The Effects of COVID-19 on Family Climate: A Fuzzy Clustering Approach to Examine Spanish Households

Round 1

Reviewer 1 Report

Based on a quantitative study, the article employs a multi-9 criteria decision-making method rooted in fuzzy logic and TOPSIS and a fuzzy clustering method in order to analyze the effect on family climate in Spain by Covid-19. Data was gathered during one month from March 27 to April 27 in 2020. Overall, 2034 citizens responded to the questionnaire, allowing, in principle, to make larger generalizations about the effect of Covid-19 on family climate. The author/s of the article demonstrate competence in recent research concerning the effects of Covid-19 on family climate (on the family in general) though it can be questioned to what extent it is possible to talk about general effects on Covid-19 on family life without contextualizing the comments and conclusion in terms of culturally and socially (also socioeconomically) specific terms. Therefore, it would be useful to mention the area or country on which the research on family that provides secondary literature for the article focuses on. Research method for the article is presented in detail. However, the overall research design of the article is complex and seems to be targeted more narrowly to practitioners of quantitative methods in sociology. The results are presented in systematic and detailed manner though the article falls somewhat short on outlining new findings based on the study – therefore rising the question to what extent the adaptation of the methodology not used for research in family research before can be justified. The article also does not quite fulfil the objective of discussing policy implications as concrete lockdown measures in effect during the time of gathering data have not been outlined.

The article would benefit from a stronger contextualization of the period when data was gathered in terms of the infection rates, lockdown measures applied as well as their duration also indicating whether there was any variation from region to region (e.g., the Canary Islands where most of the respondents were located). Taking into account the very high percentage of respondents from two regions (Andalusia and the Canary Islands) it is questionable whether the conclusions can be extended to Spain as a country. Another option would be to highlight some major differences between the findings from the most represented region, the Canary Islands that makes up half of the data upon which the study is based. It would also be worthwhile addressing the importance of standard of living of the respondents to their perception of family climate, e.g., did the respondents living in more spacious homes perceive the effects less negatively?

The article is based on the assumption that Covid-19 has only negative effect on family climate – at least this is how it is presented in the introductory section. The findings, however, prove that the effect on the family climate was negative only for one specific group – young respondent living with his/her parents. Here, it would be useful to specify what is meant by the category ‘young’ (age range, employment and/or study status, relationship status). The effects of Covid-19 on family climate should be outlined more thoroughly in the concluding sections, including positive effects and also paying attention to gendered differences. The article could also offer some suggestions of attending to the group for which the effect of Covid-19 on family life had perceivably negative effect. Can the lockdown measures be adjusted to better cater for the needs of that group? Also, are there any possibilities to retain the lockdown conditions in areas where it was perceived to have a positive effect on family climate (e.g., retaining the possibility to work partially from home)?

Based on a quantitative study, the article employs a multi-9 criteria decision-making method rooted in fuzzy logic and TOPSIS and a fuzzy clustering method in order to analyze the effect on family climate in Spain by Covid-19. Data was gathered during one month from March 27 to April 27 in 2020. Overall, 2034 citizens responded to the questionnaire, allowing, in principle, to make larger generalizations about the effect of Covid-19 on family climate. The author/s of the article demonstrate competence in recent research concerning the effects of Covid-19 on family climate (on the family in general) though it can be questioned to what extent it is possible to talk about general effects on Covid-19 on family life without contextualizing the comments and conclusion in terms of culturally and socially (also socioeconomically) specific terms. Therefore, it would be useful to mention the area or country on which the research on family that provides secondary literature for the article focuses on. Research method for the article is presented in detail. However, the overall research design of the article is complex and seems to be targeted more narrowly to practitioners of quantitative methods in sociology. The results are presented in systematic and detailed manner though the article falls somewhat short on outlining new findings based on the study – therefore rising the question to what extent the adaptation of the methodology not used for research in family research before can be justified. The article also does not quite fulfil the objective of discussing policy implications as concrete lockdown measures in effect during the time of gathering data have not been outlined. 

The article would benefit from a stronger contextualization of the period when data was gathered in terms of the infection rates, lockdown measures applied as well as their duration also indicating whether there was any variation from region to region (e.g., the Canary Islands where most of the respondents were located). Taking into account the very high percentage of respondents from two regions (Andalusia and the Canary Islands) it is questionable whether the conclusions can be extended to Spain as a country. Another option would be to highlight some major differences between the findings from the most represented region, the Canary Islands that makes up half of the data upon which the study is based. It would also be worthwhile addressing the importance of standard of living of the respondents to their perception of family climate, e.g., did the respondents living in more spacious homes perceive the effects less negatively?

