Analysis of the Predictive Role of Self-Compassion on Emotional Distress during COVID-19 Lockdown

Round 1

Reviewer 1 Report

Line 55: "Very diverse psychological... explored.."  write these variables out, rather than simply relay on a list of cited authors.

Line 91: ...having a genuine intention to relieve the suffering..."  Neff posits that SC by it's non-judgemental and mindfulness-based nature will relieve suffering, not that it creates the intention to do so. 

Line: 102  A better wording is : Despite this. there is a lack of ...

Line 307-308:  Summarize this quote to make is smoothly flow into your previous sentences.  It currently stick out poorly.

The discussion needs improvement with a statement as to why this study is of importance, how future research can build upon it and how this study is applicable and generally needed in the sciences and public. 

Overall the authors rely heavily on citing previous literature but do not summarize said points of literature, creating an article that does not read as seriously as it can.  With changes in citations, the above noted concerns, this article could be considered for publication.

 

Author Response

First of all, we would like to express our sincere thanks for all your suggestions, which have been very enriching to learn and improve the quality of our articles.

Line 55: "Very diverse psychological... explored.."  write these variables out, rather than simply relay on a list of cited authors. 

Thank you. We agree that it is a better way to describe them. Done.

Line 91: ...having a genuine intention to relieve the suffering..."  Neff posits that SC by it's non-judgemental and mindfulness-based nature will relieve suffering, not that it creates the intention to do so.

We used the article (Neff, 2003). Neff. (2003). Self-Compassion: An Alternative Conceptualization of a Healthy Attitude Toward Oneself. Self and Identity, 2(2), 85–101. https://doi.org/10.1080/15298860309032

As described by Neff at the end of page 86 and the beginning of page 87, “Compassion involves being touched by the suffering of others, opening one’s awareness to others’ pain and not avoiding or disconnecting from it, so that feelings of kindness toward others and the desire to alleviate their suffering emerge”. As we understand it, one cannot always alleviate suffering. For example, we cannot alleviate the suffering of a family in a war-torn country, even though the desire to do so may arise in us. However, we are ready to elaborate further on this concept in the article, pending your response.

Line: 102  A better wording is : Despite this. there is a lack of ... 

Thank you very much, we agree and have corrected it.

Line 307-308:  Summarize this quote to make is smoothly flow into your previous sentences.  It currently stick out poorly.

Yes, it is better as you suggest and we have changed it.

The discussion needs improvement with a statement as to why this study is of importance, how future research can build upon it and how this study is applicable and generally needed in the sciences and public. 

Yes, we also think the applicability of the results is scarcely described. We have introduced new paragraphs emphasizing both the practical and clinical relevance of the results obtained.

Overall the authors rely heavily on citing previous literature but do not summarize said points of literature, creating an article that does not read as seriously as it can.  With changes in citations, the above noted concerns, this article could be considered for publication.

We have now tried to summarize every key point. Thank you for this new suggestion.

Reviewer 2 Report

This is a review of the manuscript “Analysis of the moderator role of self-compassion on emotional distress during COVID-19 lockdown”. The topic of the manuscript is appropriate for the journal as it presents an online cross-sectional study (n=855) to determine if (or which) of the three facets of self-compassion (mindfulness vs. over-identification; self-kindness vs. self-judgment; common humanity vs. isolation) moderates the relation between perceived infectability and variables of risk of mental health. The manuscript is well-written and the objectives are clearly defined, however there are major statical problems in the link between objectives and the chosen statistical procedures, and in the correct application of these procedures.

The authors extensively, and clearly, show the state of the art on the topic. However, they did not present clearly what are the main contributions of the current study to the field. In fact, the authors just refer (lines 86-89) that the “mediating role of self-compassion on the mental impact of the COVID-19 pandemic” is understudied (though, the authors performed (supposedly) a moderation analysis). The relevance and novelty of the study in the context of existing research should be further detailed. In addition, should specify the hypotheses for their study.

One major problem I found is, indeed, related to the adequacy of the performed analysis to answer to the objectives that the authors identify.

