Document Type : Review Article
Author
Professor, Department of Psychology, Faculty of Education and Psychology, Isfahan University, Isfahan, Iran
Abstract
The current paper aimed to investigate core equations properties that preserve in any transformation of equations. One of these characteristics is the symmetry of variables in correlational studies namely equivalent equations. In this paper, first two variables research and discrimination of causality in relationship research and of causality in relationship hypotheses were explained, then toward the third variable that we refer only to mediation, while it has two equivalent models. The issue of equivalent models in applications of covariance structure analysis also discussed. The author then describes the role of equations in psychological papers and emphasizes its exploratory approach. First, he mentions Savadkouhi et al. (2020) Generalized Estimating Equation approach with an independent correlation matrix treating the two records for each subject. The author showed that researchers couldn’t confirm equations before applying them to one of two research designs, namely longitudinal or experimental design. The first rows of the paper also emphasize the longitudinal design called cross-lagged panel Analysis that is now criticized by researchers but remain significant because of confirmation approaches in longitudinal research. The author discusses the causal Inference through experimental design and demand for using mediation analysis hypothesis construction in Iranian papers. The author believes that this continues the tradition of applying exploratory statistics such as stepwise regression analysis in place of confirmatory statistics such as hierarchical regression analysis. Some aspects of the philosophy of science especially refutability was discussed. In the last section, the author introduces some rules for generating equivalent path models, with suggestions to improve the status of mediation research in Iranian papers.
Keywords