Student Adaptation, Loneliness and Mental Health Profiles during the Second Wave of the Pandemic COVID-19

Round 1

Reviewer 1 Report

Dear Authors,

I read the article with keen interest, as it aligns withing the scope of this journal. The topic is intriguing and is likely to captivate readers. The paper is well-structured, and while most sentences are clear, there are no irregular or awkward sentence constructions. The authors have generally used English correctly in terms of grammar and syntax. However, there are some areas where the authors need to provide more clarity and improve the article to meet the standards of an international journal.

Summary:

There are two dots in line 9.

 

Introduction:

Loneliness is mentioned in most of the introduction. Also, the authors state in lines 31-32: "In children and young people, the association of loneliness with mental health difficulties, such as depression and anxiety, has been established [13–15]." Could the authors provide further clarification on what was discovered through the research mentioned earlier?

Updating the title to include loneliness is essential as it is a vital variable under study.

 

Methods:

Please provide detailed information on representativity. What is the recommended minimum sample size based on the study's power and reference population, according to the literature? Explain the criteria utilized by the authors to select their sample for recruitment.

Can the Authors explain how they ensured impartiality towards choice? Also, can the Authors clarify how they eliminated any potential selection bias?

 

The conclusion of the study is missing.

Author Response

Summary: There are two dots in line 9.

This has been corrected.

Introduction: Loneliness is mentioned in most of the introduction. Also, the authors state in lines 31-32: "In children and young people, the association of loneliness with mental health difficulties, such as depression and anxiety, has been established [13–15]." Could the authors provide further clarification on what was discovered through the research mentioned earlier?

More information has been added: Specifically, depression has been found to predict levels of loneliness [14,15]. For example, one meta-analysis reports that depression at age seven is linked with higher levels of depression at 15 years [15]. Similarly, the association between loneliness and social anxiety symptoms is reciprocal, leading to a vicious cycle [13].

Updating the title to include loneliness is essential as it is a vital variable under study.

Loneliness has been added to the title.

Methods:

Please provide detailed information on representativity. What is the recommended minimum sample size based on the study's power and reference population, according to the literature? Explain the criteria utilized by the authors to select their sample for recruitment.

Can the Authors explain how they ensured impartiality towards choice? Also, can the Authors clarify how they eliminated any potential selection bias?

We appreciate your concerns and recognize the importance of these considerations in our study. The information being requested was not explicitly addressed in our study, and this limitation was discussed in the last paragraph of the Discussion. The study did not focus on representativity or recommended minimum sample size based on power and reference population. As described in the Procedure, the data was collected via online questionnaires that were emailed to a large number of schools randomly drawn from the database of schools registered in the Greek Ministry of Education. Therefore, we believe that fairness in the selection process was ensured. However, further studies with larger and more representative samples would be beneficial to address your concerns. We have included a sentence in the Discussion to address future research that could explore the recommended minimum sample size, the representativity of the sample, and strategies to enhance impartiality and minimize selection bias.

The conclusion of the study is missing. – The conclusion has been added:

The study shows that high school students’ adaptation varied during the second wave of the pandemic. The study identified three profiles of students (Resilient and Satisfied, 38%; Average, 41%; Vulnerable and Distressed, 21%) with varying levels of reported resilience, well-being, loneliness and fear of COVID-19, indicating that more than half of the students struggled to some degree during this period of the pandemic. Resilience was positively associated with life satisfaction and negatively associated with loneliness, ill-being and fear of COVID-19. Targeting loneliness and anxiety related to COVID-19 are promising areas for mental health promotion interventions during the pandemic. Socio-demographic factors (gender, age, family income) are related to student adaptation and show the heterogeneous impact of the pandemic on young people. Older age, female sex and lower family income were related to less favourable outcomes. Taking these factors into consideration might facilitate the identification of young people at risk. The findings indicate that tailored mental health interventions are needed to support adolescents adopting tiered approaches. As schools constitute the ideal setting to reach most young people, providing school-based universal and targeted interventions to promote young people’s resilience and well-being is suggested.

Reviewer 2 Report

Please see the attached PDF file for my comments.

