How do public exposure and moral beliefs impact feelings of shame and guilt?

In general there is very good practice here for highlighting original-replication comparisons and for ensuring that the final report follows the registration closely. Here are my comments for improvement.

1. Abstract: "impacted differently by exposure" should be expanded for clarity, e.g. "... by appraisals of potential exposure to disapproving others"
2. Authors should follow APA style when citing modern translators and commentators on ancient philosophers, see 7th ed. style manual, p. 325, example 36.
3. The Introduction should position the public exposure/reputation view of shame vs. guilt more clearly against a dominant rival view, the Tangney interpretation of Lewis' Shame and Guilt in Neurosis that shame is distinct from guilt in being subjectively about the whole person rather than a single transgression. This omission is much missed, for example on p. 9 where the GASP is mischaracterized as being mainly about the public/private dimension, when actually it intentionally conflates that dimension with several others derived from the Tangney viewpoint. This addition will only strengthen the case for importance of this study.
4. An important part of a case to replicate a specific article is that it has not been directly replicated already, and/or that conceptual replications/extensions are few - can the authors speak to this issue? 
5. Power analysis should make clearer that main effects and not the interaction + simple effects (because the original did not find it significant) were the main hypotheses of the study. If testing the interaction is important then the study should be powered with regard to simple effects tests interpreting the interaction.
6. Was an attempt made to contact the original authors for the missing disobeying scenarios?
7. pp. 17-18 - if I understand correctly participants can only proceed after correctly answering or re-answering attention check questions. This is a deviation from the original that will likely increase effect size and should be questioned. It might be more defensible to not force this kind of learning and look at responses only with (better test of idea) or also without (better replication of original) a correct check response.
8. Is the outlier exclusion a deviation from the original?
9. I agree it is important to test the validity of the measures as face-wise several of the shame-related measures seem conceptually and empirically shaky (e.g. I usually find anger at self loads on guilt and not shame). Rather than simply testing correlations it might be better to test omnibus reliability and also discriminant validity (e.g., factor analysis of all items) of all the shame and guilt measures, both direct and indirect, together. This would let us assess, improving on the original, whether each "related" measure is correctly classified or not.
10. Also, there is a discrepancy where the results test individual scale reliability but the analysis plan does not.
11. If transforming data into long format, multilevel (hierarchical) analysis needs to be followed with participant as random factor in order to properly account for nonindependence of observations. However, treating shame/guilt as a within-participants factor while keeping wide format data would be simpler. It is not clear from the writing which analysis will be used.
12. In general there are a few grammar and word errors noted that suggest further proofreading is needed.

signed, Roger Giner-Sorolla

This report describes a replication plan for Study 1 of Smith et al.’s 2002 paper about the effect of public exposure of a wrongdoing on ratings of guilt and shame. The paper has been pretty influential, with over 600 citations on Google Scholar, and has been published in the Journal of Personality and Social Psychology a long time ago. Therefore, other than determine whether the study is worthy of replication (which it definitely is), there is not much to say about the merits of that research, the relevance of the original research question, the rigor of the original study design etc. As long as the authors follow the research protocol employed in the original study, they are not obliged to independently satisfy the evaluation criteria of that work.

The intended replication follows the original protocol pretty closely, including the experimental design, adjusted sample size and analytical approach. The authors had to write the stimuli of one of the conditions, which was not provided in the original article. The new text seems valid: it is close enough in length and in format to the texts in the other conditions. The authors are also adding a couple of attention check items, which is a legitimate addition, and standard in such studies in 2022. One difference that is noteworthy and should be discussed in the replication manuscript is the physical settings in which participants will complete the study. The original study was conducted in the lab with about 30 people in the room per session. The replication will be conducted online, with participants completing the study on their own electronic devices, presumably in private. I recommend also coding the type of device on which participants do the study, and if possible prevent people from completing it on their mobile phones, as research in information systems finds systematic differences in user attention between tasks performed on a PC or a tablet and tasks performed on mobile phones.

The only other comments I have are about the literature review. These comments are not relevant to the current stage of the submission, but the authors might want to take them into account later when they write the full manuscript.

1.     The abstract states that guilt and shame are similar in that they both are associated with negative evaluations of oneself, but this is not entirely accurate. Guilt, unlike shame, is associated with negative evaluations of one's behavior, separately from one’s view of one’s qualities and characteristics (i.e. “I did a bad thing” rather than “I am a bad person”). This externalization of the emotion-eliciting wrongdoing is what distinguishes guilt from shame. See Tangney & Dearing, 2002. 

2.     There is more modern research in psychology about guilt vs. shame in general (e.g., Cohen et al., 2011; Tangney et al., 2007) and particularly about the public-private dimension. Some of these works are cited later in the report (p. 9 Choice of replication), but it should be discussed earlier. 

3.    One wrinkle in the public-private distinction between guilt and shame is that guilt enhances interpersonal motives such as the desire to be loved and accepted by others (Baumeister et al., 1994). These motives might moderate the effect of public exposure of one’s behavior on one’s feelings of guilt and shame. Again, this does not affect your replication research but might be a point worth addressing in your discussion.  

In sum, I think the choice of study to replicate is clever and the replication plan is sound. Good luck with your research!

Uriel Haran

References:

Baumeister, R. F., Stillwell, A. M., & Heatherton, T. F. (1994). Guilt: An interpersonal approach. Psychological Bulletin, 115(2), 243–267.

Tangney, J. P., Stuewig, J., & Mashek, D. J. (2007). Moral emotions and moral behavior. Annual Review of Psychology, 58, 345–372. https://doi.org/10.1146/annurev.psych.56.091103.070145