Hi, i've discovered a significant interaction in my 3x2 ANOVA in my psychology dissertation (3x2 Between Subjects Design, using SPSS). I was wondering which post-hoc analysis i should use to study this interaction. My tutor has provided us with example instructions, which involve splitting the file and conducting two seperate one-way ANOVAs:

To me, these instructions do not make it clear whether these two one-way ANOVAs constitute a post-hoc interaction analysis. Is this the case? If so, how would i go about interpreting them and writing my results in APA format? If not, what kind of post hoc test would i need to carry out in order to assess the interaction?

To clarify, the DV is Familiarity. The IVs are Context (3 Levels), and Conjunction (2 Levels). The indiv_index referenced by my tutor is just an example DV.

Thank you for any help!!!!

*"e) In this analysis we did not find a significant interaction of Context x Conjunction. However, we would have expected this based on the hypotheses above. If you do uncover a significant interaction for other variables (and this is likely e.g., for familiarity), you will need to undertake an interaction post hoc analyses*

However, let’s imagine the interaction of Context x Conjunction was significant (to assist you on other variables where interactions might reach significance). Go to Data > Split file. Click Compare Groups. Put ‘Context’ in to the Groups based on box. Next, you need to run two, One-way ANOVA’s. For the first one go to Analyse > Compare means > One-way. Put ‘conjunction’ in to the Factor box and ‘Individ_index’ in to the Dependent variable box. Now go again to Data > Split file. Click Compare Groups. Put ‘Conjunction’ in to the Groups based on box. For the second One-way ANOVA go to Analyse > Compare means > One-way. Again, put ‘Indvid_index’ in to the Factor box and ‘context’ in to the Dependent variable box. Now, look at both analyses – this will tell you where the significant differences lie within the interaction.

6) Repeat the above process for the remaining dependent variables you wish to analyse (but note there is no need to conduct the interaction post hoc analyses in 5e if the interaction is not significant in the main analyses)"However, let’s imagine the interaction of Context x Conjunction was significant (to assist you on other variables where interactions might reach significance). Go to Data > Split file. Click Compare Groups. Put ‘Context’ in to the Groups based on box. Next, you need to run two, One-way ANOVA’s. For the first one go to Analyse > Compare means > One-way. Put ‘conjunction’ in to the Factor box and ‘Individ_index’ in to the Dependent variable box. Now go again to Data > Split file. Click Compare Groups. Put ‘Conjunction’ in to the Groups based on box. For the second One-way ANOVA go to Analyse > Compare means > One-way. Again, put ‘Indvid_index’ in to the Factor box and ‘context’ in to the Dependent variable box. Now, look at both analyses – this will tell you where the significant differences lie within the interaction.

6) Repeat the above process for the remaining dependent variables you wish to analyse (but note there is no need to conduct the interaction post hoc analyses in 5e if the interaction is not significant in the main analyses)"

To me, these instructions do not make it clear whether these two one-way ANOVAs constitute a post-hoc interaction analysis. Is this the case? If so, how would i go about interpreting them and writing my results in APA format? If not, what kind of post hoc test would i need to carry out in order to assess the interaction?

To clarify, the DV is Familiarity. The IVs are Context (3 Levels), and Conjunction (2 Levels). The indiv_index referenced by my tutor is just an example DV.

Thank you for any help!!!!

Last edited: