Examining the factorial structure, measurement invariance and convergent and discriminant validity of a novel self-report measure of work ability: work ability – personal radar

Abstract

For identification of groups and domains for work ability promotion, brief self-report measure, work ability – personal radar (WA-PR), based on the ‘the house of work ability’ model is presented and psychometrically evaluated in the structural equation framework using data from technological sector (N = 3754). The house model had acceptable fit to the data. In addition, factor loadings in the model were invariant across groups, demonstrating metric invariance of the WA-PR. Scalar invariance of WA-PR was fully demonstrated across men and women, and partially demonstrated across age and employee groups. Comparisons between groups revealed lower levels of health and functional capacity, but higher levels of four other WA-PR dimensions in older employees. In addition, all house structures showed convergence with alternative work ability measures. WA-PR demonstrated potential for effective measurement of multiple work ability dimensions from employees' perspective. It provides means for efficient location of relevant domains and focus groups for work ability promotion.

Abstract

Practitioner Summary: A novel approach to multidimensional work ability measurement was developed to tackle the challenges of work ability promotion. The properties of the instrument were psychometrically evaluated in structural equation modelling framework. Instrument demonstrated potential for locating relevant domains and focus groups for work ability promotions at workplaces and organisations.

Keywords::

• work ability
• measurement
• self-report
• invariance
• validity

Acknowledgements

The authors would like to acknowledge the Good Work – Longer Career project of the Federation of Finnish Technology Industries (2010–2015; http://www.tyohyvinvointi.info/in-english), the Metalworkers' Union, the Federation of Professional and Managerial Staff, the Trade Union Pro and the Professional Engineers in Finland, for providing the data for this article.

Disclosure statement

The authors declare that they have no conflict of interest.

Notes

^1. It is possible to have more resources available than is demanded, but the concepts of overqualification (leftover of personal resources; Brynin 2002) or underemployment (lack of demands; McKee-Ryan and Harvey 2011), are not in the scope of this paper.

^2. It should be noted that the chosen directionality of the paths was arbitrary. Although SEM is powerful statistical method for analysing causality, the authors do not suggest that WA-PR – dimensions are the cause of WAS, RvD and RET. Actually, questions about causality cannot be answered with the current cross-sectional data set. The purpose was solely to study overlap between constructs when controlled for other relevant constructs (unique associations and nomothetic overlap). To examine robustness of the results presented in the text, the models were tested also in reversed causal manner (alternative measures predicting WA-PR constructs). The model fits and interpretations of these models were identical to the models presented in the text. Results for models not presented in the article are available from authors.