Talk:DataONE:Notebook/Data Citation and Sharing Policy/2010/07/29


 * Heather A Piwowar 13:28, 29 July 2010 (EDT): Hi Nic.  A few things:
 * the fact that the individual categories didn't have significant p-values means that you have to treat the p-value of the interaction with caution.
 * the fact that you are doing what we would call "exploratory" analysis at this point means that you are likely to get significant p-values by chance sometimes, just because you are generating a lot of p-values. That means that for something to be significant in exploratory analysis we often use a p value threshold that is even harder to reach than p<0.05, like p<0.01.
 * one way you can look at the p-values of the interactions is using the "lrm" command in the Design package.  We can experiment with this.
 * dealing with interactions in regressions is pretty advanced. I'd be happy to help you through it (and it may be necessary to analyze your data well) but we have to be careful about where we put our remaining time.  Let's talk....