Person-Oriented Profiles Can Clarify Variable-Oriented Associations: The Example of Communication with Parents and Adolescents’ Mental Health Problems

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors The topic raised in the article is interesting and relevant to improving the relationship between fathers and mothers with their children. Bibliographic references are relevant. It is suggested, however, to delve deeper and clarify other variables that, in addition to communication, may affect the mental health of adolescents in the future. From this perspective, it is suggested to reinforce aspects related to having suffered bullying, among others. From this perspective, it is suggested to consider references from studies in this line such as School factors associated with perpetrators and victims of direct and indirect bullying among adolescents. José Luis Antoñanzas Laborda, Ester Ayllón Negrillo, Nieves Moyano Muñoz, Jacobo Cano de Escoriaza
We appreciate the effort in obtaining a robust sample. The methodological aspects, however, we believe could be reinforced with a certain qualitative perspective that allows a deeper understanding of the nuances between the variables studied. In any case, perhaps it could be incorporated into the limitations or prospective part. We must point out that the conclusions can be strengthened with aspects related to family policies that administrations must deploy to improve these family relationships and strengthen the mental health of adolescents, future citizens. Finally, more robust and concrete implications could be incorporated aimed at professionals working in the educational, psychological and medical fields, with the aim of coordinating actions that allow a more effective approach to family relationships and preventing the impact on the mental health of adolescents. . We value the work done.

Author Response

Reviewer 1.

ISSUE 1: The topic raised in the article is interesting and relevant to improving the relationship between fathers and mothers with their children. Bibliographic references are relevant. It is suggested, however, to delve deeper and clarify other variables that, in addition to communication, may affect the mental health of adolescents in the future. From this perspective, it is suggested to reinforce aspects related to having suffered bullying, among others. From this perspective, it is suggested to consider references from studies in this line such as School factors associated with perpetrators and victims of direct and indirect bullying among adolescents. José Luis Antoñanzas Laborda, Ester Ayllón Negrillo, Nieves Moyano Muñoz, Jacobo Cano de Escoriaza.

ANSWER 1: There are many conditions that affect young people's mental health. There are biological factors; home and school experiences (indeed you mention school experiences); poverty; individual factors such as coping skills and body image; peer relationships (you mention bullying); social media; personality factors; and many others. The list can be long. In this study, we looked at a key marker of parental relationships - communication with mum and dad. Good communication is necessary for parental support, and communication is a key indicator of positive or negative parental relationships. In short, we needed to have a clear focus in this study..

ISSUE 2: The methodological aspects, however, we believe could be reinforced with a certain qualitative perspective that allows a deeper understanding of the nuances between the variables studied.

ANSWER 2: This study is part of a data collection exercise that has been carried out every four years since 1983 in many European and North American countries. It is based on quantitative information. However, before the quantitative measures were included in the questionnaires, in most cases their content was explored using qualitative methods.

 

ISSUE 3: Finally, more robust and concrete implications could be incorporated.

ANSWER 3: Thank you for this comment. We have been reluctant to draw strong conclusions about how the results can be used by practitioners. They can easily be misinterpreted, with the result that the reader wants to know which of the two methods gives the 'right' answer. To our knowledge, this is the first study to investigate whether person-oriented methods can provide more specific information than variable-oriented methods. Further studies are needed to cross-validate our findings in this and other areas before the discussion of practical consequences can be seriously put on the table. We believe that simply mentioning that the practical implications may be different, and giving a concrete example, can be a starting point for future discussions about these implications. That is why we have written:

“The results highlight that the practical implications may differ depending on whether analyses are based on variable or person-oriented methods. This is a conclusion that, to our knowledge, has not been widely discussed. For example, one policy implication of the variable-oriented results is that it is important to involve fathers in parental support activities, as mothers tend to make up the majority of participants. A more rigorous policy implication of the person-oriented analyses is that it should be standard practice to include both fathers and mothers in such support activities, as it is when young people have problems communicating with both parents that they experience the greatest mental health problems. Future studies are needed to cross-validate the reported findings in mental health and other domains. They will provide the basis for more general discussions about practical applications. To us, an advantage of person-oriented methods is that results can be interpreted more directly at the individual level than in variable-oriented analyses (Bergman & Trost, 2006). Person-oriented approaches tell us something about people. Are people with a particular response pattern different from people with a different configuration?”

