I have a construct with four dimensions. I name the construct as X and the four dimensions as a, b, c and d. and the dependent variable is named as Y. In this case, after my hypothesis testing, only a and b positively related to Y is supported, but not c and d. Under this situation, I was asked to test the influence of X on Y ( a must). Now, can I say that X positively related to Y, when half of the dimensions (a,b) only support the positive relationship with Y? Where can I find the reference, please? Thank you!
Dear Guo Fanghua, I hope this finds you well. If a, b,c and d are dimensions of X, you are not supposed to hypothesize that these dimensions influence Y, but that X influences Y. As an additional analysis, you can explore the previous relationships to provide a more nuanced comprehension of the phenomenon of interest. One analysis I suggest you do is a Necessary condition analysis (NCA) that provides the role of each variable for the occurrence of Y, even though the dimensions do not have an influence statistically significant on Y, because NCA does not have an additive logic as the regressive models do. For further information, see Jan Dul studies and the following video too: Dul. J. (2015) Necessary Condition Analysis (NCA): Logic
and methodology of “necessary but not sufficient”
causality. Organizational Research Methods (in press), https://www.youtube.com/watch?v=62s51xQVkY8. If I can provide any assistance on this, please let me know.
- Badges
- Science topic
- Similar topics
- Psychology