Medicine and Health Sciences Commons^™

A Positive Trait Item Response Model, Joseph F. Lucke

Joseph Lucke

All current models from item response theory (IRT) assume the latent trait follows a standard normal distribution. While this assumption is appropriate for traits such as ability or attitude, it creates both conceptual and technical problems traits such as addiction (alcohol, drugs, gambling). The distribution of an addiction trait is better assumed to be anchored at zero (no addiction) and positively skewed. A small change to the usual IRT model yields a class of positive-trait item response models (PTIRMs). I discuss PTIRMs and present one model in detail, including item characteristic curves and item information curves. I present an example …

A Critique Of The False-Positive Report Probability, Joseph Lucke

Joseph Lucke

The false positive report probability (FPRP) was proposed as a Bayesian prophylactic against false reports of significant associations. Unfortunately, the derivation of the FPRP is unsound. A heuristic derivation fails to make its point, and a formal derivation reveals a probabilistic misrepresentation of an observation. As a result, the FPRP can yield serious inferential errors. In particular, the FPRP can use an observation that is many times more likely under the null hypothesis than under the alternative to infer that the null hypothesis is far less probable than the alternative. Contrary to its intended purpose, the FPRP can promote false …

Benchmarking Patient Outcomes, Ellen B. Rudy, Joseph F. Lucke, Gayle R. Whitman, Lynda J. Davidson

Joseph Lucke

Purpose: To examine the usefulness of three types of benchmarking for interpreting patient outcome data.

Design: This study was part of a multiyear, multihospital longitudinal survey of 10 patient outcomes. The patient outcome used for this methodologic presentation was central line infections (CLI). The sample included eight hospitals in an integrated healthcare system, with a range in size from 144 to 861 beds. The unit of analysis for CLI was the number of line days, with the CLI rate defined as the number of infections per 1,000 patient-line days per month.

Methods: Data on each outcome were collected at the …