Thank you for the great tutorial on conducting simple main effects within the framework of an ANOVA in SPSS! I do have one question: Previously, I would follow up interactions and main effects by either conducing separate t-tests or separate ANOVAs and report the t- or F-statistics resulting from those analyses, so how would you report the F-statistics from the pairwise comparisons done within the ANOVA framework? Would you report the resulting univariate/multivariate F-statistics that now appear in the output after the pairwise comparison result boxes?
Student here trying to summarize/notetake what I've learnt in the video for future use. May have misunderstood. Proceed with caution. So we're testing for 2 main effects: 1. Whether violent videogames increase aggressions levels irrespective of sex. 2. Whether sex affects aggression levels, irrespective of what genre of videogame. Despite a statistically significant difference in the main interaction, we have to find out whether differences in main effects are significant too. This can be done with an ANOVA framework or two separate independent samples t-tests. The ANOVA method is covered in the first part of the video and is slightly more complicated. Some will argue the ANOVA method is how it _should be done_ since it yields more statistical power. Both will yield slightly different p-values due to different methods of calculating the standard error. From textbook: Calculate Cohen's d from simple main effect t-values.
Hey, thanks a lot for the video! But one question, sometimes I don't have an interaction effect between two factors from the table Tests of Between Subjects Effects. But when I run the tests for simple effects as you showed in this video. The results of the estimated marginal means show that the p-value between some factors is smaller than 0.05 nevertheless. Could you tell me why? Thanks in advance!
Hi, thank you for the video. I have a question. I have a two-way mixed MANOVA (1 between subjects - group) and (2 within subjects - repeated measures - pre-post treatment) design. The main effects are significant but the interaction is not. When I conduct a simple main effects for group and for time, then the results are significant? Not sure why this is? Or how to interpret this. Please help.
A significant interaction implies that the magnitude (or direction) of a main effect is different across the levels of another independent variable. In your case, the absence of a significant interaction implies that the means have changed from time 1 to time 2 in a significant way, and to about the same degree for both groups.
@@how2statsbook477 Hi, thank you for this response. So should a simple effects still be conducted if there in no significant interaction? There is a debate on this. Can you help?
Thankyou very much for the wonderful, simple, concise but still detailed video on simple main effect analysis.
Thank you for the great tutorial on conducting simple main effects within the framework of an ANOVA in SPSS! I do have one question: Previously, I would follow up interactions and main effects by either conducing separate t-tests or separate ANOVAs and report the t- or F-statistics resulting from those analyses, so how would you report the F-statistics from the pairwise comparisons done within the ANOVA framework? Would you report the resulting univariate/multivariate F-statistics that now appear in the output after the pairwise comparison result boxes?
thanks for your help!! very useful video
Student here trying to summarize/notetake what I've learnt in the video for future use. May have misunderstood. Proceed with caution.
So we're testing for 2 main effects:
1. Whether violent videogames increase aggressions levels irrespective of sex.
2. Whether sex affects aggression levels, irrespective of what genre of videogame.
Despite a statistically significant difference in the main interaction, we have to find out whether differences in main effects are significant too.
This can be done with an ANOVA framework or two separate independent samples t-tests. The ANOVA method is covered in the first part of the video and is slightly more complicated. Some will argue the ANOVA method is how it _should be done_ since it yields more statistical power. Both will yield slightly different p-values due to different methods of calculating the standard error.
From textbook: Calculate Cohen's d from simple main effect t-values.
Hey, thanks a lot for the video! But one question, sometimes I don't have an interaction effect between two factors from the table Tests of Between Subjects Effects. But when I run the tests for simple effects as you showed in this video. The results of the estimated marginal means show that the p-value between some factors is smaller than 0.05 nevertheless. Could you tell me why? Thanks in advance!
Hi, thank you for the video. I have a question. I have a two-way mixed MANOVA (1 between subjects - group) and (2 within subjects - repeated measures - pre-post treatment) design. The main effects are significant but the interaction is not. When I conduct a simple main effects for group and for time, then the results are significant? Not sure why this is? Or how to interpret this. Please help.
A significant interaction implies that the magnitude (or direction) of a main effect is different across the levels of another independent variable. In your case, the absence of a significant interaction implies that the means have changed from time 1 to time 2 in a significant way, and to about the same degree for both groups.
@@how2statsbook477 Hi, thank you for this response. So should a simple effects still be conducted if there in no significant interaction? There is a debate on this. Can you help?