Relaxing the rule of ten events per variable in logistic and Cox regression.

Department of Epidemiology and Biostatistics, University of California-San Francisco, 185 Berry Street, San Francisco, CA 94107, USA.
American Journal of Epidemiology (Impact Factor: 4.98). 04/2007; 165(6):710-8. DOI: 10.1093/aje/kwk052
Source: PubMed

ABSTRACT The rule of thumb that logistic and Cox models should be used with a minimum of 10 outcome events per predictor variable (EPV), based on two simulation studies, may be too conservative. The authors conducted a large simulation study of other influences on confidence interval coverage, type I error, relative bias, and other model performance measures. They found a range of circumstances in which coverage and bias were within acceptable levels despite less than 10 EPV, as well as other factors that were as influential as or more influential than EPV. They conclude that this rule can be relaxed, in particular for sensitivity analyses undertaken to demonstrate adequate control of confounding.

1 Follower
  • [Show abstract] [Hide abstract]
    ABSTRACT: Most epidemiology textbooks that discuss models are vague on details of model selection. This lack of detail may be understandable since selection should be strongly influenced by features of the particular study, including contextual (prior) information about covariates that may confound, modify, or mediate the effect under study. It is thus important that authors document their modeling goals and strategies and understand the contextual interpretation of model parameters and model selection criteria. To illustrate this point, we review several established strategies for selecting model covariates, describe their shortcomings, and point to refinements, assuming that the main goal is to derive the most accurate effect estimates obtainable from the data and available resources. This goal shifts the focus to prediction of exposure or potential outcomes (or both) to adjust for confounding; it thus differs from the goal of ordinary statistical modeling, which is to passively predict outcomes. Nonetheless, methods and software for passive prediction can be used for causal inference as well, provided that the target parameters are shifted appropriately.
    Annual Review of Public Health 03/2015; 36:89-108. DOI:10.1146/annurev-publhealth-031914-122559 · 6.63 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The norm of silent acquiescence on the Supreme Court was thought to have been eviscerated in the twentieth century by certain institutional reforms and the rise of dissenting opinions. Given that silent acquiescence is difficult to observe, however, the extent to which this norm persists on the modern Court remains unclear. To overcome this observational difficulty, I analyze private memoranda exchanged by justices who served during the Burger Court. The empirical results suggest that silent acquiescence is a rare but regular occurrence on the modern Court, and is more likely to occur in comparatively unimportant cases. Notwithstanding institutional and personnel changes that limited silent acquiescence and precipitated an increase in dissenting opinions, it appears that the practice of go-along voting continued throughout the Burger Court. The results have implications for our understanding of separate opinion writing, judicial decision-making, and judicial legitimacy.
    Justice System Journal, The 11/2014; 36(1):3-19. DOI:10.1080/0098261X.2014.969854 · 0.50 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: A common use of language tests is to support decisions about examinees such as placement into appropriate classes. Research on placement testing has focused on English for Academic Purposes (EAP) in higher education contexts. However, there is little research exploring the use of language tests to place students in English as a Second Language (ESL) support classes in secondary education. The present study examined the relationship between secondary school students' language test scores from a standardized English-language test and the placement of these students into ESL classes by their language teachers. Ninety-two ESL students in two English-medium schools took TOEFL (R) Junior (TM) Standard. For the same students, data collection also included teachers' judgments regarding the ESL classes the students should attend. Strong correlations between test scores and the teacher-assigned ESL levels were found. Moreover, the results from the logistic regression analysis indicated a great degree of overlap between the teacher-assigned ESL levels and the levels predicted from the TOEFL Junior Standard scores. The findings of this study provide some preliminary evidence to support the use of TOEFL Junior Standard as an initial screening tool for ESL placement. The limitations and implications of these findings for ESL placement decisions in secondary education are also discussed.
    Language Testing 04/2013; 31(2):223-239. DOI:10.1177/0265532213499750 · 1.11 Impact Factor