A Survey of Methods for Analyzing Clustered Binary Response Data

Pendergast, Jane; Gange, Stephen J.; Newton, Michael A.; Lindstrom, Mary J.; Palta, Mari; Fisher, Marian R.

doi:10.2307/1403425

Cited by 243 publications

(156 citation statements)

References 0 publications

Supporting

Mentioning

156

Contrasting

Order By: Relevance

“…[84][85][86] For objective 1 (baseline demographics and previous insurance experiences by race/ethnicity), we performed bivariate analyses to describe sociodemographic characteristics and previous insurance experiences by race/ethnicity.…”

Section: Analysesmentioning

confidence: 99%

“…84,85,87 For objective 3 (determining whether racial/ethnic disparities were explained by socioeconomic/demographic or health system factors), we used generalized linear mixed models to accommodate nested survey sampling and repeated observations for each child (measures before and again during enrollment). We included known sociodemographic and health system factors as covariates and examined the differential effects of SCHIP over time by racial/ethnic group with an interaction term (racial/ ethnic group ϫ time).…”

Section: E698mentioning

confidence: 99%

See 1 more Smart Citation

Reduction in Racial and Ethnic Disparities After Enrollment in the State Children's Health Insurance Program

Shone¹,

et al. 2005

View full text Add to dashboard Cite

ABSTRACT. Background. Racial/ethnic disparities are associated with lack of health insurance. Although the State Children's Health Insurance Program (SCHIP) provides health insurance to low-income children, many of whom are members of racial/ethnic minority groups, little is known about whether SCHIP affects racial/ethnic disparities among children who enroll.Objectives. The objectives of this study were to (1) describe demographic characteristics and previous health insurance experiences of SCHIP enrollees by race, (2) compare racial/ethnic disparities in medical care access, continuity, and quality before and during SCHIP, and (3) determine whether disparities before or during SCHIP are explained by sociodemographic and health system factors.Methods. Pre/post-parent telephone survey was conducted just after SCHIP enrollment and 1 year after enrollment of 2290 children who had an enrollment start date in New York State's SCHIP between November 2000 and March 2001, stratified by race/ethnicity (non-Hispanic white, non-Hispanic black, and Hispanic). The main outcome measures were usual source of care (USC), preventive care use, unmet needs, patterns of USC use, and parent-rated quality of care before versus during SCHIP.Results. Children were white (25%), black (31%), or Hispanic (44%); 62% were uninsured >12 months before SCHIP. Before SCHIP, a greater proportion of white children had a USC compared with black or Hispanic children (95%, 86%, and 81%, respectively). Nearly all children had a USC during SCHIP (98%, 95%, and 98%, respectively). Before SCHIP, black children had significantly greater levels of unmet need relative to white children (38% vs 27%), whereas white and Hispanic children did not differ significantly (27% vs 29%). During SCHIP, racial/ethnic disparities in unmet need were eliminated, with unmet need at 19% for all 3 racial/ethnic groups. Before SCHIP, more white children made all/ most visits to their USC relative to black or Hispanic children (61%, 54%, and 34%, respectively); all improved during SCHIP with no remaining disparities (87%, 86%, and 92%, respectively). Parent-rated visit quality improved for all groups, but preexisting racial/ethnic disparities remained during SCHIP, with improved yet relatively lower levels of satisfaction among parents of Hispanic children. Sociodemographic and health system factors did not explain disparities in either period.Conclusions. Enrollment in SCHIP was associated with (1) improvement in access, continuity, and quality of care for all racial/ethnic groups and (2) reduction in preexisting racial/ethnic disparities in access, unmet need, and continuity of care. Racial/ethnic disparities in quality of care remained, despite improvements for all racial groups. Sociodemographic and health system factors did not add to the understanding of racial/ethnic disparities. SCHIP improves care for vulnerable children and reduces preexisting racial/ethnic disparities in health care. R acial/ethnic disparities in health care delivery are associated with adverse hea...

show abstract

Section: Analysesmentioning

confidence: 99%

Section: E698mentioning

confidence: 99%

Reduction in Racial and Ethnic Disparities After Enrollment in the State Children's Health Insurance Program

Shone¹,

et al. 2005

View full text Add to dashboard Cite

show abstract

“…Estimations are carried out with the software package Stata, release 8. 10 An overview of the models and their estimators can be found in Pendergast et al (1996) or Hosmer and Lemeshow (2000).…”

mentioning

confidence: 99%

Estimating Probabilities of Default for German Savings Banks and Credit Cooperatives

Porath

2006

Schmalenbach Bus Rev

View full text Add to dashboard Cite

A healthy banking system is a fundamental condition for financial stability. When assessing the riskiness of the banking system, analysts often restrict their focus to large banks. This may create a distorted picture in countries like Germany with fragmented banking systems. In Germany, savings banks and cooperative banks taken together are important players in the market. However, little is known about their default risk. The reason is that these banks usually resolve financial distress within their own organisations, which means defaults are not observable from the outside. In this paper we use a new dataset which contains information about financial distress and financial strength of all German savings banks and cooperative banks. The data have been gathered by the Deutsche Bundesbank for microprudential supervision and have never before been exploited for macroprudential purposes. We use the data to identify the main risk drivers. To this end we estimate a default prediction model (hazard model). A second goal of the paper is to analyse the impact of macroeconomic information for forecasting banks' defaults. Recent findings for the USA have cast some doubt on the usefulness of macroeconomic information for banks' risk assessment. Contrary to recent literature, we find that macroeconomic information significantly improves default forecasts. Keywords:bank failure, default probability, time-discrete hazard rate Non-technical summaryThere is little evidence on the default risk of German savings banks and cooperative banks, although they constitute an integral part of the German banking system. Our aim is to develop a statistical system which estimates probabilities of default (PDs) for savings banks and cooperative banks. We also try to find evidence for the importance of macroeconomic developments for the estimation of PD. Since we adopt a prudential supervisory perspective, default is defined as any event that jeopardises the bank's viability as a going concern.The paper uses a dataset from the Deutsche Bundesbank that combines default events with balance sheet information, audit reports, and macroeconomic variables. The data is used to estimate a discrete hazard model. Hazard models are particularly suited for our purposes because they simultaneously include macroeconomic and microeconomic data.Our main findings are as follows.

show abstract

“…This method can be easily adapted for binary outcomes. A more comprehensive review of methods for analysing clustered binary data is presented by Pendergast et al (1996) and Aerts et al (2002).…”

Section: Introductionmentioning

confidence: 99%

A comparative study of models for correlated binary data with applications to health services research

Dilba

Aerts

2005

SEJS

View full text Add to dashboard Cite

Various methods of modeling correlated binary data are compared as applied to data from health services research. The methods include the standard logistic regression, a simple adjustment of the standard errors of logistic regression by a single inflator, the weighted logistic regression, the generalized estimating equation, the beta-binomial model, and two proposed bootstrap methods. First, these approaches are compared for a fixed set of predictors by individual tests of significance. Next, several subsets of predictors are compared through the AIC criterion, whenever applicable.

show abstract

A Survey of Methods for Analyzing Clustered Binary Response Data

Cited by 243 publications

References 0 publications

Reduction in Racial and Ethnic Disparities After Enrollment in the State Children's Health Insurance Program

Reduction in Racial and Ethnic Disparities After Enrollment in the State Children's Health Insurance Program

Estimating Probabilities of Default for German Savings Banks and Credit Cooperatives

A comparative study of models for correlated binary data with applications to health services research

Contact Info

Product

Resources

About