ABSTRACT
AIM: To assess the psychometric properties of a comprehensive tool for assessing psychosocial work characteristics(the QPSNordic), focusing on measurement invariance (MI) across occupations. METHODS: QPSNordic consists of 26 scales covering task, individual, and organizational aspects of work. Multiple group confirmatory factor analyses were performed in order to elucidate MI across a variety of occupational groups. SUBJECTS: Private sector employees and health care workers from the public sector participated. Equal size (n = 500) groups from six different occupations were drawn and entered in multiple group confirmatory factor analyses for assessment of MI. RESULTS: Nine of the 24 analysed scales functioned well across all occupational groups and could thus be used for valid comparisons across professions. The majority of the remaining scales showed MI at least across some groups. CONCLUSION: With some exceptions, the QPSNordic has good psychometric properties. Scales measuring organizational aspects of work could also be used across professional groups, enabling valid comparisons between differing workplaces and occupations.
Subject(s)
Interpersonal Relations , Surveys and Questionnaires/standards , Workplace/psychology , Adult , Female , Humans , Job Satisfaction , Male , Middle Aged , Psychometrics , SwedenABSTRACT
Psychometric properties, particularly predictive validity, of scales in the General Nordic Questionnaire for Psychological and Social Factors at Work (QPS(Nordic)) were assessed. The analysis is confined to the scales in the QPS(Nordic,) and 24 of the 26 scales are included. A large group of Swedish county council employees (n= 3,976; response rate = 65%) participated in a study and were given the QPS(Nordic). Register data for long-term sick leave (>90 days), with diagnosis, were used for predictive analysis. The following main results were obtained: Reliability was generally satisfactory, confirmatory factor analysis indicated good fit, concurrent validity was good, some less often investigated organizational variables predicted sickness absence, and scales were differentially associated with absence due to psychiatric and musculoskeletal disorders. In conclusion, the psychometric testing of the QPS(Nordic) so far suggests that it is a good instrument for assessing health-related factors at work.
Subject(s)
Burnout, Professional/diagnosis , Job Satisfaction , Occupational Diseases/diagnosis , Sick Leave , Social Adjustment , Somatoform Disorders/diagnosis , Surveys and Questionnaires , Adult , Burnout, Professional/prevention & control , Burnout, Professional/psychology , Disability Evaluation , Factor Analysis, Statistical , Female , Humans , Leadership , Male , Middle Aged , Norway , Occupational Diseases/prevention & control , Occupational Diseases/psychology , Organizational Culture , Psychometrics/statistics & numerical data , Reproducibility of Results , Risk Factors , Self-Help Groups , Somatoform Disorders/prevention & control , Somatoform Disorders/psychology , Workload/psychology , Young AdultABSTRACT
OBJECTIVES: The assessment of job demands, a pivotal concept in the study of work stress, is often problematic. Part of the problem is that people differ in their perception of whether a given task is overly demanding, or merely a welcome challenge. This study address this problem by examining alternative response formats for job demands that include the respondents' appreciation of this aspect. Another job stress dimension known to be important for health, leadership, was used for comparison and the association with health measures were examined. METHODS: A professionally homogenous group of 105-white collar workers participated in the study. A questionnaire assessing two job characteristics, demand and leadership, was used with four alternative response formats. The selected items were taken from the General Nordic Questionnaire for Psychological and Social Factors at Work (the QPS(Nordic)). Five indices from the SF-36 were used to reflect physical and mental health. Part correlations, taken from a multiple regression procedure were used to elucidate the unique contribution of response format in the association with health variables. RESULT: For the assessment of Demand, response formats that tapped the employee's experience of the demands performed best. This was in contrast to Leadership scales, where an alternative response format did not contribute any unique information. CONCLUSION: Insofar as interventions are aimed at improving worker's health, it seems reasonable to recommend that the workers should also be asked if the job demands are actually experienced as overly demanding.
