I saw the effects of sleepiness in the meditation group in the results section. The conclusion drawn by the author in the abstract is: 【follow‑up analyses showed that sleepiness uniquely moderated the effect of meditation on the LPP, such that less sleepiness during meditation, but not the control audio, corresponded to smaller Lpps to negative images】. Does this mean that sleepiness should be a moderator? However, in the presentation of the results, it seems that only the interaction of sleepiness, group, and valence was reflected. How should this be understood? I always get confused between interaction effects and moderation.

The result part: 【To parse the three-way Valence X Group X Sleepiness interaction, Time was collapsed across Valence and the data was split by Group. To test whether the interaction between Sleepiness and Valence differed as a function of Group, follow-up correlational analyses involving Fisher r-to-z tests were conducted to compare the relationship between sleepiness and the LPP at each valence across groups. Analyses revealed a significant positive correlation between sleepiness ratings and the LPP on negative (r = 0.25, p = 0.014, 95% CI [0.05, 0.45]) but not neutral trials (r = 0.11, p = 0.290, 95% CI [-0.09, 0.31]) in the meditation group. In contrast, sleepiness ratings were unrelated to the LPP on neither neutral (r = 0.07, p = 0.473, 95% CI [-0.13, 0.27]) nor negative trials (r = -0.10, p = 0.313, 95% CI [-0.30, 0.10]) in the control group. Critically, the Fisher r-to-z tests yielded a significant group difference in only the correlation between sleepiness and the negative LPP (z =|2.45|, p = 0.014), but not the neutral LPP (z =|.24|, p = 0.81). Put more simply, less sleepiness during the guided meditation was uniquely related to the predicted smaller LPP on negative trials; see Fig. 4 for waveforms separated by group and sleepiness.】The article is: https://www.nature.com/articles/s41598-020-71122-7

Many thanks in advance!