 

Reviewer 2 Report

Comments and Suggestions for Authors

The authors are congratulated for a brilliant approach to data analysis using both the person-centered and variable-centered approaches to analyzing the relationship between 15-year old communication  style with parents and mental health of the young people. Important results were revealed through the analysis the data suggesting that poor communication between 15-year olds and their parents can lead to poor mental health outcomes for the young people. There were also gender variation in these outcomes. The limitations of the study have been outlined and the vitality of the the methodological approach as beneficial to longitudinal studies has been suggested.

Author Response

Reviewer 2.

ANSWER: Thanks for overall positive comments.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

An interesting paper, referred to an actual topic as the mental health issues and the connection with other psychosocial topics such as family/social support, in this case the comunication with parents.  I would like to make some comments and suggestions for the
authors, in order to help for the publication:
- As general proposal, I highly recommend writing in third person (except
in Discussion part).
- Abstract needs a review,  the objective of the article  is not included into the abstract. I highly recommend using the structure Background-Methods-Results-Conclusions.
- In my opinion, the Introduction is correct, thank you.

- The Methods part is good in content, the methodology is a strength of the study, but there is a big lack in explanation: how the authors have identified the mental health problems is not explained in the text (because the items expressed in the tool are not enough to determinate mental health problems), and it's fundamental for this paper this explanation.
- The description/identification of possible bias is not included in the paper, and it's also a fundamental part of this research.

-The part of Conclusions doesn't have practical proposals, and in my opinion this is a paper with a clear practical motivation to decrease the psychosocial impact of lockdowns (in Uganda and in other places)

- Lines 596-599 seem to be very strange, please review them.

Author Response

REVIEWER 3.

ISSUE 1: As general proposal, I highly recommend writing in third person (except in Discussion part).

ANSWER 1: We have read the "Instructions for Authors" and have not found that the journal favors writing in the third person.

 

ISSUE 2: Abstract needs a review, the objective of the article is not included into the abstract.

ANSWER 2: We agree. Thank you. Unfortunately, this information was originally removed in order to keep the abstract to 200 words. We have now added the following sentence at the end of the first paragraph of the abstract: “A person-oriented method, cluster analysis, was used here to clarify these variable-oriented findings on the prediction of mental health problems.”

 

ISSUE 3: I highly recommend using the structure Background-Methods-Results-Conclusions.

ANSWER 3: Agree. We have done so. This is the structure that the journal recommends. Thank you.

 

ISSUE 4: How the authors have identified the mental health problems is not explained in the text (because the items expressed in the tool are not enough to determinate mental health problems), and it's fundamental for this paper this explanation.

ANSWER 4: In this study we measure indicators of problematic mental health. We do not cover more serious measures of psychopathology and psychiatric illness. We use a non-clinical symptom checklist rather than a clinical diagnosis. Because of the reviewer's comments on this, we have decided to clarify this in the manuscript. We now write: ” In this study we measure indicators of problematic mental health. We do not include specific measures of psychopathology and psychiatric diagnoses (such as depression, anxiety, suicidal ideation, psychosis, etc.).” Thank you for this comment.

 

ISSUE 5: The description/identification of possible bias is not included in the paper, and it's also a fundamental part of this research.

ANSWER 5: There are many different types of bias. We think we have highlighted three major limitations of the study. Other biases and limitations are of course possible.

