Regression for non-Euclidean data using distance matrices

Regression methods for common data types such as measured, count and categorical variables are well understood but increasingly statisticians need ways to model relationships between variable types such as shapes, curves, trees, correlation matrices and images that do not fit into the standard framework. Data types that lie in metric spaces but not in vector spaces are difficult to use within the usual regression setting, either as the response and/or a predictor. We represent the information in these variables using distance matrices which requires only the specification of a distance function. A low-dimensional representation of such distance matrices can be obtained using methods such as multidimensional scaling. Once these variables have been represented as scores, an internal model linking the predictors and the responses can be developed using standard methods. We call scoring as the transformation from a new observation to a score, whereas backscoring is a method to represent a score as an observation in the data space. Both methods are essential for prediction and explanation. We illustrate the methodology for shape data, unregistered curve data and correlation matrices using motion capture data from an experiment to study the motion of children with cleft lip.     

FAMILY SYSTEMS THERAPY AFTER OPERATIONS DESERT STORM WITH EUROPEAN-THEATER VETERANS

We describe a quasi-experimental trial of time-limited family therapy with veterans and families of veterans who served in Europe, outside the war zone, during Operation Desert Storm (ODS). Family systems therapy was provided both to individuals and conjointly to couples or families during the acute postwar readjustment period. The intervention adapted strategies from structural, strategic, intergenerational, and behavioral family therapies in a brief-treatment protocol for systemic stressor resolution. Veterans given family system therapy were able to resume functional levels of psychosocial adjustment and reduce the risk of long-term (Chronic or delayed) psychosocial impairment. Based on these preliminary findings, controlled evaluation of family systems therapy appears warranted for indi-     

Modeling the Effects of Expressed Emotion,Psychiatric Symptomology,and Marital Quality Levels on Male and Female Verbal Aggression

Previous research has linked expressed emotion (EE) levels and psychiatric symptomology, whereas other empirical efforts have examined the relationship between EE and marital relationship variables. Still other research has linked psychological conditions with acts of aggression, and the impact of marital issues within this domain of inquiry also has been explored. The study presented here examined the combined effects of EE, marital quality, and psychiatric symptomology on displays of verbal aggression by both relationship partners. Both husbands' and wives' reports of marital quality were predictive of both husbands' and wives' verbal aggression levels; here, reports of lower marital quality were associated with higher levels of verbal aggression. Additionally, husbands' and wives' reports of psychological symptomology in self and partner were important paths to marital quality, as were husbands' and wives' reports of EE levels. Further, some evidence was generated that supports the mediating role of marital quality with regard to EE and psychological symptomology. However, the overall fit indices of the models tested in this empirical effort were modest. Limitations that concern the instruments utilized in the present study are discussed in this regard.     

Social Studies Teachers’ Organization

Recent innovations in Web-based technology tools have made planning instruction with technology in mind far more doable than in years past. To aid teachers in planning with technology, tools are organized into five broad categories: communication, collaboration, presentation, organization and critical-thinking. The purpose and potential of each group of tools is explored and utilized in the development of a technology-rich unit plan on medieval China.     

Potential Selective Responding in a Parent Questionnaire Study of Post-Institutionalized Children

This study assesses selective responding in a single wave of data collection and a four-wave study. Participants were 121 parents of post-institutionalized children, identified as “never responders,” “previous responders,” or “wave 4 responders.” Parents evaluated their adopted child's family, school, peer, and behavioral adjustment. Children (47% male) were 2 to 20 years old (M = 10.79, SD = 4.59) and adopted between 5 and 54 months of age (M = 15.49, SD = 9.94). Within a single wave of data collection, no evidence suggested that selective responding contributes much bias. Over a multi-wave study, however, results may under-represent adjustment difficulties.     

Social Skills in Children Adopted From Socially-Emotionally Depriving Institutions

This study assessed social skills in post-institutionalized (PI) children with respect to age at adoption, age at assessment, and gender. Parent ratings of social skills (Social Skills Rating System) and behavior problems (Child Behavior Checklist) were obtained for 214 children and 127 adolescents who were adopted from socially-emotionally depriving Russian institutions. Results showed that children adopted before 18 months of age have better social skills than those adopted after this age; those assessed in childhood demonstrate better social skills than those assessed in adolescence. PI females, especially later-adopted adolescents, have particularly poor social skills. Children with poor social skills tend to have higher rates of behavior problems.     

A re-evaluation of random-effects meta-analysis

Summary.  Meta-analysis in the presence of unexplained heterogeneity is frequently undertaken by using a random-effects model, in which the effects underlying different studies are assumed to be drawn from a normal distribution. Here we discuss the justification and interpretation of such models, by addressing in turn the aims of estimation, prediction and hypothesis testing. A particular issue that we consider is the distinction between inference on the mean of the random-effects distribution and inference on the whole distribution. We suggest that random-effects meta-analyses as currently conducted often fail to provide the key results, and we investigate the extent to which distribution-free, classical and Bayesian approaches can provide satisfactory methods. We conclude that the Bayesian approach has the advantage of naturally allowing for full uncertainty, especially for prediction. However, it is not without problems, including computational intensity and sensitivity to a priori judgements. We propose a simple prediction interval for classical meta-analysis and offer extensions to standard practice of Bayesian meta-analysis, making use of an example of studies of 'set shifting' ability in people with eating disorders.