Regression Analysis of Multivariate Binary Response Variables Using Rasch-Type Models and Finite-Mixture Methods

A model is considered for the regression analysis of multivariate binary data such as repeated-measures data (for example, panel data) or multiple-indicators with measures of some underlying characteristic such as attitude or ability (for example, surveys or tests). The model is related to the usual Rasch model, the usual latent-class model, and other familiar models such as logistic regression. In addition to a regression specification, the model includes parameters that describe heterogeneity not accounted for by the predictors. In contrast to most other approaches, a nonparametric specification of the latent mixing distribution is used, leading to a formulation based on scaled latent classes. We examine the relationship between this model and several other models, give a tractable formulation of the likelihood function and likelihood equations, present an algorithm for maximum-likelihood estimation, and analyze marginal and conditional latent structures. The approach is illustrated with longitudinal data from the German Socioeconomic Panel.     

Matching With Multiple Controls to Estimate Treatment Effects in Observational Studies

Matching to control for covariates in the estimation of treatment effects is not common in sociology, where multivariate data are most often analyzed using multiple regression and its generalizations. Matching can be a useful way to estimate these effects, especially when the treatment condition is comparatively rare in a population, and controls are numerous but mostly unlike the treatment cases. Matching on numerous covariates is abetted by the estimation of propensity scores, or functions of the probability that cases are treatments rather than controls. This procedure is illustrated in the estimation of the effects of an organizational innovation on Medicare mortality within hospitals; the data set is very large, but innovative hospitals few, and many of the remaining hospitals are quite unlike the hospitals constituting the treatment subsample. Results are based on a variance-components model that is extended to consider the effects of an additional covariate. They show effects of the organizational innovation comparable to those estimated via multiple regression models but with substantially reduced standard errors.     

Matching deprivation mapping to urban governance in three Indian mega-cities

Recent discussions of poverty recognize its multi-dimensional character, focusing on multiple sources of deprivation that poor households experience. However, for urban planners and politicians to implement intervention programs effectively in mega-cities, knowledge on sources and spatial patterns of multiple deprivations is needed. Disaggregated analysis potentially allows them to target programs to specific locations, set priorities in line with local needs, and be efficient in the use of funding. This article maps multiple deprivations of households in three Indian mega-cities, at the electoral ward level. It uses the livelihoods approach to address the following questions: 1) Is there a spatial concentration of deprivations within cities and do poverty levels differ between them? 2) Do cities differ in the types of deprivations? and 3) do these cluster within the city? The paper compares Delhi, Mumbai and Chennai, using a multi-criteria model within a geographical information system developed for identifying hotspots of multiple deprivation. It further links the results of deprivation mapping to questions of urban governance by comparing deprivation results with data on slums currently used to target anti-poverty programs. Finally, the opportunities of using spatial data for improving priorities and efficient use of funds in poverty programs are analyzed.     

Family Intervention and Health Care Costs for Kansas Medicaid Patients with Schizophrenia

Despite a number of studies investigating the effect of pharmacotherapy on treatment costs for schizophrenia patients, there has been little attention given to the effect of family intervention. In this study, data from the Kansas Medicaid system were used to analyze healthcare costs for 164 schizophrenia patients who had participated in family intervention. Structural equation modeling was used to test two competing views of the role of family intervention in treatment. The results showed that a model including direct and indirect effects of family intervention provided a better fit to the data. Family intervention had a significant indirect effect on general medical costs (through other psychological treatment) that showed a savings of $586 for each unit increase in the provision of these services. In addition, the total indirect effects for family intervention showed a $580 savings for general medical costs and $796 for hospitalization costs (for each unit increase).     

Time and the external validity of experiments

Time is a major consideration in the external validity of both randomized and naturalistic experiments. Major threats to external validity are: (1) time unit of analysis; (2) seasonal and cyclical variation in subjects; (3) dynamic covariation; (4) cross-time intervention effects; (5) subject-time sampling problems (6) time-personological interactions; and (7) effect size changes over time. Problems associated with sampling and randomization efficacy over time are discussed.     

REGRESSION WITH MISSING YS: AN IMPROVED STRATEGY FOR ANALYZING MULTIPLY IMPUTED DATA

When fitting a generalized linear model—such as linear regression, logistic regression, or hierarchical linear modeling—analysts often wonder how to handle missing values of the dependent variable Y . If missing values have been filled in using multiple imputation, the usual advice is to use the imputed Y values in analysis. We show, however, that using imputed Y s can add needless noise to the estimates. Better estimates can usually be obtained using a modified strategy that we call multiple imputation, then deletion (MID). Under MID, all cases are used for imputation but, following imputation, cases with imputed Y values are excluded from the analysis. When there is something wrong with the imputed Y values, MID protects the estimates from the problematic imputations. And when the imputed Y values are acceptable, MID usually offers somewhat more efficient estimates than an ordinary MI strategy.     

