All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Many thanks for the changes made in response to the last round of comments. These changes have again considerably improved the manuscript, which I am now happy to accept for publication. I did identify a number of small remaining grammatical errors, which I have corrected in the attached draft.
Congratulations on this publication!
[# PeerJ Staff Note - this decision was reviewed and approved by Jafri Abdullah, a PeerJ Section Editor covering this Section #]
[# PeerJ Staff Note: Although the Academic and Section Editors are happy to accept your article as being scientifically sound, a final check of the manuscript shows that it would benefit from further English editing. Therefore, please identify necessary edits and address these while in proof stage. #]
This second revision has considerably improved the quality of the manuscript, and I thank you for taking the reviewers' comments on board in your revisions. There remain some outstanding issues that should be addressed to enhance readability and interpretation of the results (especially for readers less familiar with the types of analyses that are reported). I have outlined these - structured by section- below.
*Introduction*
The introduction has much improved with the addition of relevant literature. In addition, the research question is now much clearer.
- There remain several typos however (e.g. places where a space between a word and a bracket is missing). Please correct these, as well as some grammatical errors (e.g. “…higher prevalence … was [not were] found to be associated…”; “Studies have found association” (should be ‘associations’).
- Please clarify “Assessment of the extent that availability of cash or cash equivalent to cover a small emergency moderates (i.e., modifies the association) between experience of economic distress and behavioral health outcomes can help inform stakeholders in behavioral health and counseling on the nature of this important association”. It is unclear which relationship ‘this association’ refers to. This explanation of why the study is important would also be better placed after the following sentence (starting with ‘The objective of…’), as it would avoid having to repeat describing the moderating effect. Currently, there is a lot of overlap between the last two sentences of the introduction.
- To assist the reader in evaluating the reported results relative to the proposed research question, please specify the predictions that the proposed hypothesis generates. More specifically, given the way the data are analysed, what patterns of results would need to be observed to support the hypothesis (i.e. describing how stratified associations should significantly differ from one another).
*Results*
- Please structure your results into separate sections to assist the reader in understanding the data presented. This can consist of using subtitles for the different sections of results: sample characteristics, associations between variables of interest, and model results.
- The results section does not describe what kind of analyses were conducted (in fact, the abstract has more detail on this than the manuscript itself). Only the outcomes of these analyses are described. Please describe in the manuscript for the results reported in Tables 2, 3, and 4 which analyses were conducted and how they were conducted. In addition, please provide goodness-of-fit benchmark values appropriate for your sample size and variables, to assist the reader in interpreting the reported findings.
- The meaning of the following sentence is unclear, it seems a section of the sentence is missing “There were no significant differences in the extent that experience of economic distress and behavioral health outcomes according to Breslow-Day-Test of heterogeneity of stratified odds ratios.”
- There seems to be only one sentence which refer to the actual test of the proposed research question: “There were also no statistically significant indication of additive or multiplicative effect modifications (i.e., moderation).” These analyses should be more explicitly introduced as the primary tests of the proposed question, and the description of these null-results should be expanded.
- The results presented following this key test of the proposed question should be explicitly described as exploratory findings.
*Discussion*
- The discussion also still includes some typos and grammatical errors (e.g. “Future studies should considering” ; “should be interpret”). Please revise as PeerJ does not provide copy-editing support.
- Please break up the discussion into smaller paragraphs to assist with readability.
While the comments of two of the reviewers were addressed to a satisfactory level, there are outstanding issues identified by the third reviewer which require substantial further revision to bring the manuscript to a standard that is acceptable for publication. In particular, the introduction is extremely brief and still has a paucity of literature supporting the formulation of the hypothesis. In addition, the Results section does not present an analytic strategy that maps onto the hypothesis, as no results describing a moderation analysis are presented. Presenting the results in a manner that is more tightly aligned to the hypothesis will also help in the interpretation of the findings. Given the limited analyses currently presented, the interpretation of the data is also limited, however more importantly the proposed hypothesis is not addressed.
These two issues of
1. providing a sufficient literature background/context with adequate references, and
2. analysing, presenting and interpreting the results in a manner that is consistent with the proposed hypothesis,
in addition to the other comments raised by the reviewer, will need to be addressed in the next revision for any further consideration for publication to be made.
See initial review
See initial review
See initial review
See initial review
no comment
no comment
no comment
It would be easier for the reader if Table 3's title is more specific: showing the alternative definition explicitly in the title.