 

ISSUE 6: The part of Conclusions doesn't have practical proposals, and in my opinion this is a paper with a clear practical motivation to decrease the psychosocial impact of lockdowns (in Uganda and in other places)

ANSWER 6: We have been reluctant to draw strong conclusions about how the results can be used by practitioners. They can easily be misinterpreted, with the result that the reader wants to know which of the two methods gives the 'right' answer. To our knowledge, this is the first study to investigate whether person-oriented methods can provide more specific information than variable-oriented methods. Further studies are needed to cross-validate our findings in this and other areas before the discussion of practical consequences can be seriously put on the table. We believe that simply mentioning that the practical implications may be different, and giving a concrete example, can be a starting point for future discussions about these implications. That is why we have written:

“The results highlight that the practical implications may differ depending on whether analyses are based on variable or person-oriented methods. This is a conclusion that, to our knowledge, has not been widely discussed. For example, one policy implication of the variable-oriented results is that it is important to involve fathers in parental support activities, as mothers tend to make up the majority of participants. A more rigorous policy implication of the person-oriented analyses is that it should be standard practice to include both fathers and mothers in such support activities, as it is when young people have problems communicating with both parents that they experience the greatest mental health problems. Future studies are needed to cross-validate the reported findings in mental health and other domains. They will provide the basis for more general discussions about practical applications. To us, an advantage of person-oriented methods is that results can be interpreted more directly at the individual level than in variable-oriented analyses (Bergman & Trost, 2006). Person-oriented approaches tell us something about people. Are people with a particular response pattern different from people with a different configuration?

 

ISSUE 7: Lines 596-599 seem to be very strange, please review them.

ANSWER 7: This is actually the text that explains the first table: "Table 1. Prediction of perceived unhealthiness and psychosomatic symptoms from ease of communicating with fathers and mothers among 15-year-olds in the five survey years. Beta weights for ease of communicating with father and ease of communicating with mother.” We do not know how this text came to be included in the conclusions. However, we originally had a section of tables at the end of the manuscript. The journal moved each of the tables to where they were referred to in the text. The description of Table 1 remained.

Also, all the tables originally had the layout orientation Lying down, but the journal changed it to Standing up when they moved the tables in the tables section to where they should be in the text. This has resulted in columns being deleted in some tables.

 

 

 

Reviewer 4 Report

Comments and Suggestions for Authors

This study used large scale survey data collected across multiple years to examine the relation between adolescents’ ease of communication with mothers and fathers and their health status. The study has the strengths of having large samples across multiple years and using both the variable oriented and person-oriented approaches. Nevertheless, I have major concerns about this manuscript, that is, the lack of clear hypotheses and indepth discussion about the findings. Detailed questions and suggestions are listed below.

 

1. What did the authors expect to find regarding the associations between adolescents’ ease of communication with mothers and fathers and their health status?

2. Did the authors have hypotheses regarding the groups identified using the person-oriented approach? The authors need to review existing literature examining parent-adolescent relationship using the person-oriented approach.

3. What exact gender differences were reported by the existing studies in both health status and adolescent-parent relationship? What were the authors hypotheses regarding the gender differences examined in the current study?

4. Why did the authors choose to use cluster analyses rather than the more widely used latent profile analyses?

5. Besides summarizing the findings and briefly stating their implications, the authors need to interpret why they think the findings emerge while making connections with existing literature. Specifically, why ease of communication with fathers was more strongly associated with adolescents’ health than ease of communication with mothers? Why the “poor mother communication” group had the smallest size? Why girls had worse health status than boys and had more problems communicating with parents, especially fathers?

6. Gender should be coded as 0 and 1 in regression analyses.

Author Response

REVIEWER 4.

OVERALL ISSUE: I have major concerns about this manuscript, that is, the lack of clear hypotheses and indepth discussion about the findings. Detailed questions and suggestions are listed below.

OVERALL ANSWER: This is not a classic research study aimed at filling the gaps in knowledge about the role of adolescent-parent communication in adolescent mental health and explaining changes over the years of the study. With that aim, we would have designed and reported the study very differently. Rather, it is a study of the methodological choices available to researchers - either ordinary variable-oriented or person-oriented strategies, and what person-oriented analyses can reveal over and above the information provided by variable-oriented strategies. We can see that the reviewer expected a classical research study and asked questions in line with these expectations. However, our aim was to distinguish between two types of research methodology.