The article is based on the assumption that Covid-19 has only negative effect on family climate – at least this is how it is presented in the introductory section. The findings, however, prove that the effect on the family climate was negative only for one specific group – young respondent living with his/her parents. Here, it would be useful to specify what is meant by the category ‘young’ (age range, employment and/or study status, relationship status). The effects of Covid-19 on family climate should be outlined more thoroughly in the concluding sections, including positive effects and also paying attention to gendered differences. The article could also offer some suggestions of attending to the group for which the effect of Covid-19 on family life had perceivably negative effect. Can the lockdown measures be adjusted to better cater for the needs of that group? Also, are there any possibilities to retain the lockdown conditions in areas where it was perceived to have a positive effect on family climate (e.g., retaining the possibility to work partially from home)?

Author Response

You can see our response to the comments in the attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

The article aims to analyze the family climate during the confinement during the first wave of Covid19 in the Spanish context, based on a quantitative survey and using an original statistical approach, the fuzzy statistics methods, although quite complex.

The article is interesting and could be a potential article for Social Sciences. It is also well written and well documented about the Covid 19 on family cohesion (literature issued from sociology and psychology journals) and on the methodology. But, as a reader, we remain a bit skeptical on three aspects: first, the main orientation of the article is not very clear. Do the authors want to show the potential of the statistical methods used? And if so, why publish in Social Sciences and not in a journal of methodology in social sciences? Or (and we assume this), is it to show and interpret substantial results (3 types of family climate)? Secondly, what does the use of their methods bring to the table compared to more traditional methods (sum of scores on each item and regression analysis, or factor analysis)? The authors should therefore provide answers to these questions at various points in their text (methodology, results, conclusion) (or at least make sure that readers do not ask them or no longer ask them). Third, do the authors really measure the family climate in the interviewees' households? This last question is a bit provocative, but it sometimes seems to us, particularly when reading tables 7 and 8, that the analysis is more about the interviewee's interpretation of the family climate in her household. In other words, a mother could give different answers to the different items of the family climate scale (if she is happy that all family members are present in the household) than her teenage child (if he/she regrets not being able to see his/her friends anymore).

Another point of discussion is about the clusterization. A first question concerns the choice of three clusters. As we are not familiar with fuzzy statistics methods, we naively wondered if the choice of three clusters was related to the constitution of triangular fuzzy numbers. But if it is rather a question of two independent methodological choices, the authors should develop why the choice of three clusters, and not four or five clusters, which could give some complexity to the results. Concerning the fuzzy numbers, the authors could also detail the reasons for the choice of each of the triplets associated with each possible response of the family climate items. Finally, following Table 7, it seems that the choice of the intermediate label to designate the 3rd cluster is rather wrong. As the authors note, maximum values are observed in 11 items of the scale for this cluster, while the minimum value of the "pay attention" item is lower than in the case of the extreme negative cluster. It thus seems that this second cluster borrows a lot from the extreme positive cluster, but with exceptions concerning attention to the other or yell when angry. It seems that the family climate is overall good in this cluster, except at some moments in which climate is considered a little more "explosive" and “stormy” (misunderstanding, disorder, even friction within the family). By giving a concrete interpretation of the three main types of climates, the link with individual characteristics would make it possible to better understand how climates are interpreted according to the social groups to which one belongs (for example, whether one is young or not, etc.). But as said before, the used scale does not seem to be directly a measure of the family climate, but rather corresponds to a measure of the interpretation of the climate that is made by an interviewee, according to his or her age, social background, etc.

I also have some more specific comments, mainly of a methodological nature.

Data: are there any distortions in the interviewed sample in relation to the Spanish population (over-representation of certain social classes, for example).

Tables 4 and 5: are they essential to the subject (the question is a bit provocative)?

Table 6: is it possible to have statistical tests that would show a difference between men and women, or according to other variables?

Figure 1: is unreadable

Table 8: There seems to be an error in the probabilities of belonging to each of the clusters according to the age group to which they belong (the total row does not add up to 100%, and there is in fact a repetition in columns 4 and 5). For the other variables, it would be better to have a total of 100% rather than 99.9%, as is sometimes the case.

Author Response

You can see our response to the comments in the attached file.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Thanks to have answered to questions and to have taken into account my remarks in the text (and so promptly).