How to interpret a reverse pattern between activation and correlation in fMRI research

Hi experts,

In my fMRI experiment, two conditions were compared: a high disgust condition and a low disgust condition. The high disgust condition involved presenting participants with disgusting images, while the low disgust condition presented the same images but with the disgusting elements digitally removed. During fMRI scanning, participants passively viewed stimuli from both conditions. After scanning, participants rated the level of disgust for each set of stimuli on a scale of 0 to 10.

Three results were observed:

  1. The disgust ratings for the high disgust condition were significantly higher than those for the low disgust condition, with ratings close to 10 for the high disgust condition and close to 0 for the low disgust condition.

  2. Beta values in a specific brain region were significantly higher (t-test) for the low disgust condition than for the high disgust condition, consistent with existing references indicating a response to this type of digital image processing.

  3. When examining the relationship (Pearson correlation) between the difference in activation (beta values: high disgust condition - low disgust condition) of this region and the difference in ratings (high disgust condition rating - low disgust condition rating) across all participants, a significant positive correlation was found. Almost all activation differences were negative, while rating differences were positive.

On one hand, from the perspective of activation, this brain region appears to respond more strongly to the low disgust condition. On the other hand, from a correlation standpoint, it exhibits the opposite effect.

How can these results be interpreted?

Thank you!