In my research I hypothesise that the relationship between performance goal orientation (IV) and meaningfulness (DV) is completely dependent on reward and recognition expectancy (ModV) such that when reward/recognition expectancy is high the relationship between performance goal orientation and meaningfulness will be strong, but diminish at low levels of reward/recognition expectancy. At the same time, I argue that in a model that includes reward/recognition expectancy as a variable, the effect of performance goal orientation on meaningfulness will be non-significant. In sense, reward/recognition expectancy, is therefore considered a moderator and control variable.
According to this line of reasoning, I hypothesise that the interaction term will be significant, as well as the main effect of reward/recognition expectancy (ModV) on meaningfulness (DV); however I expect that the main effect of performance goal orientation (DV) on meaningfulness (IV) will be non-significant in a model that includes the reward/recognition expectancy (ModV). Does this makes sense?
Hello,
If the interaction is significant (i.e. there is moderation), then it is entirely feasible that there will not be a main effect visible between the IV and the DV (see this page for some good info: http://davidakenny.net/cm/moderation.htm).
If you weren't already planning on using it, I would recommend the PROCESS macro for SPSS (using Model 1), which is available here: http://processmacro.org/index.html
...it is an easy way to run the analysis, and you can select an option that allows you to graph the interaction, which is helpful to visualise what is going on.
Good luck with your research,
Rachel
Hi Francois
There is no need to have a significant main effect for interaction to take place. Interaction indicates that the effect of the IV on the DV is conditional on the level of a third variable (lets call it M). When the IV interacts with M the interest shifts from the main effect to simple effects (i.e. the effects of the IV on the DV at different levels of M). It is entirely possible simple effects could be significant even if the main effect was non-significant.
Deon
- Badges
- Science topic
- Similar topics
- Thesis Research