The paper still lacks three journal standards:
Literature references, sufficient field background/context provided.
Professional article structure
Self-contained with relevant results to hypotheses
See attached document for comments
The research objective is clearer but the research question is not well defined. The authors do not make a strong case as to why or how the question is relevant or meaningful. There is not an identified knowledge gap.
There are gaps in the method description.
See comments in attached document.
It is difficult to assess the conclusions due to the way in which the discussion/conclusions are structured as well as the overall structure of the paper.
See attached document for comments
See attached document.
I have received comprehensive reviews of three experts in the field. Each noted several strengths of the manuscript, including the importance of the research question, design of the study, and the amount of detail provided in the manuscript. However, each reviewer also identified areas where the manuscript can be improved. Please address each of the reviewers' comments when submitting a revised version of this manuscript.
This paper reports the results clearly and in sufficient detail. Clearly written throughout. Introduction shows context clearly. Raw data has been supplied.
This study appears well design to answer the research question, though it is cross sectional and a longitudinal design would have been preferable. This is primary research in line with the journal scope with a clear research question. Ethics approval and considerations is outlined and enough detail is given for replication.
Findings are clearly presented, statistics seem appropriate and demographic variables are controlled for. It’s finding about cash reserves being more important than financial shocks is relatively unique an important one for the literature. Data has been provided and the statistical analyses as far as I can tell are sound (I believe they used a difference software package from me so I cannot repeat the analysis but they appear sound from the details given). Conclusions are clear.
• Introduction line 68: ‘habits in saving’ please reword e.g. ‘usual habits in saving’.
• Introduction line 77: Change ‘leaky’ to ‘leaking’.
• Materials line 108: Please explain why the US amount for emergencies of 400 dollars was used and then this change to Thai Baht: Was there not such an estimate available for Thailand specifically to use as a benchmark for case available for emergencies?
• Materials line 128: Please add to paper that the measure of financial stress by Siapush has also been used in previous studies of mental health and finances (e.g. Richardson et al., 2007).
• Materials line 136: Please clarify if the PHQ and GAD measures have been formally validated with a Thai sample and if so provide a reference. Please also provide the chronbachs alpha of the GAD-7 in this sample to demonstrate reliability of this measure. Please also clarify how this measure was used in the analysis: I am unclear if this was used in the results i.e. using this score to determine if they had a financial shock. There are other analyses which could have been conducted e.g. correlating PHQ and GAD depression scores with scores on this financial measure: Please explain why this was not done and clarify how this financial measure was analysed in the data.
• Materials line 177: It says that missing data was excluded ‘except for questions on sensitive issues’, please justify this and explain in more detail: What constituted a sensitive question? Why was this not excluded, and what happened with this missing data?
• Materials and method line 184: Please clarify what statistical software was used for the analysis.
• Results line 220-221: Please remove ‘(n=1,555 participants)’ as this is repeated twice in this sentence.
• Results line 226: Please add the % to ‘the majority of participants’. I am also not sure if you could describe 51.7% as majority of participants unless you conducted a statistical analysis to show if this gender difference was statistically greater than expected.
• Results line 230 ‘slightly higher than those’ add % here as well please.
• Discussion line 265-272: I do not think this much discussion about prevalence of credit card use in Thailand is needed, it is not especially relevant to the results. Please edit down.
• Discussion line 293 onwards. Please add to discussion about limitations that perhaps a more global measure of stress could have been included. I think it is also important to place more emphasis on the cross-sectional nature of the study as a limitation: Reverse causality cannot be ruled out e.g. it could be that those who are prone to depression are less likely to have savings rather than not having savings increase the risk of experiencing depression symptoms.
• Discussion lines 316-323: There is a lot of discussion about the link between finances and binge drinking, including you saying that ‘those who experienced economic distress were less likely to self-report binge drinking’. Please correct this and edit down this section: There was a non-significant trend, so this needs to be clarified in the discussion and I don’t think this must discussion about alcohol us warranted giving a non-significant finding.
• Discussion line 324: You say qualitative methods should be used I think longitudinal analysis would be more helpful to look at the link between covid financial impact and binge drinking.
• Discussion line 330: You say that the characteristics of your sample ‘closely reflect’ the demographics of the Thai population. How can you be sure unless you have formally compared the demographics in the sample? I don’t think such a full formal analysis is required, but linking here to some data about broader Thai population could help with a brief analysis e.g. gender break down and average age here was X compared to general population.
