SciPost Submission Page
Facilitating better sharing quality of COVID-related headlines
by Irene Sophia Plank
This is not the latest submitted version.
Submission summary
| Authors (as registered SciPost users): | Irene Sophia Plank |
| Submission information | |
|---|---|
| Preprint Link: | scipost_202508_00069v2 (pdf) |
| Date submitted: | Oct. 14, 2025, 7:20 p.m. |
| Submitted by: | Irene Sophia Plank |
| Submitted to: | Journal of Robustness Reports |
| Ontological classification | |
|---|---|
| Academic field: | Multidisciplinary |
| Specialties: |
|
Abstract
Including accuracy prompts and digital literacy tips similarly decrease the likelihood to share COVID-related headlines, especially if they are false.
Author comments upon resubmission
Dear editor, dear reviewers,
Thank you for the positive and detailed feedback! I detail the changes with respect to the editorial recommendation below in the list of changes. I will also upload responses to each reviewer's individual comments.
Thank you for the positive and detailed feedback! I detail the changes with respect to the editorial recommendation below in the list of changes. I will also upload responses to each reviewer's individual comments.
List of changes
1. I elaborated on the differences and the reason for the re-analysis in the Goal section.
2. I fit a cumulative model with a probit link function to the full, unaggregated data. The model includes random group-level intercepts for persons (slopes for Truth), item (slopes for Condition) and country (slopes for Condition, Truth and their interaction), as suggested by the reviewer. Additionally, I fit an alternative model with Country as a population-level predictor with comparable results.
3. I extended the model to include both interventions of sharing likelihood, i.e., the prompt and the tips condition, explaining the exclusion of the accuracy condition in the manuscript. While I still use sum contrasts, I now also compare the two intervention conditions.
2. I fit a cumulative model with a probit link function to the full, unaggregated data. The model includes random group-level intercepts for persons (slopes for Truth), item (slopes for Condition) and country (slopes for Condition, Truth and their interaction), as suggested by the reviewer. Additionally, I fit an alternative model with Country as a population-level predictor with comparable results.
3. I extended the model to include both interventions of sharing likelihood, i.e., the prompt and the tips condition, explaining the exclusion of the accuracy condition in the manuscript. While I still use sum contrasts, I now also compare the two intervention conditions.
Current status:
Has been resubmitted
Reports on this Submission
Report
The author has substantially revised the Robustness Report, and addressed multiple reviewer concerns. Both the reasoning behind the analysis and the analysis itself are now clearer and more appropriate. I have no further comments, and congratulate the author on a useful reanalysis.
Only a small remark: In line 21, "with a" is duplicated.
For transparency: I took a look at the code, but did not re-run it due to time constraints.
Only a small remark: In line 21, "with a" is duplicated.
For transparency: I took a look at the code, but did not re-run it due to time constraints.
Recommendation
Publish (easily meets expectations and criteria for this Journal; among top 50%)
Report
The author has done a very good job in addressing all of my previous comments, as well as those raised by the other reviewers.
I have no further major but have listed below a few very minor points and clarifications that the author is free to address or disregard as they see fit before final publication.
Minor Comments:
• Line 20: "whose variance was 0" — It might be helpful to clarify explicitly what variance refers to here. I assume this refers to the variance in sharing likelihood ratings over all trials for a specific participant?
• Line 69: There appears to be a typo with a repeated word: "...I modelled SL with a with a cumulative Bayesian..."
• Results Section: The author reports a "posterior probability" (e.g., "posterior probability = 100%") and I guess it refers to the the Posterior Probability that the Estimate > 0, but I am not sure. Maybe the authors could clarify what estimate and posterior probability refer to.
I have no further major but have listed below a few very minor points and clarifications that the author is free to address or disregard as they see fit before final publication.
Minor Comments:
• Line 20: "whose variance was 0" — It might be helpful to clarify explicitly what variance refers to here. I assume this refers to the variance in sharing likelihood ratings over all trials for a specific participant?
• Line 69: There appears to be a typo with a repeated word: "...I modelled SL with a with a cumulative Bayesian..."
• Results Section: The author reports a "posterior probability" (e.g., "posterior probability = 100%") and I guess it refers to the the Posterior Probability that the Estimate > 0, but I am not sure. Maybe the authors could clarify what estimate and posterior probability refer to.
Recommendation
Publish (easily meets expectations and criteria for this Journal; among top 50%)