Comments for author File: Comments.pdf

Author Response

Major comments

1. Relating personal experience and life-course events to mental health status during the pandemic is impressive. We have measures of family member/friend has become ill and loss of loved one due to COVID-19. Did we also measure whether the respondent got COVID? If so, please include this personal infectious experience in analyses. Also, can we separate family member has become ill from friend has become ill ? This difference is theoretically significant because the infection of a family member is a more traumatic event than a good friend who got the disease.

We did not measure whether the respondents got COVID.
Also, we cannot separate "family member has become ill" from "friend has become ill".

2. The sample is highly unbalanced regarding biological sexes at birth and school grades. There are significantly more girls and more senior high school students. Does this unbalance reflect the reality of Greek middle schools? If not, how large is the difference? And what weighting strategy do we need to match the sample to the actual population of Greek high school students?

You are right about the differences regarding biological sex at birth and school grades. However, these are only some of the ones: similar unbalances were found regarding family income (higher income overrepresented), school type (private schools overrepresented), etc. Due to the nature of the study and because as the number of weighting variables goes up, the greater the risk that the weighting of one variable will confuse or interact with the weighting of another variable, we decided not to proceed with data weighting.

3. Cluster analysis in Table 2 is the shining point of this study. Rather than naming them as Cluster 1, 2, and 3, could you give more meaningful names to these clusters?

Thanks for this comment and your suggestion. We have named Cluster 1 as "Resilient and Satisfied), Cluster 2 as "Average", and Cluster 3 as "Vulnerable and Distressed".

4. As the shining point of this study, more discussions of these clusters further increase this creativity and better address research aim #2 (see Page 3). Such as how family backgrounds, age, biological sex at birth, school grades, and personal experience during the pandemic predict cluster membership. The multinomial logistic regression model using cluster memberships as the dependent variable helps to achieve this analytical goal. 2

A multinomial logistic regression was performed during the analyses stage. The analysis results were not significant and, therefore, were not reported in the paper. However, we followed your suggestion and created an orderable discrete variable with three levels: a relative or friend getting infected, a relative or friend getting hospitalised, and the loss of a relative or friend. We run a multinomial logistic regression to test whether age, biological sex at birth, school grades, family backgrounds (family status and parents' educational level), and personal experience during the pandemic. The results were again not statistically significant.

5. From Table 3 to Table 6, we see how biological sex at birth, school grade, family income, and losing a beloved one impact each mental health indicator and higher-order factor independently. These tables are very informative. However, we need an overall picture: how do those demographic, family, and personal experience factors jointly affect mental health status? Treating each variable in Table 2 as a dependent variable, conduct an OLS regression using independent variables of demographic, family, and personal experience factors. Organize five OLS models in a new table to fully address research aim #3 (see Page 3). Then, we may move Tables 3 to 6 to the appendix.

This was also our idea. However, in most cases the F ratio computed failed to reach statistical significance (therefore, the regression model is entirely useless in these cases). Thus, we preferred to present the impacts of the sociodemographic variables individually on the study’s variables.

6. The measures of personal experience during the pandemic can be further refined. I suggest two variables. The first variable is a dummy variable showing if the respondent ever got COVID (Yes/No). The second variable is an orderable discrete variable with three levels: Loss of a loved one due to COVID-19, a family member getting infected, and a friend getting infected. These two variables and the possible interaction between these two variables will be independent variables in regression models mentioned in major comment 4 and 5.

Unfortunately, we did not ask whether the respondent ever got COVID. The second variable that you suggested (orderable discrete variable) has been created and used in the new analyses.

 

Minor comments

These minor comments relate to wording and formatting issues.

P1-L30 associated with various health and mental health difficulties- Define mental health difficulties before using it for the first time.

This has now been added to the text: “the latter defined as symptoms that meet diagnostic criteria for mental disorders as well as subclinical levels of symptoms that indicate poor mental health, for example, behavioural or emotional difficulties”

P2- L91 In the face of the numerous challenges, research on resilience installs hope- Define resilience before using it for the first time.