• Discussion line 337: Add to future studies about the importance of longitudinal studies in the area.
no comment
no comment
A summary of research
Employing a nationally representative telephone-based survey in late April 2021 with a sample size of 1,555 participants, this study assessed the extent that mental health outcomes and health behaviours were associated with the availability of emergency cash reserves and experience of economic distress in the general adult population in Thailand. The authors found that participants who experienced economic distress but had emergency cash reserves had a lower prevalence of anxiety, depressive symptoms, and a history of gambling than participants who did not experience economic distress but had no emergency cash reserve.
Comments
Abstract
1. In the Background, it makes more sense to link the importance of emergency cash reserve to economic problems caused by the COVID19 pandemic rather than link it directly to the COVID19 pandemic.
2. It seems the Cis in Table 2 do not correspond to the Conclusion. The CIs of the two mentioned categories are overlapped.
Methods
3. The Sampling Methods should be put close to the Study Population and Sample Size Calculation.
4. The study population should be spelled out clearly whether they are people who have household phone numbers or not. Please make it clear that the phone number list covers the cell phone or household phone. Please provide information regarding the coverage of Thai people who have phone numbers used in the study, and how well the people who have phone numbers represent the general population. The poor people who do not have any phone number may get worse economic and psychological impacts from the pandemic as compared to those who have one. The authors mentioned the sampling list and the replacement sampling model are belong to the research firm; however, this information should be described adequately.
5. Please explain more about the sampling lists of both sampling levels. For example, the first sampling level uses the individual list or household list; the second level uses what phone number list. Please give more detail on what cumulative systematic sampling is for both sampling levels.
6. Please provide a bit of information describing the quality of the field researchers (e.g., university students or the research company’s regular researchers) in order for the reader could assess the quality of the study data.
Results
7. Line 218, please consider changing the term ‘the participant rate’ to ‘the response rate.’
8. Line 243, please change ‘or’ to ‘for.’
9. It would be good to show the ORs for all controlled variables in order to know how much those variables influence the outcomes.
10. The reader might understand all parameters easier by 95%CI than S.E.
11. It would be good to do cross-tabulations between ‘personal monthly income’ and several other variables, including credit card use, emergency cash reserve, the experience of economic distress, and all health outcomes. These tabulations could show how the baseline income is important and in what dimensions. Or, to do the current analyses but stratified by personal income because it could illustrate the effect of personal income on the outcome measures.
Discussion
12. The authors should compare the sample characteristics with the Thai population characteristics to explicitly show how well the sample represents the population. Do not have to do an explicit table.
13. The authors might consider using a different classification for identifying the ‘emergency expense variable’ in the analysis. For example, treating those who use a credit card as ones who have no cash may cause misclassification bias because using a credit card may not mean they have no cash. These people may use their credit cards because it is a better option for them (e.g., save cash to use for other things, no interest, get points, etc.; poor people could not have a credit card).
14. It might be better to demonstrate sensitivity analysis of the effect of a different definition of the cash reserve than to explain that there was a small prevalence of those who used credit cards in this study.
no comment
The authors address an interesting and important topic related to the relationship between an individual's financial situation and their experience of anxiety and depression. In reading the paper, I am not entirely sure of their conceptual argument or the exact research question they are attempting to answer. Lines 25 to 27 present the objective as assessing "the extent to which mental health outcomes and health behaviors were associated with the availability of emergency cash reserves and experience of economic distress." Lines 84 to 87 present the objective as assessing "the proportion of the general adult population of Thailand with cash reserves for a small emergency, and the extent that the availability of such emergency cash reserves was associated with mental health and behavioral health outcomes during the COVID-19 pandemic." Lines 252 to 255 present the objective as describing "the availability of emergency cash reserves . . . and experience of economic distress uring the COVID-19 pandemic, and their associations with mental health and behavioral health outcomes." The analysis examines the interaction of economic distress and emergency cash reserves although the authors do not discuss moderation.
It seems that a conceptual model and explicitly stated research question might be helpful in clarifying the objective of the study. From my reading, it seems that the question is whether the relationship between the experience of economic distress and mental health depends on the individual's level of emergency cash reserves. For example, the authors state that "participants who experienced economic distress but had emergency reserve had lower prevalence of anxiety, depression, . . . . " (Lines 45-46) This is just a guess, however, and might be incorrect.
