Apologies for not reading these archives as carefully as I should have done before posting!  I can see from a historical post, that the answer to my first question is (a):  The saliences tell me where voxels show the same (or opposite) correlation depending on the colour of the salience.

I think so. In my case, I'm looking at correlations between accuracy on a 1-back task and brain activity in 3 stimulus conditions (untrained faces, untrained houses, and trained houses). The scatter plots show negative correlations for the two house sets, and zero correlation for the face set. You state: "or to the difference in the correlation if your behavior profile (weight or LV) indicates a task or group difference" So, in my case, I assume that because the face condition doesn't seem to contribute to the LV at all (no correlation, and a weight of near-zero in the behaviour LV plot), and both other conditions are negatively weighted, then yellow would indicate negative correlations and blue positive correlations between brain activity and behaviour. Correct? A second question: As I was scanning these archives, I discovered that my earlier task pls was conducted incorrectly, but I'm at a loss of how to fix it: The task pls that I ran compared these three conditions in session 1 (prior to training the trained house set) versus session 2 (after training). I ran this by treating session 1 and session 2 as separate groups, but I see from your response to others that because these are the same subjects across sessions, I should have entered all 6 conditions (stimulus x session) into the same session file (treating the analysis as repeated-measures). The trouble is: I have different data files for the 2 sessions, and when defining the data files associated with each of the runs, I don't see any way of providing different data files for different conditions (as the gui is set up, the files you select for a given run apply to all conditions). I suppose I could just run the task pls as two separate analyses: one for each session, and look at whether the condition pattern changes across sessions... but what I really want to know is: do I get a significant LV that appears to be weighting session 1 vs session 2 (i.e. some overall effect of learning)?

Not quite. Let me try to clarify. I have two fMRI datasets for each person (session1: pre-learning and session2: post-learning). In each session, there were 6 runs that contained 3 conditions each. I want to compare those three conditions with each other, but also look for session differences (i.e. pre-learning/ post-learning differences). Before, I ran this analysis by setting up two session files for each person (pre and post). In each session file, I defined the six runs and the three conditions in each run and assigned the associated img files that made up each run. Then, I ran the analysis by assigning all the session1 (pre) files to group 1 and all the session2 (post) files to group 2. BUT, I had the same subjects across these two groups, so a repeated measures design would be more appropriate. In an earlier post, you suggested to another poster that to run a repeated measures design, all conditions should be assigned within a single session file, so that the analysis could be conducted with a single group. But, I don't think I can do that in this case. I can append 3 new conditions (for the post data) to my pre session files but, when assigning the img files to each run, I would need to assign different files to the pre and post conditions (because they come from different scanning sessions), and I don't see any way to do that in the gui.

Ok.  That makes sense now.  I'll give that a try.  Thank you both for your help!!