The definition has been added to the manuscript:  Resilience is defined as “the process and outcome of successfully adapting to difficult or challenging life experiences, especially through mental, emotional, and behavioral flexibility and adjustment to external and internal demands”

P3-L129 Selective items are: Please explain: Why we don’t use all 20 items to measure loneliness? What 17 items are omitted?

All 20 items were used to measure loneliness

P3-L136 study and 6 items were finally used - Please explain: Which item is excluded from the measure of the fear of COVID-19? Why must we exclude this item?

One item was excluded following the suggestion of the Institute of Educational Policy, which granted permission for the study. We have added a line there to explain this.

P3-L145-148 Positive attitude towards life, Self-esteem, Joy in life and (Lack of) Depressive mood are included in a higher order positive factor called Satisfaction (=.93) while Somatic Complaints and Problems are included in a higher order negative factor called Illbeing (Cronbach’s a=.87) - Please double-check the main text and tables to ensure that Satisfaction and Ill-being are used consistently. For example, Row #7 of Table 1 says, BSW/Y Mean of problems. Does it mean Illbeing ? If so, we need to make terms of the same meaning consistent across this paper.

Corrections were made accordingly.

P4-L153 Selective items include Please explain: Why we use the entire 15 items to measure resilience? What 12 items are omitted?

All items have been used to measure resilience. It seems that the term “selected” caused confusion. Instead of “selective items”, we wrote “example items”.

P4-L157 gender Do we use biological sex at birth (or say sex in short) or the socially constructed concept of gender in this study? Please be very clear about this crucial difference

We have used biological sex at birth. We have replaced all instances of “gender” with “sex” or “biological sex at birth”.

P4-L157 parental educational level, SES Please show how we measure these two concepts. A frequency table for parental education levels and summary statistics of SES should be given (maybe in an appendix).

We have included details about the measurement of these two variables in the final paragraph of part 2.2 (Instruments).

P4-L172 parent’s organizations Please define this term since international readers may not know what these organizations are.

This has been added to the manuscript:

Parent organizations are groups of parents/guardians of each school's pupils who are represented by a parent council. Parent organizations represent the views of par-ents/guardians, facilitate cooperation with the school, support school operations, and support student education and development.

P4-L190 BSWY-problems Does it mean Ill-being? If so, we need to make terms of the same meaning consistent across this paper.

Corrections were made accordingly.

P6-L254 P7-L266 P9-L274 P9-L284 - BSWY - Mean of problems & somatic complaints in Tables 3, 4, 5, 6 Does it mean Ill-being? If so, we need to make terms of the same meaning consistent across this paper.

Corrections were made accordingly.

P7-L258 Grade had a significant effect on all BSWY Define how we measure grades. And what do S and HS mean?

The Note that explained the meaning of S and HS was under Table 3. It has been moved under the Table 4.

P8-L268 Family income had a significant effect Please show how we measure this concept. A frequency table or summary statistics of income should be given (maybe in an appendix).

Family income was measured by a single question with three choices: low, middle, high. This has been explained in the final paragraph of part 2.2 (Instruments).

Frequencies (absolute) appear in Table 5.

P6-L254 Table 3 formatting This table is very hard to follow. Consider a new table formatting like the Suggested new Table 3 below. Show two higher order factors of Satisfaction and Ill-being first. Each one is followed by lower-order factors with at least one-unit indent. Use * to show significance.

DONE

P10-L284 This table is very hard to follow. Consider a new table formatting like the Suggested new Table 6 below. Show two higher order factors of Satisfaction and Ill-being first. Each one is followed by lower-order factors with at least one-unit indent. Use * to show significance.

Table 6 has been deleted because we realized that its results were on a similar question (Has a relative, close friend, or loved person died recently?) and not on the question “Have you experienced the death of a beloved person from COVID-19?”

 

We are thankful for your comments and hope that we have comprehensively responded to all the issues raised.

Round 2

Reviewer 1 Report

Dear Authors,

thank you for improvements

The English language is clear

Reviewer 2 Report

I am very glad to have your replies and revisions. This draft is significantly better than the previous one. All my comments and questions are carefully responded and fully addressed. I have no further suggestions.