 

ISSUE 1. What did the authors expect to find regarding the associations between adolescents’ ease of communication with mothers and fathers and their health status?

ANSWER 1: This is a relevant question. Thank you. In the Introduction we now write: “We did not have specific hypotheses about the role of quality of communication with mothers versus fathers for adolescent mental health, as some studies showed that adolescents' communication and relationship with mothers had stronger effects on adolescents' mental health than communication and relationship with fathers (Levin & Currie, 2010; Pantaleao & Ohannessian, 2019; Van Lissa, et al., 2019), while other studies showed that adolescents' communication and relationships with fathers had stronger effects on adolescents' mental health than communication and relationships with mothers (Flouri & Buchanan, 2003: Kernis, Brown, & Brody, 2000; Papaleontiou-Louca, & Al Omari 2020). Our results from the Swedish HBSC dataset showed that ease of communication with fathers had stronger effects on adolescents' mental health than ease of communication with mothers in all survey years between 2002 and 2018. This consistency across years was the starting point for comparing whether similar findings from common variable-oriented analyses would also appear when person-oriented analyses are applied to the same data.

If the results of variable-based analyses are the same as the results of person-based analyses, there should be no need to complement variable-based analyses with person-based analyses. They will produce approximately the same results. But if the results do not converge, there may be a good argument for complementing linear variable-oriented methods with non-linear person-oriented methods.”

 

ISSUE 2. Did the authors have hypotheses regarding the groups identified using the person-oriented approach? The authors need to review existing literature examining parent-adolescent relationship using the person-oriented approach.

ANSWER 2:  To our knowledge, there are no previous person-oriented studies linking parent-adolescent communication to adolescent mental health. There are person-oriented studies linking parent-adolescent relationships to other outcome variables, for example in the political socialization literature. However, these cluster configurations are unique to these particular studies and do not provide important information for the present study.

 

ISSUE 3. What exact gender differences were reported by the existing studies in both health status and adolescent-parent relationship? What were the authors hypotheses regarding the gender differences examined in the current study?

ANSWER 3: In the Introduction, we mention sex differences in mental health, “In these analyses, we examine gender differences because many previous studies have documented greater mental health problems among adolescent girls than boys (Bor, Dean, Najman, & Hayatbakhsh, 2014; Hagquist, 2009; Hagquist et al., 2019; Högberg et al., 2020; Ottová-Jordan et al., 2015).”.

 

ISSUE 4. Why did the authors choose to use cluster analyses rather than the more widely used latent profile analyses?

ANSWER 4: We agree that LPA (Gaussian Mixture Models) is more flexible than K-means. In K-means, a given point belongs to one and only one cluster, whereas in LPA/GMM a point belongs to each cluster to a different degree. LPA/GMM makes fewer assumptions. This does not mean that LPA/GMM is always to be preferred or most common.

When we started thinking about the analyses, we wanted to use a method that most social scientists would intuitively understand, and where we could easily make comparisons between years and genders. We decided to use hierarchical cluster analysis to identify the number of clusters and then K-means analysis to identify the cluster configurations. Cluster analysis provides simplicity and clarity and the results are interpretable at the individual level. With little variability between survey years, this gave us the advantage of using the same centroid when comparing changes in clusters over age and between genders.”

ISSUE 5. Besides summarizing the findings and briefly stating their implications, the authors need to interpret why they think the findings emerge while making connections with existing literature. Specifically, why ease of communication with fathers was more strongly associated with adolescents’ health than ease of communication with mothers? Why the “poor mother communication” group had thesmallest size? Why girls had worse health status than boys and had more problems communicating with parents, especially fathers?