Recovery coaches and the stability of reunification for substance abusing families in child welfare

Substance abuse is a long-standing challenge for child welfare systems. Parental substance abuse disrupts family stability, family cohesion, and jeopardizes the well-being of children. In the current study we test an intervention to improve child welfare outcomes for substance abusing families, specifically the probability of families achieving a stable (at least 12 months) reunification. The intervention was an integrated case management model where recovery coaches were appointed to substance abusing parents associated with an open foster care placement. A diverse group of families (n = 1623) were randomly assigned to either a control group (services as usual) or an experimental group (services as usual plus a recovery coach). Multinomial logistic regression indicated that substance abusing parents associated with a recovery coach were significantly more likely to achieve a stable reunification as compared with similar families in the control group.     

AVERAGE PREDICTIVE COMPARISONS FOR MODELS WITH NONLINEARITY, INTERACTIONS, AND VARIANCE COMPONENTS

In a predictive model, what is the expected difference in the outcome associated with a unit difference in one of the inputs? In a linear regression model without interactions, this average predictive comparison is simply a regression coefficient (with associated uncertainty). In a model with nonlinearity or interactions, however, the average predictive comparison in general depends on the values of the predictors. We consider various definitions based on averages over a population distribution of the predictors, and we compute standard errors based on uncertainty in model parameters. We illustrate with a study of criminal justice data for urban counties in the United States. The outcome of interest measures whether a convicted felon received a prison sentence rather than a jail or non-custodial sentence, with predictors available at both individual and county levels. We fit three models: (1) a hierarchical logistic regression with varying coefficients for the within-county intercepts as well as for each individual predictor; (2) a hierarchical model with varying intercepts only; and (3) a nonhierarchical model that ignores the multilevel nature of the data. The regression coefficients have different interpretations for the different models; in contrast, the models can be compared directly using predictive comparisons. Furthermore, predictive comparisons clarify the interplay between the individual and county predictors for the hierarchical models and also illustrate the relative size of varying county effects.     

MEANINGFUL REGRESSION AND ASSOCIATION MODELS FOR CLUSTERED ORDINAL DATA

Many proposed methods for analyzing clustered ordinal data focus on the regression model and consider the association structure within a cluster as a nuisance. However, the association structure is often of equal interest—for example, temporal association in longitudinal studies and association between responses to similar questions in a survey. We discuss the use, appropriateness, and interpretability of various latent variable and Markov models for the association structure and propose a new structure that exploits the ordinality of the response. The models are illustrated with a study concerning opinions regarding government spending and an analysis of stability and change in teenage marijuana use over time, where we reveal different behavioral patterns for boys and girls through a comprehensive investigation of individual response profiles.     

Use of a Parent Management Training Intervention with Urban Foster Parents: A Pilot Study

The development of effective interventions for foster children with behavior problems is essential given the consequences of behavior problems for children's placement stability and permanency outcomes. This article presents findings from a pilot study of an intervention providing parent management training (PMT) and support to foster parents in groups and home visits. The intervention was an adaptation of the KEEP (Keeping Foster Parents Trained and Supported) group intervention, provided in a large urban child welfare agency serving predominantly African American foster parents. The study used an intent-to-treat design, with 25 foster parents of 31 children (ages 4-12) in specialized foster care assigned to either an intervention or treatment as usual control group. Longitudinal outcomes were analyzed using random effect regression models. Over time, children's behavior problems were significantly lower in the intervention group relative to the control group, and the effect of the intervention was partially mediated by parents' understanding of how to appropriately use the intervention parenting skills. These results provide support for the effectiveness of KEEP with urban foster children with significant behavior problems.     

Quasi-experimental evaluation using confirmatory procedures: A case study of an economic and social empowerment intervention to reduce girls’ vulnerability to HIV in rural Mozambique

PurposeOur goal was to determine whether an economic and social empowerment intervention implemented in Zambézia Province, Mozambique reduced girls’ vulnerability to HIV. We use this experience to discuss challenges of evaluating real-world interventions.MethodsTwo rounds of data were collected from 885 girls, 13–19 years, for this clustered, non-equivalent (two-stage) cohort trial. We used multi-level exact matching and difference-in-differences estimation to estimate intervention effects on two outcomes: girls’ knowledge of gender-based violence and school attendance.ResultsEstimates of two outcomes analysed indicated no statistically significant intervention effects. Preliminary analysis of data from the intervention group revealed this study was unable to obtain accurate measures for five outcomes related to HIV vulnerability.ConclusionsAlthough our study did not find evidence of impact on the a priori selected outcomes, we report on our experience implementing this robust methodologic design and describe how the challenges encountered in this program setting affected our ability to attain results. We recommend prospective evaluation designs with random allocation be accommodated early during planning. When not possible, quasi-experimental studies should collect data from large samples. To reduce measurement bias, biological endpoints such sexually transmitted infections should serve as primary outcomes for programs intending to reduce sexual behaviors.     

