Aim of this study is to encourage the using of structured means modelling as a measurement error-free method, and culturally invariant approach (i.e., measurement invariance at strong level), for evaluating between-group differences in latent means and controlling for the inclusion of latent covariates. Although this technique is not novel in academic literature, applications to quantitative psychology are still uncommon. To compensate this gap an application of latent means and latent ANCOVA, latent covariates models is proposed to the Schwartz�s taxonomy of basic human values theory. Data were collected in June 2011 on a regional basis and age classes and analysed by means of structural equation modelling for a representative sample of roughly 3,000 Italian food consumers. Empirical applications of the Schwartz�s elliptical taxonomy often leads to the overlapping of its adjacent domains both at whole sample and at group level. Hence, someone may come to bias conclusions that no between-group differences exist in those overlapped domains. This study confirms that there is still room for between-group differences at latent mean level. Standardized effect sizes of structured means differences were estimated for all the ten motivational value domains of the Schwartz�s taxonomy across the main four Italian macro-regions. Self-direction, stimulation, power and achievement factors were successively regressed on the hedonism latent covariate, with using latent ANCOVA models, due to the ambivalent nature of the hedonism domain that the Schwartz�s theory stipulates to share (strongly overlaps) elements of both openness to change (i.e., self-direction and stimulation) and self-enhancement (i.e., power and achievement).
Marco Vassallo, Council for Agricultural Research and Economics (C.R.A.), Italy
Stream: Qualitative/Quantitative Research in any other area of Psychology
This paper is part of the ACP2015 Conference Proceedings (View)
View / Download the full paper in a new tab/window
Comments & FeedbackPlace a comment using your LinkedIn profile
Share this Research