ANSWER 5. We have included the following paragraph in the Discussion; “Finally, we need to pay attention to the particular relationship between adolescents' mental health and their communication problems with their fathers. It is one thing to show that poor communication with fathers is associated with poor adolescent mental health, but it is another to show that communication with fathers may be more important for adolescent mental health than communication with mothers. One explanation for the study findings may lie in the different roles that mothers and fathers play in their relationships with their adolescents. The mother typically has the role of caregiver in the family. Looking back on their upbringing, samples of Swedish young adults living in a suburb of Stockholm in 1958, 1981 and 2011 were asked "Who did you turn to most for support?" and given three alternatives: only mother, only father or both. The majority of young adults answered only the mother (61, 71 and 58%, respectively) (Trifan, Stattin, & Tilton-Weaver, 2014). Consistent with the finding that the mother is the primary provider of emotional support, both boys and girls in the present study perceived that they had an easier time communicating with their mothers than with their fathers. One explanation for the study findings is that particularly poor communication with the father is associated with perceived low paternal emotional support and low quality interactions, which negatively affect adolescents' mental health (cf. Foster, et al., 2023; Kernis, Brown, & Brody, 2000; Papaleontiou-Louca, & Al Omari 2020; Zhang, Pan, Zhang, & Lu, H., 2021). In the present study, adolescents' perceptions of poor communication with fathers were associated with more mental health problems (perceived general health and psychosomatic symptoms) than poor communication with mothers across survey years and for both sexes in variable-oriented analyses. Good communication with both parents is considered crucial for adolescents' mental health (Grusec & Davidov, 2021). Not surprisingly, person-oriented analyses showed that the cluster group who had problems communicating with both their fathers and their mothers generally reported more overall poor health and psychosomatic symptoms than other cluster groups.”

Why the “poor mother communication” group had the smallest size? Poor communication with mother is considerably less frequent that poor communication with the father. Compare Tables 2, 3, and 6.

 

ISSUE 6. Gender should be coded as 0 and 1 in regression analyses.

ANSWER 6: If girls are coded 0 and boys are coded 1, or if girls are coded 5 and boys are coded 10, will give identical results in a regression analysis. But the 0-1 coding is most often reported, so we follow this rationale. Thank you.

Reviewer 5 Report

Comments and Suggestions for Authors

The authors examined mother and father communication and health, and compared differences when using regression and cluster analysis results. The idea was interesting, the use of multiple methods was a strength and the longitudinal nature of the data was strong. The paper had significant limitations that limited its contribution, which are outlined below.

Introduction:

the authors did a nice job outlining the importance of communication. More updated cites on communication would strengthen the paper. Also more information on how communication might relate to general health (more similar to the measures used here) would strengthen the paper.

The authors described the difference between variable and person oriented methods. Overall this was strong, but there were some areas where more clarity is needed. For example, the sentence “in regression studies these weights are often used to infer how the variables function within individuals”. I think this is rather vague. The coefficients show the relationship between variables on average; but don’t say anything about within person processes.  

Furthermore, the methods used here don’t assess stability in person oriented clusters (as the authors describe) because the people in the clusters could change over time. In order to understand this, the authors would need to use a latent transition type of analysis.

The language in several areas lacks precision. Is the outcome general health or mental health (page 2). The measures list general health, yet the paper in many places refers to mental health (193)

The introduction would be stronger with more discussion of their hypotheses as to which parent would matter more for health.

Methods:

Measures: the study was based on one non-specific item of health (overall perceptions of health on a likert scale). The authors vacillated between calling this health and mental health. Yet the item was so nonspecific it was difficult to understand what it would predict.

All of the tables are not formatted properly and are difficult to read. In Table 2, the subscripts are not clear- in many rows there are not matching subscripts.

All of the models would be stronger if the authors added control variables. Many demographic variables could affect health and none of them were included in the analysis.

More information is needed about their methods that led them to conclude that fathers were more important in their variable oriented approach.

I was surprised that the clusters only had two variables. Usually a benefit of these approaches is to capture the complexity of many variables. Because they only had two variables, I wasn’t quite convinced of the utility of the approach, especially because the interaction term was not significant in regression analyses. It also ran counter to their original article to use the clusters to create variable oriented classes and then put them in the same regression model.

Discussion

The discussion would be strengthened by connecting their substantive findings to the literature in terms of which parent/combination of parents is most important

Comments on the Quality of English Language

Overall the english was strong. As noted above, the paper should be edited for precision in some places.

Author Response

REVIEWER 5.

