Multiple imputation of a randomly censored covariate improves logistic regression analysis

Randomly censored covariates arise frequently in epidemiologic studies. The most commonly used methods, including complete case and single imputation or substitution, suffer from inefficiency and bias. They make strong parametric assumptions or they consider limit of detection censoring only. We employ multiple imputation, in conjunction with semi-parametric modeling of the censored covariate, to overcome these shortcomings and to facilitate robust estimation. We develop a multiple imputation approach for randomly censored covariates within the framework of a logistic regression model. We use the non-parametric estimate of the covariate distribution or the semi-parametric Cox model estimate in the presence of additional covariates in the model. We evaluate this procedure in simulations, and compare its operating characteristics to those from the complete case analysis and a survival regression approach. We apply the procedures to an Alzheimer's study of the association between amyloid positivity and maternal age of onset of dementia. Multiple imputation achieves lower standard errors and higher power than the complete case approach under heavy and moderate censoring and is comparable under light censoring. The survival regression approach achieves the highest power among all procedures, but does not produce interpretable estimates of association. Multiple imputation offers a favorable alternative to complete case analysis and ad hoc substitution methods in the presence of randomly censored covariates within the framework of logistic regression.     

Frequent Card Playing and Pathological Gambling: The Utility of the Georgia Gambling Task and Iowa Gambling Task for Predicting Pathology

The current investigation examined performance on two laboratory-based gambling tasks, the Georgia Gambling Task (GGT; Goodie, 2003. The effects of control on betting: Paradoxical betting on items of high confidence with low value. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29, 598–610) and the Iowa Gambling Task (IGT; Bechara, Damasio, Damasio, & Anderson, 1994. Insensitivity to future consequences following damage to human prefrontal cortex. Cognition, 50, 7–15), as well as self-reported markers of gambling pathology using the Diagnostic Interview for Gambling Severity (DIGS; Winters, Specker, & Stinchfield, 2002. The downside: Problem and pathological gambling (pp. 143–148). Reno, NV: University of Nevada, Reno) among a sample of undergraduate students who are frequent card players. Two hundred twenty-one participants (55 female and 166 male; mean age 19.21 years) who self-classified as playing cards at least once per month completed these measures. Performance on GGT and IGT systematically related to gambling-related pathology in several ways. Overconfidence and bet acceptance on the GGT, and myopic focus on reward on the IGT, predicted gambling related pathology. GGT and IGT performance correlated with each other, but both contributed independently to predicting gambling pathology. Card playing frequency predicted gambling pathology but not GGT or IGT performance. Discussion focuses on the role of biases of judgment and risky decision making in pathological gambling.     

Measuring and evaluating school performance: an investigation of status and growth-based achievement indicators

Several school performance indices were evaluated using achievement data from multiple longitudinally matched middle school student cohorts in a large district in the southwestern United States. Measures designed to index the within-cohort achievement status and growth of students as well as the status and growth-based changes that occur over years between different student cohorts were studied. Within- and between-cohort status and growth-based estimates of school performance were generated by applying simple and relatively complex statistical models to the longitudinal time-series data obtained on students. Results indicated that within-cohort status-based estimates were closely related to student demographics and between-cohort estimates were associated with cohort enrollment size and initial performance level. These results suggest that schools may often be rewarded and penalized on the basis of factors over which school personnel have limited or no control. Implications for the measurement and evaluation of school performance are discussed.     

Does housework matter anymore? The shifting impact of housework on economic inequality

In recent years, American women’s housework time has declined while American men’s housework time has risen. We examine how these changes have affected economic inequality in the United States. Using time-diary data from the Time Use in Economic and Social Accounts, 1975–1976 (N = 1,484) and the American Time Use Survey, 2003 (N = 5,534), we value adults’ housework using two alternative methodologies and assess its influence on households’ real access to goods and services in both years. Results suggest that housework reduces economic inequality in both years. But, between 1975–1976 and 2002–2003, overall economic inequality rose largely because of the growing wage inequality and also, in part, because of growth in housework inequality. Socio-demographic change partially inhibited the overall growth in economic inequality.

Caregiving Load and Respite Service Use: A Comparison between Older Caregivers and Younger Caregivers

Study aims were to explore differences in predictors of respite care use between older (aged 65+) and younger caregivers (aged 18–64 years), and associations between caregiving load and respite care use using multivariate logistic regression analysis and unpaid caregiver (n= 10,500) data from the 2009 California Health Interview Survey. Caregiving load comprised number of care recipients, weekly hours in caregiving, and caregiving duration. Variables with a significant association with respite care for older caregivers were female gender, income, and health insurance. For younger caregivers, respite care use associations were with ethnicity, caregiving relationship, education, and availability of substitute help.     

Categorizing Risks for Risk Ranking

Any practical process of risk ranking must group hazards into a manageable number of categories. Defining such categories requires value choices that can have important implications for the rankings that result. Most risk-management organizations will find it useful to begin defining categories in terms of environmental loadings or initiating events. However, the resulting categories typically need to be modified in light of other considerations. Risk-ranking projects can benefit from considering several alternative categorization strategies and drawing upon elements of each in developing their final categorization of risks. In principle, conducting multiple ranking exercises by using different categorizations could be interesting and useful. In practice, agencies are unlikely to have either the resources or patience to do this, but other groups in society might. Done well, such additional independent rankings could add valuable inputs to democratic risk-management decision making.     

The development of a universal approach to testing of walkway slip resistance in the U.S

The measurement of walkway slip resistance is required to assess the risk of slipping. Unlike other countries, the U.S. has not decided upon a single approach to the measurement of slip resistance. Several types of tribometers are available to measure slip resistance however, the measured value of the slip resistance of a given surface has been found to be tribometer specific. The introduction of ASTM International Standard F2508 Standard Practice for Validation and Calibration of Walkway Tribometers Using Reference Surfaces, has produced a method which allows validation of each type of tribometer and the values generated during testing.