Patrick C Hardigan1, Manuel J Carvajal. 1. Statistical Consulting Center, Nova Southeastern University, Fort Lauderdale, Fla 33328, USA. patrick@nova.edu
Abstract
OBJECTIVE: To demonstrate the application of Rasch analysis in the study of job satisfaction among practicing pharmacists. DESIGN: Cross-sectional survey study. SETTING: Community independent, community chain, hospital, and other pharmacies in the United States in fall 2005. PARTICIPANTS: 790 practicing pharmacists. INTERVENTION: Mailed survey based on past studies of job satisfaction. MAIN OUTCOME MEASURE: Job satisfaction as measured using the Rasch rating scale model (which considers responses on Likert scales as categorical data) and traditional estimation techniques (which assume a continuum among responses on these scales by analyzing data as interval). RESULTS: A gain in precision for the Rasch technique was observed for the constructs distributive justice, job autonomy, job ambiguity, role conflict, positive affectivity, job resources, and supervisory support. No gain in precision was found for the constructs negative affectivity and work involvement. Traditional estimates offered more precision for the constructs job repetition, promotional opportunity, and workload. CONCLUSION: Compared with traditional estimates, Rasch estimates provided more precise scores of job satisfaction on certain subscales, especially those whose mean scores were large or responses more diverse.
OBJECTIVE: To demonstrate the application of Rasch analysis in the study of job satisfaction among practicing pharmacists. DESIGN: Cross-sectional survey study. SETTING: Community independent, community chain, hospital, and other pharmacies in the United States in fall 2005. PARTICIPANTS: 790 practicing pharmacists. INTERVENTION: Mailed survey based on past studies of job satisfaction. MAIN OUTCOME MEASURE: Job satisfaction as measured using the Rasch rating scale model (which considers responses on Likert scales as categorical data) and traditional estimation techniques (which assume a continuum among responses on these scales by analyzing data as interval). RESULTS: A gain in precision for the Rasch technique was observed for the constructs distributive justice, job autonomy, job ambiguity, role conflict, positive affectivity, job resources, and supervisory support. No gain in precision was found for the constructs negative affectivity and work involvement. Traditional estimates offered more precision for the constructs job repetition, promotional opportunity, and workload. CONCLUSION: Compared with traditional estimates, Rasch estimates provided more precise scores of job satisfaction on certain subscales, especially those whose mean scores were large or responses more diverse.
Authors: Vincent Chung; Marc Chong; Lau Chun Hong; Polly H X Ma; Samuel Y S Wong; Sian M Griffiths Journal: Chin J Integr Med Date: 2012-07-22 Impact factor: 1.978
Authors: Richard J Holden; Neal R Patel; Matthew C Scanlon; Theresa M Shalaby; Judi M Arnold; Ben-Tzion Karsh Journal: Res Social Adm Pharm Date: 2010-02-11