strongly influenced friendship, which, in turn, affected depression through its strong association with loneliness. It appears that popularity is important for setting the stage for relationship development, but that it is dyadic friendship experiences that most directly influence feelings of loneliness and depression.

Ethnic Differences in the Effect of Parenting on Gang Involvement and Gang Delinquency: A Longitudinal, Hierarchical Linear Modeling Perspective.

This study examined the relative influence of peer and parenting behavior on changes in adolescent gang involvement and gang-related delinquency. An ethnically diverse sample of 300 ninth-grade students was recruited and assessed on eight occasions during the school year. Analyses were conducted using hierarchical linear modeling. Results indicated that, in general, adolescents decreased their level of gang involvement over the course of the school year, whereas the average level of gang delinquency remained constant over time. As predicted, adolescent gang involvement and gang-related delinquency were most strongly predicted by peer gang involvement and peer gang delinquency, respectively. Nevertheless, parenting behavior continued to significantly predict change in both gang involvement and gang delinquency, even after controlling for peer behavior. A significant interaction between parenting and ethnic and cultural heritage found the effect of parenting to be particularly salient for Black students, for whom higher levels of behavioral control and lower levels of lax parental control were related to better behavioral outcomes over time, whereas higher levels of psychological control predicted worse behavioral outcomes.

Adolescent problem behavior: the effect of peers and the moderating role of father absence and the mother-child relationship.

Examined the effect of peer problem behavior, the absence of a father or equivalent in the home, and the mother-adolescent relationship as predictors of adolescent problem behavior in a sample of 112 African American adolescents. Statistical analyses compared a moderator model to a mediational model and a cumulative risk model. As predicted, the moderator model was superior to the alternative models. Specifically, whereas the mediational model predicted that the effect of father absence and the mother-child relationship upon adolescent problem behavior would be mediated by peer problem behavior, neither effected peer problem behavior or adolescent problem behavior. Similarly, a cumulative risk index did not predict either child or parent reports of problem behavior and was not sensitive to specific contingencies that existed between the predictor variables. In contrast, an interactive, moderator model described the data quite well. This model suggested that father or equivalent absence magnifies the negative impact of peer problem behavior, while a positive mother-adolescent relationship attenuates this risk. A strong mother-adolescent relationship also served to protect adolescents in father-absent homes from the risk of peer problem behavior.

The Use of Epidemiological Methodology as a Means of Influencing Public Policy

To best influence policymakers, researchers need to provide information and measures of effects that reflect the nature of policy decisions. Specifically, policymakers are often interested in factors associated with changes in the number of cases or rate of disorders in a community. Regression/analysis of variance (ANOVA) models, which focus on the prediction of means, slopes, and variances, do not directly address such questions. In contrast, epidemiological statistics, which focus on differences in proportions of cases, do provide such information. Three epidemiological measures of effect (the risk-ratio, the odds-ratio, and the population attributable fraction) are reviewed; their value as tools for informing public policy is discussed; and examples are provided illustrating their use. Researchers are encouraged to consider adopting an epidemiological perspective as part of their work.

Early risk factors for mental retardation: role of maternal age and maternal education.

The predictive value of maternal age and education in relation to rates of administratively defined mental retardation in a 3-year birth cohort (N = 267,277) was studied. Low maternal education placed individuals at increased risk for both educable mentally handicapped (EMH) and trainable mentally handicapped (TMH) placements. Older maternal age was associated with increased risk of mental retardation, but for individuals with EMH, this age effect was only seen in the lowest education group. In terms of population-level risk, it was younger mothers with 12 years of education or less whose births were associated with the greatest proportion of mental retardation. From a public policy viewpoint, children born to mothers with low levels of education are an important group to target for prevention/early intervention efforts.

The identification of early risk factors for severe emotional disturbances and emotional handicaps: An epidemiological approach

Epidemiological methodology is used to examine the relationship between early childhood risk factors and future identification as having a Severe Emotional Disturbance or as having an Emotional Handicap (SED/EH) at age 13. Data were obtained from 1979/1980 Florida birth records that were electronically linked with 1992/1993 Florida school records. An epidemiological perspective was chosen due to its ability to model both individual and community-level risk. In regards to increasing an individuals risk of SED/EH, two factors, gender (being male) and low maternal education (mother not completing high school at the time of the childs birth), were found to have particularly strong effects. When examining effects of these risk factors upon overall rates of SED/EH in the community, maternal education and marital status (being unmarried at the time of the childs birth) were associated with a large proportion of the cases. Health/biological markers were moderately associated with SED/EH on the individual level, but were related to a relatively small percentage of cases in the population. In addition, effects varied based upon ethnic/cultural heritage. Researchers are encouraged to consider using an epidemiological perspective and its potential utility in the field of community psychology and public policy is discussed.

Assessing Moderator Variables: Two Computer Simulation Studes

The present article proposes a strategy for conceptualizing moderating relationships based on their type-strictly correlational and classical correlational-and form continuous, noncontinuous, logistic, and quantum. The results of two computer simulation studies comparing the relative power of three statistical approaches for assessing moderator variables are presented: moderated multiple regression, multiple regression with a dichotomized moderator, and correlational analysis. As predicted, moderated multiple regression was generally found to be equal or superior to these alternative approaches at detecting moderator relationships. Although the alternative approaches did potentially offer greater statistical power under limited circumstances, this was only true in more extreme cases and required fairly accurate estimates of specific characteristics of the joint distribution of the predictor, moderator, and outcome variables.

A Review of Some Individual- and Community-Level Effect Size Indices for the Study of Risk Factors for Child and Adolescent Development.

This article discusses the computation and application of various epidemiological measures of effect in educational and developmental research. Specifically, epidemiology provides a potentially important perspective for studies identifying risk factors for healthy child development. For example, it allows for the examination of both individual-level risk (the impact of risk factors on individuals experiencing them) and community-level risk (the impact of risk factors on the overall number of cases within a population). In terms of individual-level risk, issues related to the use and interpretation of the risk-ratio, the odds-ratio, and the logistic regression odds-ratio are reviewed. In addition, community-level measures of effect, such as the population-attributable fraction percentage, are examined. Implications of the design methodology (cohort study, case-control study, or representative study) on the choice and use of these measures of effect are discussed. Data from a large-scale ongoing project in developmental epidemiology are presented throughout the article for illustrative purposes.

Impacting re-arrest rates among youth sentenced in adult court: An epidemiological examination of the Juvenile Sentencing Advocacy Project.

Examines the impact of a program aimed at reducing re-offending among juveniles transferred to adult court in Miami-Dade County, Florida. Initiated in 1998, the Juvenile Sentencing Advocacy Project (JSAP) worked to increase the degree to which defense lawyers, prosecutors, judges, and police officers considered the developmental status of youth charged with crimes, as well as the contextual basis for their behavior and their potential for rehabilitation. Through such activities, the goal was to increase the use of juvenile sanctions, rather than traditional adult sentences. Based on previous research, it was predicted that increased use of juvenile sanctions would be associated with fewer youth re-offending. This article examines 162 youth who were transferred to and sentenced in adult court during 1999. Re-offense patterns were monitored through June 2001. Analyses using epidemiological measures of effect found that the use of juvenile sanctions significantly increased following implementation of JSAP and that youth receiving adult probation or boot camp were 1.74 to 2.29 times more likely to re-offend than were youth receiving juvenile sanctions. The increased use of juvenile sanctions following implementation of JSAP corresponded to an 11.2% to 15.3% decrease in the number of youth one would have anticipated would re-offend had previous patterns of sentencing continued.