Building work engagement: A systematic review and meta‐analysis investigating the effectiveness of work engagement interventions

• DOI: http://doi.org/10.1002/job.2167
• Author: Malcolm Patterson,Caroline Knight,Jeremy Dawson
• Published date: 01 July 2017
• Date: 01 July 2017

Building work engagement: A systematic review

and meta-analysis investigating the effectiveness

of work engagement interventions

CAROLINE KNIGHT*, MALCOLM PATTERSON AND JEREMY DAWSON

Institute of Work Psychology,University of Sheffield Management School, Sheffield, South Yorkshire, UK

Summary Low work engagement may contribute towards decreased well-being and work performance. Evaluating,

boosting and sustaining work engagement are therefore of interest to many organisations. However, the evi-

dence on which to base interventions has not yet been synthesised. A systematic review with meta-analysis

was conducted to assess the evidence for the effectiveness of work engagement interventions. A systematic

literature search identified controlled workplace interventions employing a validated measure of work

engagement. Most used the Utrecht Work Engagement Scale (UWES). Studies containing the relevant quan-

titative data underwent random-effects meta-analyses. Results were assessed for homogeneity, systematic

sampling error, publication bias and quality. Twenty studies met the inclusion criteria and were categorised

into four types of interventions: (i) personal resource building; (ii) job resource building; (iii) leadership train-

ing; and (iv) health promotion. The overall effect on work engagement was small, but positive, k= 14, Hedges

g= 0.29, 95%-CI = 0.12–0.46. Moderator analyses revealed a significant result for intervention style, with a

medium to large effect for group interventions. Heterogeneity between the studies was high, and the success

of implementation varied. More studies are needed, and researchers are encouraged to collaborate closely

with organisations to design interventions appropriate to individual contexts and settings, and include evalu-

ations of intervention implementation. © 2016 The Authors. Journal of Organizational Behavior published

by John Wiley & Sons, Ltd.

Keywords: work engagement; interventions; meta-analysis; systematic review; intervention implementation

Introduction

Work engagement is currently a popular topic within many organisations, given its association with employee well-

being and performance (e.g. Christian, Garza, & Slaughter, 2011; Halbesleben, 2010). Evaluating, boosting and sus-

taining work engagement are therefore a prime concern of many organisations, and many studies have investigated

the possible antecedents and consequences of engagement (e.g. Halbesleben, 2010; Crawford, LePine, & Rich,

2010), leading researchers to consider the field sufficiently well developed to warrant the development and testing

of work engagement interventions (e.g. Leiter & Maslach, 2010). However, the evidence on which to base interven-

tions is limited, although a variety of intervention studies are emerging (Biggs, Brough, & Barbour, 2014;

Ouweneel, Le Blanc, & Schaufeli, 2013). No study has yet assessed the effectiveness of these interventions;

however, it is hoped that doing so will stimulate debate and direct future research and practice. The aim of this study

is therefore to conduct a narrative systematic review and meta-analysis of the evidence for the effectiveness of

controlled work engagement interventions.

*Correspondence to: Caroline Knight, Institute of Work Psychology, University of Sheffield Management School, Conduit Rd, Sheffield, South

Yorkshire, S10 1FL, UK. E-mail: caroline.knight@sheffield.ac.uk

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in

any medium, provided the original work is properly cited.

© 2016 The Authors. Journal of Organizational Behavior

published by John Wiley & Sons, Ltd.

Received 28 April 2016

Revised 13 October 2016, Accepted 9 November 2016

Journal of Organizational Behavior, J. Organiz. Behav. 38, 792–812 (2017)

Published online 13 December 2016 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/job.2167

Research Article

Work engagement

Kahn (1990) originally pioneered the concept of employee engagement, proposing that engaged employees are

physically, cognitively and emotionally involved in their work roles, and experience a sense of meaning (reward

for investing in role performance), psychological safety (a sense of trust and security at work) and availability (a

sense of having the physical and psychological resources necessary for the job). Saks (2006) developed this view

by distinguishing between job and organisational engagement to reflect the different roles of employees. Maslach

and Leiter (1997) approached engagement from the field of burnout, characterising it in terms of high energy,

involvement and efficacy, the polar opposite of burnout (exhaustion, cynicism and inefficacy), and therefore mea-

surable using the Maslach Burnout Inventory (MBI; Maslach & Jackson, 1981). Schaufeli, Salanova, Gonzalez-

