STATISTICAL POWER OF TRAINING EVALUATION DESIGNS

Sample size requirements needed to achieve various levels of statistical power using posttest-only, gain-score, and analysis of covariance designs in evaluating training interventions have been developed. Results are presented which indicate that the power to detect true effects differs according to the type of design, the correlation between the pre- and posttest, and the size of the effect due to the training program. We show that the type of design and correlations between the pre- and posttest complexly determine the power curve. Finally, an estimate of typical sample sizes used in training evaluation design has been determined and reviewed to determine the power of the various designs to detect true effects, given this sample-size specification. Recommendations for type of design are provided based on sample size and projected correlations between pre- and posttest scores.     

Time-Limited Brief Therapy with Families: A One-Year Follow-Up Study

Families from a previously reported study of time-limited brief treatment were followed up one year after termination. Results based on questionnaires completed by the parents provide no evidence for deterioration after termination for families that had received six-session time-limited therapy, twelve-session time-limited therapy, or treatment without time limits. At one year, there were no significant differences in outcome between time-limited and unlimited approaches, between six-session and twelve-session treatment, or between treated families and a waiting list control group (most of whom had entered therapy by this time). As in the original study, the results do not support claims of increased therapeutic effectiveness for time and session limits. However, time limits did succeed in shortening treatment without significantly reducing its effectiveness or the durability of outcome.     

Perception of Cognitive Ability in Families of Adolescents

The families of 20 hospitalized adolescents were compared with a control group of 20 non-pathological families on the accuracy with which family members could predict each others' performance on two tests of intellectual ability. Accuracy of prediction was measured by the discrepancies between actual scores obtained by each family member and the scores other family members predicted he or she would obtain. The control and experimental groups were found to differ significantly on socioeconomic status and intellectual ability, and when these variables were controlled, a covariance analysis revealed no significant difference between the groups on accuracy of predictions. The results suggest that variables other than psychopathology must be taken into account in explaining apparent differences between healthy and disturbed families on measurements of interpersonal judgments. Implications for research and family therapy are discussed.     

THE MODERATING EFFECT OF SITUATIONAL CONTROL OF PERFORMANCE VARIANCE ON THE RELATIONSHIP BETWEEN INDIVIDUAL DIFFERENCES AND PERFORMANCE1

The present study investigated the existence of a moderating effect for situational control of performance variance on the relationship between individual differences and performance. An experimental simulation was conducted and validity coefficients were calculated. Results supported the presence of the predicted moderating effect. The implications of these data for validation research and testing programs are discussed.     

An Empirically Derived Typology of Families: I. Relationships with Adult Health

In this report we describe the development and partial validation of an empirically derived typology of families based on 11 family variable composites derived from the California Family Health Project. Our goal was to use the typology to condense and integrate the findings from previous analyses of a large group of family variables and to account for differences in the self-reported health of adult family members. Exploratory and confirmatory cluster analyses conducted separately by gender classified 97% of the sample into four parallel types for husbands and wives: Balanced, Traditional, Disconnected, and Emotionally Strained. A 1-way MANOVA indicated that all 11 family variable composites significantly differentiated the four family types for husbands and wives. Significant differences among the four family types were also found on 10 demographic and other family variables, using ANOVA. Using MANOVA, we compared the four family types on 12 self-reported health and well-being variables by gender. Both husbands and wives from Balanced and Traditional families reported higher health scores than spouses from Disconnected and Emotionally Strained families, but no single profile of health scores was unique to a particular family type. The four family types provide an integrated and comprehensive framework for describing the family in health research.     

The California Family Health Project: VI. Multidomain Analyses

The preceding articles in this series have reported associations between each of four domains of family variables (World View, Emotion Management, Structure/Organization, Problem Solving) and the Health of adults in a community-based sample of 225 families. In this article, we explore the relationships among all four family domains (73 variables) and between all four family domains and adult Health. The results of both principal components and multidimensional scaling analyses suggested the viability of the four-family-domain framework. As expected, the pattern of relationships among the variables in each family domain was maintained when Health was added to the analyses. Using multidimensional scaling analysis, we also found that the relationship between some family domains and Health changed when viewed in the context of the other family domains. These shifts suggested the importance of family context on the relationship of that family domain and Health. Family World View and Emotion Management maintained their relationships with Health; family Structure/Organization shifted relatively moderately; and Problem Solving shifted substantially. Family World View and family Emotion Management may be relatively more independent in their relation to family member health than family Structure/Organization and Problem Solving.