Original Research
A latent class analysis of job satisfaction and turnover among practicing pharmacists

https://doi.org/10.1016/j.sapharm.2009.03.002Get rights and content

Abstract

Background

Research on job satisfaction and turnover using latent class analysis (LCA) has been conducted in other disciplines. LCA has seldom been applied to social pharmacy research and may be especially useful for examining job situation constructs in pharmacy organizations.

Objective

The objective of the study was to determine the probability of turnover among practicing pharmacists using LCA.

Methods

Using a cross-sectional descriptive design, 2400 randomly selected pharmacists with active licenses in Florida were surveyed. A model was created using LCA, then fit indices were used to determine whether underlying “job satisfaction clusters” were present. Once identified, these clusters along with the covariate practice site were modeled on a distal outcome turnover.

Results

A 5-class model appeared to best fit the data: a “pseudo-satisfied” class that contained 8% of the sample, a “career-goal” class that contained 11% of the sample, a “satisfied class” that contained 44% of the sample, a “job-expectation” class that contained 3% of the sample, and an “unsatisfied class” that contained 17% of the sample. In terms of predicting the distal outcome “turnover,” the calculated odds ratios indicate that compared with class 3 or the satisfied group, class 2 was 14 times more likely, class 4 was 17 times more likely, and class 5 was 26 times more likely to state that they do not intend to be employed with their current employer 1 year from now.

Conclusion

The LCA method was found to be effective for finding relevant subgroups with a heterogeneous at-risk population for turnover. Results from the analysis indicate that job satisfaction may be parsed into smaller, more interpretable and useful subgroups. This result holds great promise for practitioners and researchers, alike.

Section snippets

Background

Latent class analysis (LCA) is a statistical method for finding subtypes of related cases (latent classes) from multivariate categorical data.1 The most common use of LCA is to discover case subtypes (or confirm hypothesized subtypes) based on multivariate categorical data.1, 2, 3, 4 LCA is well suited to many health applications for which one wishes to identify disease subtypes or diagnostic subcategories.1, 2, 3, 4 LCA models do not rely on traditional modeling assumptions (normal

Objectives

Research using LCA, job satisfaction, and turnover has been conducted in other disciplines.13, 14 Using LCA, Shockey observed that 4 classes explained his job satisfaction data quite well.13 The conclusion was that the true prevalence of job satisfaction is constant over groups once response errors have been accounted for.

In reference to job satisfaction and turnover, researchers have found a causal link between satisfaction, organizational commitment, and turnover.14 For example, Currivan

Methods

The study's theoretical framework is based on the relationship between job satisfaction and organizational commitment in models of employee turnover.15 Furthermore, included in the model is the covariate “practice site.” Research shows a strong relationship between practice site and job satisfaction.16 Demonstrating the causal order between job satisfaction, practice site, and turnover requires specifying the individual characteristics and workplace structures as determinants of turnover.16 A

Results

The adjusted overall response rate was 23% (n = 533/2353). For the purpose of this study, which was focused on actively practicing pharmacists, a total of 429 surveys were usable and were included in the final data analysis. There were no significant differences between the characteristics of early and late respondents. The respondents' average age was approximately 45 years and more than half (55%) were female. The majority worked full time (89%), practiced in a retail setting (57%), and

Discussion

This study demonstrated the utility of LCA in the measurement of job satisfaction. This LCA arrived at a statistically sound and relatively straightforward interpretable number of classes. The basic idea underlying LCA is a very simple one: some of the parameters of a postulated statistical model differ across previously unrecognized subgroups.19 These subgroups form the categories of a categorical latent variable. Outside social sciences, latent class models are often referred to as finite

Conclusion

The LCA method was found to be effective for finding relevant subgroups with a heterogeneous at-risk population for turnover. The minimal cross-classification between classes may be the result of the small sample size. Furthermore, in reviewing the model fit indices, a 4-class solution was certainly possible, although not as strong as a 5-class solution. Hence, it is possible that the similarity in responses to specific items (ie, items 1 and 5) make it difficult for the model to perfectly

References (19)

There are more references available in the full text version of this article.

Cited by (0)

1

Tel.: +1 954 262 1332; fax: +1 954 262 2278.

View full text