More than scientists: How message and messenger attributes influence viewers’ climate change intentions

PLOS One | https://doi.org/10.1371/journal.pone.0331672 September 10, 2025 12 / 20 Our initial Level 2 analysis on the LIWC summary variables produced no significant cross-over interaction effect. The inclusion of words reflecting rational thinking, clout, authenticity, and emotional tone were all ineffective in motivating view- ers, regardless of their political orientation. Our second Level 2 analysis focused on whether political orientation moderated the impact of LIWC drive variables on viewer motivation to act on climate change. There were two significant Level 2 effects for the drive variables, one for affil- iation and one for achievement. As predicted, the effect of including affiliation-related language on viewer motivation was stronger for viewers with left-leaning political orientations than in viewers with right-leaning orientations. The opposite was true for achievement messages. The inclusion of more achievement-related words or phrases was more strongly predic- tive of climate change motivation in right-leaning viewers than in left-leaning viewers. Graphical representations of these interactions are presented in Figs 1 and 2. Our final Level 2 analysis, focusing on whether political orientation moderated the effects of the messenger attribute variables on viewer’s climate change motivation also produced no significant cross over interactions indicating that the Level 1 messenger effects (i.e., older, attractive male messengers in nature were most motivating) did not vary as a func- tion of viewers’ political orientation. Supplementary exploratory analyses

In response to a reviewer suggestion that LIWC message attributes, not included in our original analysis set, may be more predictive of motivation, we conducted supplementary Level 1 and Level 2 analyses investigating (1) the predictive effects of message length (total word count), average words per sentence (a rough indicator of verbosity), six letter words (a proxy for language sophistication, and temporal perspective (past, present or future focused). The Level 1 analysis produced two significant predictors; longer messages and future focused messages were found to be more motivating (p < .01). The analysis explained only 3% of within subject variance in motivation to act, once again a very small amount relative to the variance explained by the messenger variables (over 21%). Level 2 analyses, assessing whether the effects Fig 2. Cross-level interaction indicating that achievement messages are more motivating for right-leaning viewers. Note: Raw data points are shown for all participants, grouped by political orientation: Left-leaning (≤ 25th percentile), Moderate (> 25th and < 75th percentile), and Right-leaning (≥ 75th percentile) based on self-reported political orientation scores. Regression lines are plotted only for the Left-leaning and Right-leaning groups to illustrate modelled trends at the ideological extremes; no line is plotted for Moderates. https://doi.org/10.1371/journal.pone.0331672.g002