1) The authors claim that they want to study the protective role of self-compassion in the negative effects of covid-19 and mental distress, and for that they decided to conduct a moderation analysis using a hierarchical regression analysis with three steps 1) demographic; 2) perceived infectability; 3) self-compassion. In these steps, the authors never introduced an interaction term between perceived infectability and self-compassion, required to test the moderator effect of this variable on DASS-21. A moderation effect occurs when a third variable changes the strength or direction of the relationship between a predictor and an outcome. In the context of linear regression models, it is tested by creating an interaction term variable (IV x moderator variable) and running the multiple regression including the IV, moderator, and interaction term in the model. This was not done in this manuscript.

2) When using a hierarchichal model as the authors used, a mediation model was performed instead of a moderation. Even though this was not the objective of the author, I should also mention that, in case the authors wanted to perform a mediation, this is not the most correct method to study mediation effects because it does not allow to test if the relationship between the IV and DV has been significantly reduced after inclusion of the mediator variable. This is done by testing the significance of the indirect effect of the model. However, I believe that this was not the intention of the authors given that the relationship between perceived infectability (as predictor) and self-compassion (as mediator) is not expectable.

Regarding this confusion in the statistical method used the authors need to clarify: 1) the objectives and hypothesis to best understand and clarify what type of analysis best respond to their hypotheses and should perform the analysis adequately, either using linear regression SPSS menu or, for instance, the SPSS add-on Process Macro (Hayes, 2017).

Furthermore, and regarding the second part of the results, when analyzing the separate effects of the three facets of Self-Compassion, the authors mention that they conducted a three-step regression, however they seem to have conducted three separate multiple regressions. However, while conducting one regression per facet, they are not having into account the correlation between the three dimensions of self-compassion. The omission of these variables in each specific regression model may introduce a bias in the regression coefficients and, of course, in the conclusions. The correlations between these three facets are not presented, but I am assuming that being different dimensions of the self-compassion there should be no multicollinearity problems in the overall model. I consider that the authors should revise this method.

Of minor revision, Table 1 title indicate that Sociodemographic and clinical characteristics of the sample will be presented in the table. However, we just can see sociodemographic ones (gender, age and level of education). I consider that it makes sense to include in this table the same variables that are to be included as control variables in the model (i.e., the variables included in the first block of the hierarchical regression). Authors may also present the correlations between the variables used in the moderation model in their preliminary analysis of the data.

As the results presented in the article should be majorly revised, currently it does not make sense to deeply review the discussion section.

 

Hayes, A.F. (2017) Introduction to Mediation, Moderation, and Conditional Process Analysis A Regression-Based Approach. Guilford Press, New York.

 

Author Response

• The authors claim that they want to study the protective role of self-compassion in the negative effects of covid-19 and mental distress, and for that they decided to conduct a moderation analysis using a hierarchical regression analysis with three steps 1) demographic; 2) perceived infectability; 3) self-compassion. In these steps, the authors never introduced an interaction term between perceived infectability and self-compassion, required to test the moderator effect of this variable on DASS-21. A moderation effect occurs when a third variable changes the strength or direction of the relationship between a predictor and an outcome. In the context of linear regression models, it is tested by creating an interaction term variable (IV x moderator variable) and running the multiple regression including the IV, moderator, and interaction term in the model. This was not done in this manuscript.

We wish to thank you for these suggestion. Now, that interactional variable was created and included in regression analysis. As a consequence, mindfulness / over-identification the moderating facet lost is hypothesized moderating role. We have changed it result section (in reds) and its implications in discussion section.

2) When using a hierarchichal model as the authors used, a mediation model was performed instead of a moderation. Even though this was not the objective of the author, I should also mention that, in case the authors wanted to perform a mediation, this is not the most correct method to study mediation effects because it does not allow to test if the relationship between the IV and DV has been significantly reduced after inclusion of the mediator variable. This is done by testing the significance of the indirect effect of the model. However, I believe that this was not the intention of the authors given that the relationship between perceived infectability (as predictor) and self-compassion (as mediator) is not expectable.

