Abstract

The aim of the paper is to present the Polish adaptation of Survey Work-Home Interaction - Nijmegen (SWING). Materials and Methods: The analyses were based on the survey results from two groups of subjects, a sample of workers, representative in terms of sex and age, living in urban areas (N = 600) and a group of 59 employees examined twice with a help of SWING to assess the stability of the obtained results over a month time. Results: The analyses performed proved that the Polish version of SWING is a reliable tool for studying work-home interactions. Correlation coefficients of items with total result of negative work-home interaction (WHI) subscale varied from 0.51 to 0.74, with positive WHI subscale from 0.26 to 0.60, negative home-work interaction (HWI) subscale, from 0.54 to 0.68 and positive HWI subscale from 0.31 to 0.59. Cronbach's alpha for the whole survey was 0.79, and for subscales varied from 0.73 to 0.89. The results of factorial analysis confirmed a our-factor structure of SWING. Factors I, items had loading from 0.58 to 0.81; II, from 0.29 to 0.78; III, from 0.60 to 0.80; and IV, from 0.28 to 0.74. The values of fit index for a four-factor model, were 0.91 (NNFI), 0.06 (RMSEA), and 0.92 (CFI), which means that this model is characterized by a good fit to empirical data. The correlation coefficient between two measurements at one month interval were also high and reached the range of 0.63 to 0.84 Conclusions: The results obtained are comparable to the psychometric characteristic of the English version of SWING. Med Pr 2012;63(3):355-369

Key words