Employee well-being profiles and service quality: a unit-level analysis using a multilevel latent profile approach

Miriam Benitez, Riccardo Peccei*, Francisco J. Medina

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)

Abstract

This study is designed to contribute to a better theoretical and empirical understanding of the unit-level relationship between employee well-being (job satisfaction and burnout, in the form of emotional exhaustion and cynicism) and service quality in two ways. First is by adopting a person-oriented approach to identify distinct well-being profiles as a way of capturing the conjoint effect of satisfaction and burnout on service quality. Second is by employing an adapted form of multilevel latent profile analysis (MLPA) to examine the extent to which the distribution of employee well-being profiles in a work group affects the level of service quality provided by the group. These issues were examined using combined employee-customer data based on a sample of 396 employees and 1233 customers from 91 restaurant and reception units in 42 hotels in Spain. In line with theoretical expectations and consistent with a win-win model of the unit level relationship between employee well-being and service quality, the results showed that work groups with a greater proportion of employees with high levels of job satisfaction combined with low levels of burnout (emotional exhaustion and cynicism), significantly outperformed work groups with a less favourable internal distribution of well-being profiles. Theoretical and practical implications are discussed.

Original languageEnglish
Pages (from-to)859-872
JournalEUROPEAN JOURNAL OF WORK AND ORGANIZATIONAL PSYCHOLOGY
Volume28
Issue number6
Early online date19 Oct 2019
DOIs
Publication statusPublished - 2019

Keywords

  • Well-being profiles
  • multilevel
  • service quality

Fingerprint

Dive into the research topics of 'Employee well-being profiles and service quality: a unit-level analysis using a multilevel latent profile approach'. Together they form a unique fingerprint.

Cite this