Regarding this confusion in the statistical method used the authors need to clarify: 1) the objectives and hypothesis to best understand and clarify what type of analysis best respond to their hypotheses and should perform the analysis adequately, either using linear regression SPSS menu or, for instance, the SPSS add-on Process Macro (Hayes, 2017).

We have discussed this commentary and we agree about the potential of our analysis. In that sense, we have change our initial purpose for a more precise purpose. Now we have changed and we have substitute the moderating effect for predictive role. We only speak of moderators in discussion section.

Furthermore, and regarding the second part of the results, when analyzing the separate effects of the three facets of Self-Compassion, the authors mention that they conducted a three-step regression, however they seem to have conducted three separate multiple regressions. However, while conducting one regression per facet, they are not having into account the correlation between the three dimensions of self-compassion. The omission of these variables in each specific regression model may introduce a bias in the regression coefficients and, of course, in the conclusions. The correlations between these three facets are not presented, but I am assuming that being different dimensions of the self-compassion there should be no multicollinearity problems in the overall model. I consider that the authors should revise this method.

Now, we have added correlation coefficient between the three self-compassion facets. And yes, three separate regression analysis were performed. The correlation coefficients are high and statistically significant. The multicollinearity problems can be present, but each facet was used for each regression analysis. For us, the problem can also be because of the three facet are used as tau equivalent, and we are not sure each facet has identical weight in self-compassion total score. We introduce this limitation in discussion section.

Of minor revision, Table 1 title indicate that Sociodemographic and clinical characteristics of the sample will be presented in the table. However, we just can see sociodemographic ones (gender, age and level of education). I consider that it makes sense to include in this table the same variables that are to be included as control variables in the model (i.e., the variables included in the first block of the hierarchical regression). Authors may also present the correlations between the variables used in the moderation model in their preliminary analysis of the data.

Yes, we are sorry. Now table includes the sociodemographic data introduced in regression analysis

As the results presented in the article should be majorly revised, currently it does not make sense to deeply review the discussion section.

Discussion section have been changed according new data.

Round 2

Reviewer 1 Report

This article is well improved and I recommend for publication.

Author Response

Thanks so much for your previous commentaries and suggestions. They gave us an opportunity to a significant improve of the manuscript.

Reviewer 2 Report

In my revision of the article "Analysis of the predictive role of self-compassion on emotional 2 distress during COVID-19 lockdown" in its present form, I consider that almost all of my major concerns were addressed.

However, I still miss a more in depth consideration of the relevance and novelty of the study in the context of existing research. The literature in this domain is extensive, and I think that your paper would benefit if you state clearly what is your contribution for the field. However, I now think that there is more clarity in the link between objectives and the chosen statistical procedures. 

Furthermore, and given that the authors claim in their introduction of the paper that they "consider that it is essential to conduct research analyzing the mediator and moderator roles of SC during the COVID-19 pandemic to be better prepared for future disasters" (lines 102 to 104) I think the results about the no moderation effects should be further discussed in the discussion section.  

 

Author Response

I still miss a more in depth consideration of the relevance and novelty of the study in the context of existing research. The literature in this domain is extensive, and I think that your paper would benefit if you state clearly what is your contribution for the field. However, I now think that there is more clarity in the link between objectives and the chosen statistical procedures.

Furthermore, and given that the authors claim in their introduction of the paper that they "consider that it is essential to conduct research analyzing the mediator and moderator roles of SC during the COVID-19 pandemic to be better prepared for future disasters" (lines 102 to 104) I think the results about the no moderation effects should be further discussed in the discussion section.

Response to Reviewer:

We wish to thanks the commentaries and critical suggestions of Reviewer2. We agree that, with the data re-analysis, the hypothetical moderator effect of SC is questionable. These findings have changed our result valuations, and this was an unexpected data. In that sense, it was difficult for us discussing this new situation. Nevertheless, now we have tried to understand the results (non-moderator effect) and we introduce an explanation in lines 334 to 351.