Analysis of Location and Spatial Distribution of Elderly Women Victims of Gender Violence

Round 1

Reviewer 1 Report

This is highly relevant and important research that has been very long overdue! I especially appreciate its clear and sound outline of applicability. Excellent and very much needed piece of research.

Author Response

Thank you very much for your attention and kind words.

Reviewer 2 Report

This is an interesting study on the geo-locations of domestic violence experienced by elderly women in a specified set of regions in Spain. The interesting aspect of the study is in the finding that lower population density is associated with higher levels of violence specifically in older disabled women -- which entails that there are some mechanisms at play which bring about the outcome. 

The main issue with the study, however, is that it does not elucidate enough mechanistic activity of what is actually causing violence, or potentially causing it. I recommend that the authors delve deeper into the literature on this topic, through reference to how the dependent variable has been observed in other contexts and countries. 

Another issue is observable in the conclusion -- in the point on the lack of spatial data. While I agree that this is indeed factual and necessary to pursue for greater prevention strategies, but there needs to be greater conceptual and theoretical depth to simply identifying correlations through geo-location data. 

Author Response

Thank you very much for your valuable assessment. We have revised the text based on all your suggestions and we believe that it has helped us in our reflection and in-depth analysis. We have improved the theoretical and empirical evidence, and we have tried to facilitate the understanding of the statistical methods.

Author Response File: Author Response.docx

Reviewer 3 Report

The paper focuses on an important and under-reported topic of violence against older women and uses an innovative approach to explore the extent of violence against older women in Spain. However, it does not situate the hypotheses or conclusions in existing theoretical or empirical evidence on risks of violence against women and violence against older women. In addition, the methods and results are not presented in a clear manner that enables ease of understanding and interpretation by the reader. Conclusions are drawn in the discussion which are not supported by the findings presented in the results. Major revisions are needed before this paper can be adequately reviewed for and considered for publication. I have included specific suggestions by section below.

Overall paper:

·       Some of the language used throughout the paper is stigmatising and is not aligned with best practice for research on violence against women or gender-based violence.  For example, consider using “survivor” instead of “victim”, and try to avoid making statements that make assumptions about women’s situations and their agency or self-perception that are not supported by evidence or your study results, such as stating that older women in general do not know they are even experiencing violence.

·       The paper could benefit from copyediting in English to ensure that concepts are communicated clearly.

·       Most of the citation links in the paper did not work (I assume this is the issue? They all show up as question marks: [?]).  I cannot view the majority of the citations in the document and confirm that foundational evidence and theory are referenced.  Please fix the citation formatting.

Abstract:

·       The first sentence of the abstract (repeated again in the Introduction) is an assumption that is not rooted in evidence and should be removed or revised. While some older women may not be aware that their experiences are considered violence or abuse, many may be aware but social norms, their socioeconomic situation, and other factors may prevent them from acknowledging this publicly or removing themselves from their situation.

·       The abstract would be strengthened by including more information on the study design and the results.

Introduction:

·       Page 1, line 15: Please define the type of violence that you are examining more clearly in the introduction.  Elder abuse is not always gender-based and can be against both older men and women.  If you are speaking specifically about gender-based violence against older women, then please make this more clear in the introduction.

·       I recommend including more evidence on rates of violence against older women and potential risk factors.  While it is true that this is an under-reported issue, there is existing evidence that should be included in the introduction, such as these reviews:

o   Meyer, S. R., Lasater, M. E., & García-Moreno, C. (2020). Violence against older women: A systematic review of qualitative literature. PloS one, 15(9), e0239560.

o   Straka, S. M., & Montminy, L. (2006). Responding to the needs of older women experiencing domestic violence. Violence against women, 12(3), 251-267.

o   Bows, H. (2018). Sexual violence against older people: A review of the empirical literature. Trauma, Violence, & Abuse, 19(5), 567-583.

·       Or this multi-country prevalence study: Luoma, M. L., Koivusilta, M., Lang, G., Enzenhofer, E., Donder, L., Verté, D., ... & Penhale, B. (2011). Prevalence study of abuse and violence against older women: results of a multi-cultural survey conducted in Austria, Belgium, Finland, Lithuania, and Portugal. National Institute for Health and Welfare.

·       I also recommend further reflection on why violence against older women is underreported.  For example, many approaches to measuring violence against women only include women of reproductive age. There is a lot of discourse on this that is missing from this paper and should be included either in the introduction or in the discussion. For example, this commentary: Crockett, C., Brandl, B., & Dabby, F. C. (2015). Survivors in the margins: The invisibility of violence against older women. Journal of Elder Abuse & Neglect, 27(4-5), 291-302.

Materials and methods:

·       Page 2, line 50: The second paragraph of this section should be moved to the introduction, as it is a reflection of why there is a dearth of data on gender-based violence against older women and does not explain how it was measured for the purposes of your study.

·       Page 3, line 68: The prevalence of gender-based violence in Spain should be included in the introduction, not in the methods section. I suggest moving it to the section above.

·       Your methods section should include more than just your statistical inferences and equations. It should include subsections with more detail on the following aspects of methods:

o   Study design – is this cross-sectional data? Where is the data from? Is the data from more than one source? If so, how was it combined for your analyses? This is extremely unclear from the way you have structured this section.

o   Sample – from how many municipalities in Galicia is your data? What characteristics of this region are important? What is the sociodemographic or geopolitical history?

o   Measures – How were your key variables chosen and measured? How do they represent the concepts that you are examining? For example, what are complaints of gender-based violence and why was this used rather than prevalence from the CIS? Why is this number divided by the number of women by judicial partner (as stated in Step 6)? Why do you use percentage of foreigners or social protection actions? How are these relevant? What are they and what do they tell us as constructs? The theoretical justification for using these constructs should also be included in the introduction. