Search

The quality of the home environment, as a predictor, is related to health, education, and emotion outcomes. However, factors influencing the quality of the home environment, as an outcome, have been understudied--particularly how children construct their own environments. Further, most previous research on family processes and outcomes has implemented between-family designs, which limit claims of causality. The present study uses kinship data from the National Longitudinal Survey of Youth to construct a maternal sibling-comparison design to investigate how maternal and child traits predict the quality of home environment. Using a standard between-family analysis, we first replicate previous research showing a relationship between maternal intelligence and the quality of the home environment. Then, we reevaluate the link between maternal intelligence and the home environment using differences between maternal sisters on several characteristics to explain differences between home environments for their children. Following, we evaluate whether child intelligence differences are related to home environment differences in the presence of maternal characteristics. Results are compared with those from the between-family analysis. Past causal interpretations are challenged by our findings, and the role of child intelligence in the construction of the home environment emerges as a critical contributor that increases in importance with development.

J. C. DeFries and D. W. Fulker's (1985) regression model (since named DF analysis) used kinship pair data to separate heredity and shared environmental influences. This article extends DF analysis to include measured indicators of the nonshared environment. These indicators represent specific sources of environmental influence that cause related children to be different from one another. We present two empirical studies using twin, full-sibling, half-sibling, and cousin pairs from over 7,000 5- to 11-year-old children in the National Longitudinal Survey of Youth. Study 1 is a validity analysis of kinship height and weight data. Study 2 is a DF analysis of problem behavior scores. Spanking, reading, and quality of the home environment are shown to account for nonshared variance.

Applies the confluence model to a longitudinal intact family data set. Also tests the data with simple linear models. Results suggest that the confluence model should be rejected as a useful model for these data based on principles of parsimony and predictability. (Author/AS)

Examines the contiguity of siblings within family structure--a concept reflecting how much of the family environment siblings share with one another--in relation to within family IQ scores. Results fail to demonstrate family structure effects. (HOD)

Analysis of data in the ADSEX (Adolescent Sexuality) Dataset showed that younger siblings have consistently higher levels of sexual activity at a given age than do older siblings. The effect does not interact with the race or sex-match of siblings. A sibling influence interpretation is considered to be the most plausible explanation of the findings. (RH)

Reviews and comments on Stoolmiller's (1998) criticisms of an epidemic model of the onset of social activities (EMOSA) and about nonlinear modeling in general. Discusses the idea of social contagion as a general theoretical tool. (Author)

Analyzed the correlation of nonsexual deviance and physical, sexual behavior using a sibling design. Hypothesized that both types of behavior are partly determined by a latent trait of deviance proneness, d. In two separate studies--one based on an Oklahoma dataset of college students and their siblings, and the other, on the Adolescent Sexuality Project (ADSEX) dataset of high-school students and their siblings in Tallahassee, Florida--found a strong relationship between relatively early sexual intimacy and nonsexual forms of deviance. Siblings were more alike than chance in deviance and in physical sexual behavior. Most critical for the model, there was also an association between one sibling's sexual intimacy with a partner and the other's deviance. Using LISREL, tested the latent-trait model statistically and accepted it as consistent with the obtained correlations.