Using a hierarchical logistic regression model to establish the validity of an *examination with a dichotomous criterion: *Policy implications for nursing education.
Item
-
Title
-
Using a hierarchical logistic regression model to establish the validity of an *examination with a dichotomous criterion: *Policy implications for nursing education.
-
Identifier
-
AAI9959226
-
identifier
-
9959226
-
Creator
-
Schmidt, Amy Elizabeth.
-
Contributor
-
Adviser: David Rindskopf
-
Date
-
2000
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Education, Educational Psychology | Health Sciences, Nursing | Health Sciences, Education | Education, Tests and Measurements
-
Abstract
-
This paper presents a study that employed two-level hierarchical logistic regression models to establish the degree to which scores on the Diagnostic Readiness Test (DRT) predict success or failure on the National Council Licensure Examination for Registered Nurses (NCLEX-RN), and to determine how this relationship varies as a function of student-level and school-level variables. At the first level of these models, the predictors of NCLEX-RN performance included individual DRT, and Nursing Pre-Admission Examination scores. At the second level, or school level, type of nursing program was used to explain how the relationship between DRT and NCLEX-RN varies from school to school.;For the final model, Empirical Bayes (EB) estimates of the parameters were obtained for each school, and were compared to the classical logistic regression coefficients. In addition, the overall predictive validity of the DRT was obtained, as well as the differential predictive validity for each type of nursing program. Results indicated that DRT scores were the only significant individual level predictors of NCLEX-RN performance, and this relationship varied significantly by type of nursing program. In addition, the EB estimates were superior to the estimates obtained from the traditional logistic regression analysis, in that none of the EB slope estimates were negative.;Because this is the first systematic attempt to explain school-level variability in predictors of NCLEX-RN performance, policy implications for nursing education are also discussed and specific recommendations are presented. These recommendations address, for example, ways in which this type of hierarchical logistic regression analysis can benefit nursing education by identifying factors that influence NCLEX-RN performance for different types of schools. Implications for the further use of hierarchical analysis in psychometric applications are also discussed, as is the appropriateness of traditional validity models in evaluating the validity and utility of a diagnostic measure.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
