Write a comment

Thank you for asking me to review "Adverse weather amplifies social media activity" [1]. I have now read it with interest and provide my evaluation below. While the manuscript targets an interesting question and has methodological merit, I cannot recommend it for publication at Psychological Science in its present form.

The authors examine large corpora of Twitter and Facebook posts and relate their daily counts across US cities to concurrent weather events between 2009 and 2016. They find that days when precipitation or temperature levels deviated from local averages were those with increased posts on both platforms. Moreover, authors argue that their regression equation identifies these associations as causal to claim that extreme temperatures and precipitation levels cause more social media posts.

I thank the authors for describing this observation. In my view, descriptions of large-scale behavioral patterns could strengthen the empirical base on which our understanding of digital activities rests. I also appreciated the comparisons between their contrast of interest to those related to e.g. NYC New Year's eve.

However, I don't think the manuscript in its present form satisfies the three journal review criteria [2]. First, on my reading it is not of sufficient "general interest to psychological scientists" but rather to a more specialist crowd investigating either social media, or the behavioral consequences of weather patterns.

More importantly, the "Theoretical and empirical significance" of this work in its present form is not substantial: I am not sure what kind of a theory is supported or undermined by the argument. I think readers would prefer a presentation in which authors clearly lay out the significance and implications of their work to how the mind works and how humans behave rather than leaving that thoughtwork to the readers. Even ideas about behavioral displacement--an obvious point of discussion regarding the key variables in this study--are prominently absent from this manuscript.

Third, I think the authors could improve the transparency of their reporting by directly linking to a repository of their analytic code with synthetic data if necessary. Authors will post their code online at time of publication but that means I cannot review those and therefore cannot comment positively on this aspect of the methodology. Authors also state that the raw data are not available due to the platforms' terms of service. I would like to see the formulation of this in the respective terms of service. Moreover authors state that the intermediate data are available upon request. I don't believe that to be up to date with the journal guidelines [3] nor with current gold standard practices regarding transparency and data sharing.

Therefore in summary while I think the quantitative investigation appears interesting, I cannot recommend the manuscript for publication at Psychological Science at this time. I leave it to the editor's consideration whether they think a revision could address my comments above (in a manner that doesn't turn an interesting empirical investigation into post-hoc HARKing).

All the best and sorry that I did not provide a more positive review at this time. I hope this does not discourage the authors from finding success in their otherwise interesting work.

Respectfully signed,

Matti Vuorre

[1]: https://arxiv.org/abs/2302.08456

[2]: https://www.psychologicalscience.org/publications/psychological_science/ps-submissions#CRIT

[3]: https://www.psychologicalscience.org/publications/psychological_science/ps-submissions#data

This review of "Adverse weather amplifies social media activity" [1] is contributed by Matti Vuorre under CC-BY to Psychological Science and the PREreview platform.

Competing interests

The author declares that they have no competing interests.