Back to PLS HelpPosted on 08/05/19 19:56:56
Number of posts: 9
Hi there,
Can I please if we are allowed to report a LV which is not statistically significant (p=.06) but at cluster level, all of the regions are significant (BSR > 3). Would we be able to report those regions based on their cluster level significance and if yes, can you recommend any published papers that have reported the results this way?
Many thanks!
Cheers,
MZ
Replies:
Posted on 08/06/19 18:21:36
Number of posts: 394
Hi there,
Can I please if we are allowed to report a LV which is not statistically significant (p=.06) but at cluster level, all of the regions are significant (BSR > 3). Would we be able to report those regions based on their cluster level significance and if yes, can you recommend any published papers that have reported the results this way?
Many thanks!
Cheers,
MZ
Whether it's allowed or not depends on the reviewer. I can't find an example, but we have reported PLS LV's that do not pass the conventional threshold and focus on reliability, I think so long as you are clear that the effect is marginal and not try to over-interpret you should be okay.
Login to reply to this topic.