ISSUE 1: More updated cites on communication would strengthen the paper. Also more information on how communication might relate to general health (more similar to the measures used here) would strengthen the

paper.

ANSWER 1: In the third paragraph we have added: “A recent review of the association between adolescents' ratings of the quality of parent-adolescent communication and various mental health constructs in clinical and community samples found small to moderate negative associations, with moderate associations more common in the domains of general health and depression (Zapf, et al., 2023). A meta-analysis found that among 15 common psychosocial risk factors for adolescent depression, poor parent-child communication produced the strongest association (Tang, Tang, Ren, & Wong, 2020).”

In the Introduction we also write: “We did not have specific hypotheses about the role of quality of communication with mothers versus fathers for adolescent mental health, as some studies showed that adolescents' communication and relationship with mothers had stronger effects on adolescents' mental health than communication and relationship with fathers (Levin & Currie, 2010; Pantaleao & Ohannessian, 2019; Van Lissa, et al., 2019), while other studies showed that adolescents' communication and relationships with fathers had stronger effects on adolescents' mental health than communication and relationships with mothers (Flouri & Buchanan, 2003: Kernis, Brown, & Brody, 2000; Papaleontiou-Louca, & Al Omari 2020). Our results from the Swedish HBSC dataset showed that ease of communication with fathers had stronger effects on adolescents' mental health than ease of communication with mothers in all survey years between 2002 and 2018. This consistency across years was the starting point for comparing whether similar findings from common variable-oriented analyses would also appear when person-oriented analyses are applied to the same data.

If the results of variable-based analyses are the same as the results of person-based analyses, there should be no need to complement variable-based analyses with person-based analyses. They will produce approximately the same results. The H0 hypothesis is that these two ways of handling data should produce roughly the same findings. But if the results do not converge, there may be a good argument for complementing linear variable-oriented methods with non-linear person-oriented methods.”

ISSUE 2: the sentence “in regression studies these weights are often used to infer how the variables function within individuals” is rather vague. The coefficients show the relationship between variables on average; but don’t say anything about within person processes.

ANSWER 2: We do not agree that this is a vague sentence. The literature is filled with many examples where the beta weights in a normal regression analysis in the Discussion section is translated into a discussion of how the findings can be interpreted in terms of differences between individuals (see Bergman L.R., Magnusson D., & El-Khouri B. (2003). Studying Individual Development in an Interindividual Context: A Person-Oriented Approach, Vol. 4. Mahwah, NJ: Lawrence Erlbaum Associates). But if this reviewer considers the sentence vague, others might do the same. Hence, we deleted the sentence. It is not important for our argumentation.

 

ISSUE 3: The methods used here don’t assess stability in person oriented clusters (as the authors describe) because the people in the clusters could change over time. In order to understand this, the authors would need to use a latent transition type of analysis.

ANSWER 3: We have been careful not to talk about stability over time for the same persons, but to use the term “structural stability” all the way from the abstract and throughout the study. For example; “We looked at each of the survey years to examine the structural stabilities over time for the communication clusters, i.e., changes in the response patterns (see Table 3).”

 

ISSUE 4: The language in several areas lacks precision. Is the outcome general health or mental health (page 2). The measures list general health, yet the paper in many places refers to mental health (193)

ANSWER 4: We have gone through the study and we recognized that at some places we were not consistent in our terminology. They have been changed. Thank you.

 

ISSUE 5. The introduction would be stronger with more discussion of their hypotheses as to which parent would matter more for health.

ANSWER 5: We now write “We did not have specific hypotheses about the role of quality of communication with mothers versus fathers for adolescent mental health, as some studies showed that adolescents' communication and relationship with mothers had stronger effects on adolescents' mental health than communication and relationship with fathers (Levin & Currie, 2010; Pantaleao & Ohannessian, 2019; Van Lissa, et al., 2019), while other studies showed that adolescents' communication and relationships with fathers had stronger effects on adolescents' mental health than communication and relationships with mothers (Flouri & Buchanan, 2003: Kernis, Brown, & Brody, 2000; Papaleontiou-Louca, & Al Omari 2020). Our results from the Swedish HBSC dataset showed that ease of communication with fathers had stronger effects on adolescents' mental health than ease of communication with mothers in all survey years between 2002 and 2018. This consistency across years was the starting point for comparing whether similar findings from common variable-oriented analyses would also appear when person-oriented analyses are applied to the same data.