Roma, and Bakker (2002) refuted this, arguing that while engagement is the positive antipode to burnout, it is

a separate, distinct concept, and therefore cannot be measured on a burnout scale. They defined work engagement

as a state of mind characterised by vigour, dedication and absorption (Schaufeli et al., 2002), and developed the

Utrecht Work Engagement Scale (UWES) to measure it. Vigour refers to high energy and mental resilience while

working, dedication to being intensely involved in work tasks and experiencing an associated sense of signifi-

cance, enthusiasm, and challenge, and absorption to a state of full concentration on work and positive engrossment

in it.

Schaufeli et al. (2002) view of engagement as a distinct concept from burnout was contested by Cole, Walter,

Bedeian, and O’Boyle (2012), who found in their meta-analysis, involving 50 independent samples, that burnout

and engagement were highly correlated, similarly associated with correlates, and that controlling for burnout re-

duced the effect sizes of engagement substantially. They concluded that whether burnout and engagement can be

viewed as separate dimensions was questionable.

Further academic (e.g. Crawford et al., 2010; May, Gilson, & Harter, 2004), and lay and practitioner (e.g.

MacLeod & Clarke, 2009; Robertson-Smith & Markwick, 2009) definitions and measures of engagement exist,

and scholars have also questioned the existence of engagement, arguing that it is redundant with other, established

job attitudes such as job satisfaction, organisational commitment and job involvement (for a good review, see Macey

& Schneider, 2008; see also Byrne, Peters, & Weston, 2016; Christian et al., 2011). Questions have also arisen over

the factorial validity of the UWES, with some studies suggesting that a three-factor model is superior to a one factor,

unidimensional model (e.g. Bakker & Demerouti, 2008; Schaufeli, Bakker, & Salanova, 2006), and others suggest-

ing that the three factor structure is ambiguous (e.g. Sonnentag, 2003), or that the models are equivalent (Hallberg &

Schaufeli, 2006). The field is clearly divided over the meaning of engagement and how best to measure it. Never-

theless, Schaufeli et al. (2002) perspective appears to be the most popular and well researched to date (Hakanan

& Roodt, 2010), and tends to underlie engagement interventions.

The key model underlying Schaufeli et al. (2002) perspective is the Job Demands–Resources model (JD-R;

Bakker & Demerouti, 2008; Bakker & Demerouti, 2007). This proposes that work engagement is driven, either in-

dependently or together, by both job and personal resources. Job resources refer to physical, social or organisational

aspects of the job (e.g. feedback, social support, development opportunities) that can reduce job demands (e.g. work-

load, emotional and cognitive demands), help employees to achieve work goals, and stimulate personal learning and

development. Personal resources refer to ‘positive self-evaluations that are linked to resiliency and refer to individ-

uals’sense of their ability to control and impact upon their environment successfully’(Bakker & Demerouti, 2008

p.5). These include self-esteem, self-efficacy, resilience and optimism. The motivating potential of job and personal

resources are proposed to lead to positive individual and organisational outcomes such as work engagement, well-

being and performance, whereas few resources and high work demands are proposed to lead to poor health

outcomes, such as burnout, stress and depression, as well as turnover, sickness absence and poor performance. This

model has been supported by numerous studies (e.g. Xanthopoulou, Bakker, Demerouti, & Schaufeli, 2009;

Hakanen, Bakker, & Demerouti, 2005; Simbula, Guglielmi, & Schaufeli, 2011), including meta-analyses (Crawford

et al., 2010; Halbesleben, 2010; Nahrgang, Morgeson, & Hofmann, 2011), all of which have served to advance the

model and work engagement theory.

EFFECTIVENESS OF WORK ENGAGEMENT INTERVENTIONS 793

© 2016 The Authors. Journal of Organizational Behavior

published by John Wiley & Sons, Ltd.

J. Organiz. Behav. 38, 792–812 (2017)

DOI: 10.1002/job