Exploring Dimensions of Culturally Sensitive Clinical Styles with Latinos

Abstract

This study sought to identify dimensions of culturally sensitive clinical practice by comparing the personal clinical styles of Latino and non-Latino clinicians, and their clinical decisions in respect to Latino immigrant clients. A combination of qualitative and quantitative methods were used. Sixty Latino American clinicians and 47 non-Latino (White) clinicians in Northern California completed the Achieving Styles Inventory (ASI) which yields information on direct, instrumental and relational styles. Demographic data were obtained as well as information on language proficiency and years of clinical experience. Statistical analysis included uses of t-test, ANOVA, and multiple regression analysis and revealed significant differences between the two groups in personal achieving styles within all three domains. There were no significant differences on the demographic variables. Nine clinician members of each ethnic group were interviewed in-depth through non-directive interviewing styles to assess their clinical decision making via use of a case vignette. Content analysis pointed to similarities and differences in proposed assessment and intervention strategies. The Latino group displayed more culturally relevant interpretations. The non-Latino clinicians were more directive and instrumental than the Latinos who put more emphasis on self disclosure, case management, use of Spanish with the client, and crisis intervention. This finding is consistent with both the quantitative data analysis of ASI hypotheses and the expected linguistic/cultural findings of the qualitative content analysis. Bilingual non-Latino White clinicians showed more cultural sensitivity and competence than the monolingual non-Latino clinicians. These findings suggest that linguistic/cultural match and linguistic/cultural compatibility are important in cross-cultural clinical practice.     

The role of positive academic self-concept in promoting school success

Identity research argues for enhancing students' current and future positive academic self-concepts to strengthen educational success. However, multiple factors from youths' home and school ecologies, as well as structural disadvantage, influence this relationship. Using the data from the Beyond High School Study (N = 9658), this analysis examines the role of academic self-concept in predicting school success over and above co-occurring contributors. The effects of positive academic self-concept on future educational aspirations, accessing educational guidance counseling, and student GPA were tested using stepped linear regression, controlling for student socio-demographics, school environment factors, and parental support. Results confirmed hypotheses for each academic indicator, with positive academic self-concept demonstrating the strongest coefficient. Implications for school-based intervention are discussed, linking to social psychological literature on future-oriented self-cognitions and strengthening motivational and regulatory function, particularly among youth facing systemic challenges.     

Evaluating unintended program outcomes through Ripple Effects Mapping (REM): Application of REM using grounded theory

Several evaluation models exist for investigating unintended outcomes, including goal-free and systems evaluation. Yet methods for collecting and analyzing data on unintended outcomes remain under-utilized. Ripple Effects Mapping (REM) is a promising qualitative evaluation method with a wide range of program planning and evaluation applications. In situations where program results are likely to occur over time within complex settings, this method is useful for uncovering both intended and unintended outcomes. REM applies an Appreciative Inquiry facilitation technique to engage stakeholders in visually mapping sequences of program outcomes. Although it has been used to evaluate community development and health promotion initiatives, further methodological guidance for applying REM is still needed. The purpose of this paper is to contribute to the methodological development of evaluating unintended outcomes and extend the foundations of REM by describing steps for integrating it with grounded theory.     

Using Mixture Regression to Identify Varying Effects: A Demonstration With Paternal Incarceration

The most widely used techniques for identifying the varying effects of stressors involve testing moderator effects via interaction terms in regression or multiple‐group analysis in structural equation modeling. The authors present mixture regression as an alternative approach. In contrast to more widely used approaches, mixture regression identifies varying effects without reliance on tests of moderator variables, such as using interaction terms or multiple group analyses. In many instances, the use of mixture regression also more effectively tests higher order and multiple interactions. A mixture regression example is presented using 214 families from the Fragile Families and Child Wellbeing study, half of whom had experienced paternal incarceration. Whereas typical regression and moderator analyses fail to find an effect or varying effects, mixture regression identified 4 classes uniquely influenced by the incarceration.     