If the results of variable-based analyses are the same as the results of person-based analyses, there should be no need to complement variable-based analyses with person-based analyses. They will produce approximately the same results. The H0 hypothesis is that these two ways of handling data should produce roughly the same findings. But if the results do not converge, there may be a good argument for complementing linear variable-oriented methods with non-linear person-oriented methods.”

 

ISSUE 6: the study was based on one non-specific item of health (overall perceptions of health on a likert scale). The authors vacillated between calling this health and mental health. Yet the item was so nonspecific it was difficult to understand what it would predict.

ANSWER 6: In the Measures section we now write:

“In this study we measure indicators of problematic mental health. We do not include specific measures of psychopathology and psychiatric diagnoses (such as depression, anxiety, suicidal ideation, psychosis, etc.).

The HBSC Symptom Checklist, also referred to as a measure of psychosomatic symptoms, has been used in all HBSC surveys since 1986 (Inchley et al., 2018). The scale is a non-clinical measure of subjective health complaints. It consists of the stem question "In the last 6 months, how often have you had the following...?" followed by eight items: "headache", "stomachache", "backache", "feeling low", "irritability or bad temper", "feeling nervous", "difficulty falling asleep" and "feeling dizzy". The response categories are: (1) rarely or never, (2) approximately every month, (3) approximately every week, (4) more than once a week, and (5) approximately every day. Alpha reliability was .84.

Perceived overall unhealth is a measure that captures adolescents' overall perception of their health status, measured by the single item "Would you say your health is ...?". Participants were asked to rate their overall health by selecting one of the response categories (1) poor, (2) fair, (3) good, and (4) excellent, coded inversely (Inchley et al., 2018). Multiple-item measures are preferred to single-item measures, primarily for reasons of reliability. However, this item is an attempt to understand how adolescents perceive their own health status without asking for specificity. It should be added that the correlation between perceived overall unhealth and psychosomatic problems ranged between .38 and .43 (p <.001) over the years of the survey.”

It should be added that the perceived overall unhealth measure has been used in the surveys of HBSC in at least 34 European and North American countries since 2002. There are many published studies that have included this item. We mention some.

 

ISSUE 7: All of the tables are not formatted properly and are difficult to read. In Table 2, the subscripts are not clear- in many rows there are not matching subscripts.

ANSWER 7: We had all tables in the end of the manuscript, but the journal put the tables where they were supposed to be in running text. They forgot that our tables were in Lying not Standing format.

 

ISSUE 8: All of the models would be stronger if the authors added variables. Many demographic variables could affect health and none of them were included in the analysis.

ANSWER 8: We do not see the strong argument for adding variables as controls. In a sense, they are likely to affect the variable- and the person-oriented analyses similarly. Note that our main objective is the direct comparison between two different methodologies for comparing the ease of talking to one’s mother and father with mental unhealth indicators.

We have never seen a person-oriented study where researchers “control” for particular variables, nor do we understand the statistics behind such a rationale. These “control” variables always enter as one other main cluster variable.

 

ISSUE 9: More information is needed about their methods that led them to conclude that fathers were more important in their variable oriented approach.

ANSWER 9: We do not understand. The regression coefficients show that fathers’ communication was more important than mothers’ communication to understand the adolescents mental health problems.

