文章基本信息

标题：On the Structure of Affect
本地全文：下载
作者：Nadine Wedderhoff ; Timo Gnambs ; Oliver Wedderhoff 等
期刊名称：Zeitschrift für Psychologie
印刷版ISSN：2190-8370
电子版ISSN：2151-2604
出版年度：2021
卷号：229
期号：1
页码：24-37
DOI：10.1027/2151-2604/a000434
出版社：Hogrefe Publishing
摘要：The Positive and Negative Affect Schedule (PANAS; Watson et al., 1988) is a popular self-report questionnaire that is administered all over the world. Though originally developed to measure two independent factors, different models have been proposed in the literature. Comparisons among alternative models as well as analyses concerning their robustness in cross-national research have left an inconclusive picture. Therefore, the present study evaluates the dimensionality of the PANAS and differences between English and translated versions of the PANAS using a meta-analytic structural equation modeling approach. Correlation matrices from 57 independent samples ( N = 54,043) were pooled across subsamples. For both English and non-English samples, a correlated two-factor model including correlated uniquenesses provided the best fit. However, measurement invariance analyses indicated differences in factor loadings between subsamples. Thus, cross-national application of the PANAS might only be justified if measurement equivalence was explicitly tested for the countries at hand.
其他摘要：The Positive and Negative Affect Schedule (PANAS; Watson et al., 1988) is a popular self-report questionnaire that is administered all over the world. Though originally developed to measure two independent factors, different models have been proposed in the literature. Comparisons among alternative models as well as analyses concerning their robustness in cross-national research have left an inconclusive picture. Therefore, the present study evaluates the dimensionality of the PANAS and differences between English and translated versions of the PANAS using a meta-analytic structural equation modeling approach. Correlation matrices from 57 independent samples ( N = 54,043) were pooled across subsamples. For both English and non-English samples, a correlated two-factor model including correlated uniquenesses provided the best fit. However, measurement invariance analyses indicated differences in factor loadings between subsamples. Thus, cross-national application of the PANAS might only be justified if measurement equivalence was explicitly tested for the countries at hand.
关键词：PANAS;Positive and Negative Affect Schedule;factor analysis;meta-analytic structural equation modeling;measurement invariance