Abstract:
It would be helpful to have the mode of survey administration noted in the methods section of the abstract.
Introduction:
The authors make several interesting points in the introduction but do not tie those points together in a clear conceptual argument. There are several leaps of logic. For example, lines 68 and 69 read: "habits in saving for emergencies were less likely to experience economic hardship (Gjertson, 2016). In this regard, maintenance of cash reserve for emergencies may help to lower the likelihood of negative mental and behavioral health outcomes . . . . " Gjertson was examining the relationship between having emergency savings and the experience of economic distress (material hardship). How does her study support a hypothesized relationship between cash reserve and mental health outcomes? The use of the Gjertson article would suggest that the authors are arguing that economic distress mediates the relationship between cash reserves and mental health outcomes?
Lines 60 and 61 state: "Identification of other factors associated with depression, anxiety, and other behavioral health outcomes thus have implications for public health during this pandemic." This statement appears to be related to the intended contribution of the research. It would benefit from an explanation of the implications. Why is identifying other factors important? Relatedly, lines 83 and 84 state: "Empirical data on such prevalence and associations can provide useful basic information for stakeholders in economics and behavioral health." Are there examples the authors can provide? Is there an explanation of why such individuals need this basic information?
There are also several typographical errors and some spacing issues in this section and throughout the paper.
The reader would benefit from an explanation of the behavioral health disorders (gambling, drinking, smoking). No literature was presented to support the inclusion of these outcomes. No attempt was made to explain why mental health (anxiety and depression) and behavioral health (gambling, drinking, smoking) were both included or how they relate to each other.
No hypotheses or research questions were presented for either outcome.
Materials and Methods:
This section is second in length to the discussion (about 30% of the paper). There is tremendous detail related to the data collection methods, training the data collection staff, etc. It seemed a little disorganized as some of this information was presented in study design and setting. Then again in sampling methods, data collection and quality control, monitoring and data cleaning, and ethical considerations. While these subsections have different purposes, those purposes could be made a little clearer. I think that might help the reader appreciate what was done and why it is important for them to know.
In the methods section, two things are sprung on the reader without advance notice in the introduction. One is the behavioral health outcomes and the other is the past 30 day economic distress. These constructs should be introduced in your introduction and incorporated into your conceptual argument.
Were the data collection teams assigned to a region? Or did they all work across regions? If assigned to a region, how did you control or adjust for that?
How did you identify your weights? What adjustments were made to address the complex survey?
Results:
I was expecting a much more complete explanation of your findings. Other than the more robust description of findings related to sample characteristics, this section did not provide more detail than the abstract. I was also expecting to see the details of your logistic regression. Where is your pseudo-R-square? -2LL? Percent classified correctly? You have created an interaction between economic distress and emergency cash reserves yet you do not present the results of the interaction. You also do not consider main effects of economic distress and emergency cash reserves. These details would help your reader understand what you found and set them up to understand what it means in your discussion. Having one or more research questions would help you present the results in a more intentional way.
Discussion:
The discussion would also benefit from having research question(s).
The organization of this section could be improved to support the reader's ability to digest the points being made. For example, the authors discuss limitations of the measures beginning on line 294. Then they present limitations again starting on line 327.
Lines 294 to 300 should be examined. It is difficult to understand the point(s) being made in the current language.
Conclusion:
The conclusion re-presents the findings rather than providing a summary or conclusion for the paper.
Other than the typographic errors, the spacing issues and a few lines, the paper is written clearly.
This paper reports on a cross-sectional study rather than an experiment.
Additional explanation of the methods and alignment between the methods, research question, and results would be helpful. What the authors describe as their method: "We used multivariate logistic regression models to assess the extent that the prevalence of the outcomes differed between each population groups with adjustments for experience of economic distress within 30 days prior to the survey (if applicable) and demographic and socioeconomic characteristics" (lines 187 to 190); is not derived from a research questions or related to what was presented in the results.
There is no research question.The gap and the contribution (intended or actual) are not clear.
There are no research questions.
Presentation of the findings seems rushed and limited. The discussion has considerable detail but does not draw important conclusions (perhaps due to the lack of a clear research question).
The paper has potential. The points I have raised are more related to the presentation of a conceptual argument, methods, results, etc. than to the underlying data.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.