Can the job demand-control-(support) model predict disability support worker burnout and work engagement?

ABSTRACT

Background Research shows that up to 43% of disability support workers (DSWs) report poor psychosocial work outcomes (e.g., stress, job burnout, low job satisfaction). This study examined whether the job demand-control-(support) model offers a valid explanation of DSW burnout and work engagement.

Method 325 DSWs completed online measures of burnout, work engagement, workload, job control, and supervisor or colleague support.

Results Significant three-way interactions between workload, control and colleague support were found for emotional exhaustion and personal accomplishment (burnout), and vigour (work engagement). High workload, low job control and low colleague support was related to higher burnout and lower work engagement, and high colleague support or job control reduced the impact of workload on these outcomes.

Conclusions Given the promising findings in relation to the job demand-control-(support) model, organisations looking to enhance DSW wellbeing in the workplace should address issues around job control, workload and support in combination as opposed to separately.

KEYWORDS:

• disability support workers
• job demand-control-support model
• intellectual disability
• job burnout
• work engagement

Acknowledgments

Jane Geltch and Sarah Fordyce from National Disability Services are acknowledged for their support with participant recruitment. Sarah Bourchier and Elise Jones are also acknowledged for their literature review support.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 A large effect size was deemed appropriate for sample size calculation based on R² statistics for JDC related analyses reported by Devereux et al. (2009b) and Gray-Stanley et al. (2010).

2 In the interest of limiting the size of the article, full statistical results are not provided.

3 The study has adequate power to test these expanded models. Using Cohen’s (1992) sample size guides, for a model with eight predictor variables (those that will include gender) to observe a large effect size at power = .80 and α = .05, a sample of 50 participants is needed. For a model with 13 predictors (those that will include employment setting), a sample of 63 participants is needed.

4 In the interest of limiting the size of the article, full statistical results are not provided.

5 R² values were converted to r values $(r=\sqrt[2]{R^2})$ in order to make this comparison.

6 In the interest of limiting the size of the article, full statistical results for these models are not provided.

Additional information

Funding

This study was supported by the University of Queensland’s New Staff Research Start-Up Fund.