Overcoming Pluralistic Ignorance—Brief Exposure to Positive Thoughts and Actions of Others Can Enhance Social Norms Related to Climate Action and Support for Climate Policy

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

It is a very interesting paper. Thank you for allowing me to learn about such an original project addressing vital issues for our planet. The experiment described in the paper was designed correctly and brings very important results. I must admit that the discussion of the results was very interesting and insightful.

However, there is some room for improvement. I have some suggestions concerning the section 1.3 Hypotheses. In my opinion, some parts of the section need clarification, and hypotheses might be articulated differently.

I understand that the experiment was designed as 3 (conditions) x 2 (samples), as you stated in section 2.3. Procedure. So, there are two independent variables (IV): conditions and samples, and eight dependent variables (DV) that you refer to as psychological variables. The first IV, i.e. Exposure conditions, is quite understandable. But the other IV could lead to a sort of confusion. A reader might infer from this section that the other IV is  Exposure to regionally derived CV content, which should have at least two levels: regionally derived content and non-regionally derived content.  Obviously, it is not the case. As far as I understand, the content presented in all three conditions depicts situations specific to Ohio residents, so it would have a different meaning for non-Ohio residents, that is, a national sample. I think such a clear statement is needed about the reason behind having an Ohio sample and a national sample, and how this relates to the variable named Exposure to regional content.

In my view,  the section 1.3. Hypotheses  provides a rationale for hypotheses, which is very good, but it is difficult to follow all your hypotheses. I suggest writing a sort of summary of hypotheses at the end of the section (without rationale), enumerating them one by one, or grouping,  focusing mostly on differences in DVs. Just an example of what I mean:

H1 predicts that there will be a difference in the adherence to norms between respondents exposed to the various types of exposure in such a way that climate-focused CV content will increase norms in comparison with the pro-social CV content and no-CV content.

For the sake of brevity, you may use abbreviations such as CF-CV content for exposure to climate-focused CV content, PS-CV content for exposure to pro-social CV content, NO -CV content for control condition – something like that.

 

My next comments are mostly addressed to technical issues:

• Presented numbers in tables should be shortened to two decimal places, e.g., 0.07 instead of 0.073.
• In Table 3: Is it SE (standard error) or SD (standard deviation)? We present with means usually SD, not SE. Check this in each table.
• You may save space in the tables if you put SD in the brackets like this: Mean (SD) – 3.97 (0.05), instead of 3.971 …..0.046.
• The titles of some Tables should be changed:
  • Table 2. Correlation between demographic variables and dependent variables
  • Table 3. Exposure condition effects on dependent variables. The title should go above the table.
  • Table 5. The title is too long.
• Other changes are much needed:
  • Texts below tables should be as concise as possible and start with the word “Note” written in Italic (APA rules!)
  • In Table 3 – there is a lack of df (degrees of freedom) for F; what does the letter (M) stand for? Superscripts should be written in small letters, not in capital letters.
  • In Table 4 should be presented all main effects of the “sample” variable and all interaction effects.

The last issue: References. It is a good thing to consult software such as Mendeley or Zotero to find the right style to make references, accepted by the Publisher. 

Good luck!

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript describes a study about the effectiveness of an intervention in changing perceived norms related to climate change. I think this work touches on an issue of significant real world importance and has the potential to contribute towards the body of evidence for policy making. Overall, I think the manuscript is generally written clearly and the methodology is mostly sound. I do have a few concerns and suggestions as stated below.  

1. Comparison with other intervention. There is little description of previous work that had tested interventions aimed at combating pluralistic ignorance. A quick search showed that there are indeed several of such studies (e.g., https://journals.sagepub.com/doi/pdf/10.1177/09567976251335585). I think the manuscript would benefit from a brief literature review of several existing interventions and an explanation of how the CV intervention in this study differs from those interventions. On a related note, the manuscript also briefly describes a previous study with partly similarly methodology (Franz et al., 2021) where they examined how the CV intervention shifted perceived norms. I think it would be helpful to articulate how the present study adds to the 2021 study. Together, these could help highlight the originality of the current work. 
2. Positive emotions as mediator. The authors suggested that a possible reason why the pro-social CV content has a similar effect as climate-focused CV content: pro-social CV content leads to more positive emotions (lines 681-683). I think this explanation is plausible and makes sense. But why not put this explanation to the test with the data in this study? Can't positive emotions be included as a mediator in a mediation model? This is especially since this explanation is heavily featured in the concluding paragraph (the reference to Seeger).
3. Significance of indirect effect. In lines 594-596, it was stated that "policy support increases because of descriptive norms, prescriptive norms, and psychological distance change as a result of pro-social CV exposure” (p. 15). But the indirect effect of descriptive norm is (.011, with a 95%CI [-0.001, 0.027]. Since the CI includes 0, doesn’t this means that this particular indirect effect is not significant, albeit ‘marginally’?
4. Direct vs. total effect. Possible mistake(s) in lines 598-600: “We did not find evidence of mediation for the effect of climate-focused CV content on policy support, as the direct relationship (an increase in policy support resulting from exposure to climate change CV content) was not significant”. I am guessing this direct relationship refers to the b = .082, p = .316 effect in Table 5. But shouldn’t one look at the total effect when evaluating if there is “an increase in policy support resulting from exposure to climate change CV content” (total effect is the overall effect, direct effect is the effect of IV that is not accounted for by the mediators). Could the ‘direct effect’ in Table 5 be a typo? Also, even if there is no significant total effect, there may still be mediating effect (see Zhao et al., 2010; DOI: 10.1086/651257; section ‘No need for an effect to be mediated’).
5. Power to detect interaction effect. An interesting finding is that the hypothesis for a stronger effect of region-specific content was not supported. That is, the interaction tests were mostly non-significant. As interaction effects can sometimes require a very high sample size to detect (see: https://approachingblog.wordpress.com/2018/01/24/powering-your-interaction-2/) and that the authors noted that most of the significant effects had small effect sizes, I recommend that the authors consider and discuss how likely is it that the absence of significant interactions is due to low statistical power.