If the question is how to explain the link between poor father communication and adolescent mental health problems, we have added the following paragraph in the Discussion: “Finally, we need to pay attention to the particular relationship between adolescents' mental health and their communication problems with their fathers. It is one thing to show that poor communication with fathers is associated with poor adolescent mental health, but it is another to show that communication with fathers may be more important for adolescent mental health than communication with mothers. One explanation for the study findings may lie in the different roles that mothers and fathers play in their relationships with their adolescents. The mother typically has the  role of caregiver in the family. Looking back on their upbringing, samples of Swedish young adults living in a suburb of Stockholm in 1958, 1981 and 2011 were asked "Who did you turn to most for support?" and given three alternatives: only mother, only father or both. The majority of young adults answered only the mother (61, 71 and 58%, respectively) (Trifan, Stattin, & Tilton-Weaver, 2014). Consistent with the finding that the mother is the primary provider of emotional support, both boys and girls in the present study perceived that they had an easier time communicating with their mothers than with their fathers. One explanation for the study findings is that particularly poor communication with the father is associated with perceived low paternal emotional support and low quality interactions, which negatively affect adolescents' mental health (cf. Foster, et al., 2023; Kernis, Brown, & Brody, 2000; Papaleontiou-Louca, & Al Omari 2020; Zhang, Pan, Zhang, & Lu, H., 2021). In the present study, adolescents' perceptions of poor communication with fathers were associated with more mental health problems (perceived general health and psychosomatic symptoms) than poor communication with mothers across survey years and for both sexes in variable-oriented analyses. Good communication with both parents is considered crucial for adolescents' mental health (Grusec & Davidov, 2021). Not surprisingly, person-oriented analyses showed that the cluster group who had problems communicating with both their fathers and their mothers generally reported more overall poor health and psychosomatic symptoms than other cluster groups.”

 

ISSUE 10: I was surprised that the clusters only had two variables. Usually a benefit of these approaches is to capture the complexity of many variables. Because they only had two variables, I wasn’t quite convinced of the utility of the approach, especially because the interaction term was not significant in regression analyses. It also ran counter to their original article to use the clusters to create variable oriented classes and then put them in the same regression model.

ANSWER 10: The clusters only had two variables: There is nothing wrong with clustering few variables. There are ample of examples in the adolescent literature that few, rather than many variables, are simultaneously examined. It depends on the research question. Here is one recent example: Stattin, H., Eckstein, K., & Amnå, E. (2022). Why some adolescents are open to their parents’ political communication. Journal of Youth and Adolescence. doi: 10.1007/s10964-022-01653-x

The interaction term was not significant in regression analyses: Variable-oriented methods are built on the assumption of linear relationships among the variables. Also, the interaction term is estimated on the same assumption of linear relations. Person-oriented analyses do not have the linearity assumptions (or constraints). Hence cluster combinations may appear independently if there exist linearity or not among the measures. An A*B interaction term that is not significant in regression analysis does not automatically mean that such cluster combination of A and B cannot exist.

It is clear to us that we have to emphasize more in the study that variable-oriented methods only deal with linear relationships whereas person-oriented methods allow for both linear and non-linear relationships – and that it is the non-linear relationships that primarily differentiate the two types of methods. In this revised version we have spelled this out throughout the study.

It also ran counter to their original article to use the clusters to create variable oriented classes and then put them in the same regression model. This was a convenient way to test (a) how much variance the three groups explained, and (b) how much variance the cluster groups accounted for compared to the original two communication variables. We were careful to point out that  “It should be noted, however, that these predictions were for linear rather than non-linear relationships.”

 

ISSUE 11:  Discussion section: The discussion would be strengthened by connecting their substantive findings to the literature in terms of which parent/combination of parents is most important.

ANSWER 11: As reported in Table 4, adolescents who had problems communicating with both fathers and mothers suffered from highest levels of perceived overall unhealth and psychosomatic symptoms. For clarification for the readers, we have entered the sentence “Thus, highest level of perceived overall unhealth and psychosomatic symptoms characterized the cluster group with poor communication with both their fathers and mothers.”

We now highlight more the cluster group with poor communication with both their fathers and mothers in the Discussion.

 

 

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Lot of thanks for your explanations and efforts. I have only missed a comment about the bias on mental health perspective.

Author Response

We believe that this issue about bias is not a major one.

Reviewer 4 Report

Comments and Suggestions for Authors

The authors have addressed my concerns.

Author Response

Fine