Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
Modelling Growth Trajectories of Job Satisfaction with Total Pay Using the Muthén-Roy Pattern-Mixture Model Approach
Maria de Fatima Salgueiro

Last modified: 2013-06-16

Abstract


Modelling longitudinal survey data often requires dealing with missing data. If a missing at random mechanism cannot be assumed, a missing not at random approach should be considered. This paper proposes an extended pattern-mixture model approach to modeling longitudinal growth trajectories of job satisfaction with total pay. A latent growth mixture model with latent trajectory classes is considered, in combination with a latent class variable that summarizes the missing data patterns. Longitudinal survey data from the British Household Panel Survey on job satisfaction with total pay for the period 2002 to 2008 are used. The statistical modeling is conducted with the program Mplus.

Full Text: PDF