Minor points

7. On page 2, the authors wrote that “the magnitude of Republican’s pluralistic ignorance is greater than that of Democrats and Independents (Sparkman et al., 2022)”. I suggest that the authors revise the text to make clear that this finding pertains specifically to the issue of climate change as shown by the Sparkman paper rather than a general tendency across many/most issues. If they believe in the latter, then they should cite addition papers in support of this.

8. Why are the reliability coefficients for positive and negative emotions not presented when they are presented for other DVs?

9. Inconsistency in the duration of the CV intervention. It was stated as 2.5 minutes on pages 9 & 13 but 3 minutes on page 10 & 17.

10. The presentation of indirect effects in Table 5 could be confusing to readers without experience with PROCESS macro. I understand that PROCESS does not provide p-values for indirect effects and that statistical significance is inferred from the 95% CI (whether or not it includes 0). However, readers who are not familiar with PROCESS may not know this as some other software/packages (e.g., lavaan in R) do provide p-values for indirect effect. So readers may get confused if they see * to denote significance for the a and b paths, but not the indirect effect. They may wrongly infer that the indirect effects are not significant. I suggest revising the table to make clear how significance for indirect effect should be interpreted. Alternatively, you could also use an online Sobel test calculator to derive the p-values for the indirect effects.

11. Page 19, line 732: “..fully explained by increased norms and decreased psychological distance”. I think the word fully is problematic. Many people, such as the creator of the PROCESS macro (Hayes) recommends not using terms such as full (vs. partial) mediation (see: https://www.processmacro.org/faq.html)

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In the manuscript entitled “Overcoming pluralistic ignorance - Brief exposure to positive thoughts and actions of others can enhance social norms related to climate action and support for climate policy”, the authors present a well-designed experimental study examining how exposure to pro-social and climate-focused “Community Voices” content influences normative perceptions, psychological distance, and policy support. The research leverages robust methodological approaches and provides valuable insights into combating pluralistic ignorance in climate communication. This manuscript could be published in “Sustainability” after major revision. The concerns which the authors should consider are as follow:

1. The non-significant improvement in collective efficacy should be more fully discussed, including possible explanations such as the brevity of exposure or the nature of content, and future research design suggestions should be proposed.
2. The sample size (N) for each experimental condition should be clearly reported, especially after excluding participants who failed the attention checks.
3. The finding that pro-social content enhanced policy support and positive emotions more than climate-focused content warrants deeper discussion. This can be systematically interpreted through frameworks such as self-affirmation theory or emotion-buffering mechanisms.
4. Since the observed effect sizes (η²) were generally small, the discussion should further elaborate on their practical significance, particularly in the context of only a single, brief exposure.
5. Although the reward incentive measures adopted in the manuscript help to improve the response rate, it may induce the deviation of the answer, leading to the tendency of participants to cater to the researchers and express their opinions in line with social expectations rather than real thoughts. It is suggested that the author explain this potential deviation or supplement relevant control measures.
6. There are inconsistencies in the format of references, for example, some entries contain DOI numbers, while others are not provided.
7. The author points out that “pro-social content” is more persuasive than “climate content”. However, the explanation of this mechanism is still speculative at present. It is suggested that the author express this conclusion more cautiously in the revision, or supplement relevant data analysis to enhance the rigor of the argument.
8. It is suggested that the layout and content layout of the table should be adjusted uniformly to enhance the alignment effect and overall readability of the data.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have addressed all of my comments. I would recommend it for publication.