Components of Work-related Wellbeing
MetadataShow full metadata
Background Assessing wellbeing in the workplace is becoming increasingly important as organisations recognise the influence of wellbeing on key work-related outcomes (Jarden & Jarden, 2017). However, the development of such assessments in practice is often haphazard (Donaldson, Dollwet, & Rao, 2015). Many of the available wellbeing assessments are comprised of items and scales drawn from various sources that are not based on any theoretical models. The Wellbeing360™ is one such assessment tool used in practice. It is unclear whether the items within the Wellbeing360™ are psychometrically sound and reflect a unidimensional construct. Because the focus of this thesis will be on the components of work-related wellbeing, the first aim of this research will be to test the underlying structure of the Wellbeing360™ work-related items. The second aim will be to test the reliability and validity of any resulting factors/scales. Determining the underlying structure of the work-related items will also allow the predictors of work-related wellbeing to be determined. Understanding the most significant predictors of work-related wellbeing will highlight any key components of the work environment that may be important for improving employee wellbeing. Thus, the third aim of this research will be to determine the most significant predictors of work-related wellbeing. Methods A secondary data set consisting of employee responses to the Wellbeing360™ was used. Participants were aged 18 years or over from 20 different organisations (in New Zealand and Australia) spanning nine industries. The questionnaire consisted of 116 items measuring aspects of wellbeing, health and lifestyle variables, and socio-demographics. The specific variables of interest concerned work-related wellbeing (Work-related Affect, Job Resources, and Job Demands), Age, Gender, Country of Birth, Resilience, Flourishing, Depression, Anxiety and Stress. In Paper 1, 20 of the work-related wellbeing items were subjected to Exploratory Factor Analysis (EFA) to determine their underlying structure. The resulting scales were tested for internal consistency using a Cronbach’s alpha coefficient. The construct validity of the scales was also evaluated using Pearson’s correlation coefficient. In Paper 2, the most significant predictors of high work-related affect were determined using binary logistic regression. Odds ratios were estimated for each of the predictor variables (e.g. job demands and job resources). Results The EFA revealed 16 of the 20 items loaded onto three factors: Factor 1 (Work-related Affect; six items), Factor 2 (Job Demands; three items), and Factor 3 (Job Resources; seven items). The Job Resources and Work-related Affect factors had acceptable levels of internal consistency (α = 0.85) and were deemed reliable scales. The internal consistency of the Job Demands scale was below the acceptable level (α = 0.64) and was, therefore, deemed an unreliable scale. The Work-related Affect Scale demonstrates some evidence of convergent validity as it is highly correlated with the Flourishing Scale (r = 0.51). All three scales showed some evidence of discriminant validity as they demonstrated low correlations with unrelated scales such as the Brief Resilience, Flourishing, and the Depression, Anxiety, and Stress 21 (DASS21) Scales. After adjusting for all of the variables in the binary logistic regression model, the most significant predictors of high Work-related Affect were six of the seven Job Resources items (development opportunities (OR 1.82, 95% CI 1.67, 1.99), job control (OR 1.77, CI 1.63, 1.93), appreciation (OR 1.73, CI 1.57, 1.89), workplace relationships (OR 1.70, CI 1.53, 1.89), resources (OR 1.51, CI 1.39,1.65), belonging (OR 1.43, CI 1.32, 1.54), respectively) followed by the three Job Demands items (work extended hours (OR 0.84, CI 0.80, 0.89), letting down friends/family (OR 0.84, CI 0.80, 0.90), work-life balance (OR 1.17, CI 1.10, 1.25), respectively). The seventh Job Resources item (supportive supervisor) was the weakest predictor of high work-related affect, although still statistically significant (OR 0.91, CI 0.83, 0.98). Three items were negative predictors of high work-related affect (work extended hours, letting down friends/family, and supportive supervisor). Conclusion Paper 1 provides practitioners and academics with two valid and reliable scales of work-related wellbeing (Work-related Affect and Job Resources Scales). However, further refinement and testing of the Job Demands factor is needed. The items within these scales fit previous models of wellbeing (e.g. Job Demands-Resources model). The findings from Paper 2 provide practical knowledge that may assist development of workplace wellbeing programmes and inform effective policies that target the wellbeing needs of their employees. The results indicate that programmes or policies should focus on employee resources to have the most significant impact on employee wellbeing. Specifically, resources associated with development opportunities, job control, and appreciation.