Hi all,
When using the Johnson-Neyman method for the PROCESS macro for SPSS to determine the region of significance for moderation (i.e., jn = 1) the default p-value is set to 0.05. I am wanting to change this default to a different p-value (p = 0.0167).
Is anyone aware of syntax which would allow me to do this? I have tried using /intprobes, however, that only impacts when PROCESS will produce output for conditional effects.
Any advice would be much appreciated! Thanks ~ Seth
First you mean the default alpha is 0.05. a p-value is a sample statistic. Second the cutoff you wish is a very strange number. I'm wondering if this is chosen to make your test stat significant. If so this is called p-hacking and is regarded as as a ethics violation I suggest you look all of this up and if this is what you are doing I strongly advise against it. Best, David Booth
Hi David-- Yes, I meant default alpha. I set my cutoff to 0.0167 a priori using a Bonferroni correction to adjust for inflated familywise error rate from running multiple analyses on the same data. This alpha was set prior to running any analyses, after consultation with my committee and as is written in my analytic plan.
The question I posted is in regards to the Johnson-Neyman region of significance produced by the PROCESS macro for SPSS. I would like to adjust the default alpha in this output in order to determine the percentage of values of my moderator falling outside/within the region of significance. I hope this clarifies and appreciate your engagement. ~ Seth
Given what you said I have no idea of what you are doing.. perhaps the attached paper would be worthwhile. Best wishes David Booth
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS