Abstract
Short abstract
Becoming a parent is both a rewarding and taxing experience for most parents. Caring for an infant affects stress levels. The Parental Stress Scale (PSS) is a short measure of perceived stress pertaining to the parenting role. The current study examined the psychometric properties of the Danish language version of the PSS in a community sample of 1110 mothers of children aged 0 to 12 months employing the Rasch family of IRT models, and emphasizing the issues of unidimensionality and equal item functioning (no DIF) relative to the age and educational levels of the mothers. No adequate fit to the pure Rasch model could be established for the full PSS scale with dichotomized items. After eliminating two items (2 and 11), two subscales measuring different aspect of parental stress; a 9-item scale measuring parental stress and a 7-item scale measuring lack of parental satisfaction (reversed items) were found each to fit so-called graphical loglinear Rasch models: The parental stress subscale fit a model adjusted for local response dependence between some item pairs, as well as DIF for one item relative to mothers’ level of education and DIF for another item relative to age and educational level of the mothers. The parental satisfaction subscale fit a model adjusted only for local response dependence. The findings are in line with the original interpretation of the PSS. We recommend that the scoring of the PSS is changed to reflect the two subscales and the dichotomization of response categories in future applications.
Becoming a parent is both a rewarding and taxing experience for most parents. Caring for an infant affects stress levels. The Parental Stress Scale (PSS) is a short measure of perceived stress pertaining to the parenting role. The current study examined the psychometric properties of the Danish language version of the PSS in a community sample of 1110 mothers of children aged 0 to 12 months employing the Rasch family of IRT models, and emphasizing the issues of unidimensionality and equal item functioning (no DIF) relative to the age and educational levels of the mothers. No adequate fit to the pure Rasch model could be established for the full PSS scale with dichotomized items. After eliminating two items (2 and 11), two subscales measuring different aspect of parental stress; a 9-item scale measuring parental stress and a 7-item scale measuring lack of parental satisfaction (reversed items) were found each to fit so-called graphical loglinear Rasch models: The parental stress subscale fit a model adjusted for local response dependence between some item pairs, as well as DIF for one item relative to mothers’ level of education and DIF for another item relative to age and educational level of the mothers. The parental satisfaction subscale fit a model adjusted only for local response dependence. The findings are in line with the original interpretation of the PSS. We recommend that the scoring of the PSS is changed to reflect the two subscales and the dichotomization of response categories in future applications.
Original language | English |
---|---|
Publication date | 15 Jan 2018 |
Number of pages | 1 |
Publication status | Published - 15 Jan 2018 |
Event | Seventh International Conference on Probabilistic Models for Measurement Developments with Rasch Models - University of Western Australia, Perth, Australia Duration: 17 Jan 2018 → 19 Jan 2018 |
Conference
Conference | Seventh International Conference on Probabilistic Models for Measurement Developments with Rasch Models |
---|---|
Location | University of Western Australia |
Country/Territory | Australia |
City | Perth |
Period | 17/01/2018 → 19/01/2018 |