Assessment of psychological and social factors in the workplace : Psychometric aspects

Psychosocial aspects of work have received increasing attention, partly because of their postulated importance for the workers health. Among the techniques developed for measurement of the psychosocial work environment, questionnaires are most commonly used, but their psychometric quality is often questioned and rarely known. The general aim of this thesis was to investigate the psychometric properties of one recently developed such questionnaire, the General Nordic Questionnaire for Psychological and Social Factors at Work (QPSNordic), and to contribute to its validation, particularly in reference to its ability to predict future ill health test its usefulness for comparison over different occupations identify possible shortcomings of the instrument and to suggest and test ways of improvement. The QPSNordic is a large comprehensive instrument covering 13 content areas arranged in a number of scales of the Likert type, accompanied by single items of information not included in a scale. The instrument was given to two large groups of employees, one drawn from the private sector and the other from employees in the public sector, together with scales measuring aspects of physical and mental health. Experimental studies of response formats were performed in subject samples from two insurance companies. The QPSNordic scales were generally shown to be reliable, and the measurement model for 19 of the 24 scales included in our analysis had acceptable model fit. These 19 scales were invariant over different occupational groups. Nine scales, most of them reflecting organizational factors, were invariant over all occupational groups investigated, which means that they could be used in comparisons between different professions. The ability to predict future illness was tested in the public sector employees, whose records were retrieved in the database of their insurance company. Statistically significant correlations were found between several QPSNordic scales and future long term sickleave. There was a differential pattern of correlations, insofar that some scales were predictive of psychiatric, and others of musculo-skeletal disorders. Some weaknesses were also identified, particularly in the operationalization of the job demand concept. An attempt to remediate this by adding response formats tapping satisfaction with the type of demand was successful, and it is suggested that this addition might be useful, both for theoretically oriented studies and when the instrument is used as a basis for workplace interventions aiming to improve worker health