Some Determinants of Work Alienation*

This study represents an attempt to examine the nature and sources of self-estrangement in work. The concept is defined using a discrepancy approach. Questionnaire data were collected from 1,313 employees representing seven occupational groups. The results suggest that individual work values interact with perceptions of the work situation to produce different levels of self-estrangement within each occupational group. Furthermore, multiple regression analysis indicates that job conditions are significantly more important than background factors for explaining feelings of self-estrangement. In particular, lack of control over task activities (powerlessness) and lack of meaningful work (meaninglessness) are found to be the most powerful predictors across all occupational groups. The implications of these findings are discussed.     

Types of female orgasm

This study was designed to evaluate the relationship between child sexual abuse (CSA), family violence, and other negative childhood experiences within the home and adult adjustment (TSC‐40 and Rosenberg's Self‐Esteem Scale) in a sample of first‐year female psychology students (n = 199). An experience of CSA was reported by 23.6% of women. Results of a discriminant function analysis revealed that level of family violence, father's traditional family values and mother's and father's sexual punitiveness were the variables that best discriminated between sexually abused and nonabused women. After other differences in family background were accounted for by using hierarchical multiple regression, level of family violence significantly improved the prediction of adult psychological adjustment. Standard multiple regression revealed that characteristics and severity of CSA were not significant predictors of adult adjustment. These results were interpreted as support for the use of an integrated risk‐sequelae model of CSA in evaluating the relationship between CSA and adult adjustment. Within this model, the interrelationships among CSA, violence, and dysfunctional familial environments should be simultaneously assessed.     

Self‐actualization and sexual permissiveness,satisfaction, prudishness,and drive among female undergraduates

The relationship between self‐actualization and aspects of female sexuality was examined. Ninety‐five white, single female undergraduates from a regional southeastern state university volunteered as subjects. Three questionnaires were administered: the Personal Orientation Inventory (POI), Eysenck's Inventory of Attitudes to Sex, and a two‐item religiosity measure. The data were analyzed via correlation and hierarchical multiple regression analysis which employed religiosity, age, major, grade, residence (on/off campus), and the order of test presentation as covariates. The results indicated that, when the effects of the covariates were controlled, there was a significant relationship between self‐actualization and sexual permissiveness, sexual satisfaction and prudishness. Although self‐actualization and sexual libido were significantly correlated, self‐actualization did not account for a significant portion of the variance in sexual libido after the effects of the covariates were controlled. Implications for future research include using different populations to investigate the relationship between self‐actualization and sexual attitudes and behaviors, using different measures of self‐actualization to investigate its relationship to sexual attitudes and behaviors, and examining the relationship between self‐actualization and other aspects of sexuality.     

Disparities in Access to Health Care Among Middle-Aged and Older Adults with Disabilities

Disability is increasing among middle-aged adults and, reversing earlier trends, increasing among older adults as well. Disability is experienced disproportionately by Black and lower socioeconomic status (SES) individuals. We used Medical Expenditure Panel Survey data to examine health care disparities in access to health care for middle-aged (31 to 64 years of age) and older (65+ years of age) adults with disabilities by race and ethnicity, education, and income (n = 13,174). Using logistic regression, we examined three measures of potential (e.g., usual source of care), and three measures of realized (e.g., counseling related to smoking) access. Middle-aged and older minority individuals with disabilities had lower relative risks of having usual sources of care and higher relative risks of having suboptimal usual sources of care (e.g., a place rather than a person) than White adults with disabilities. There were SES effects observed for middle-aged adults with disabilities across most measures that were, for certain measures, more pronounced than SES effects among older adults with disabilities. These findings are important, since health resources (e.g., a usual source of care) may mediate relations among disability, morbidity, and mortality. Policy actions that may mitigate the disparities we observed include financial incentives to support access to an optimal usual source of care and mechanisms to foster behavioral interventions related to smoking and exercise. Ensuring that these actions address the specific concerns of individuals with disabilities, such as physical accessibility and provider cultural competency, is essential.     

Evaluation of the Stress Adjustment and Adaptation Model Among Families Reporting Economic Pressure

ABSTRACT

This research evaluates the Stress Adjustment and Adaptation Model (double ABCX model) examining the effects resiliency resources on family functioning when families experience economic pressure. Families (N = 128) with incomes at or below the poverty line from a rural area of a southern state completed measures of perceived economic pressure, family hardiness, social support, and the McMaster Family Assessment device. The family resiliency model was evaluated using regression and two path analyses. Results of the model evaluation indicate that when family resiliency resources are introduced economic pressure is not a significant stressor influencing family functioning. Family resiliency is a key to the healthy functioning of the family unit moderating distress and decreasing the negative effects associated with economic pressure. Implications for social work practice to enhance family resiliency using community and personal supports are discussed.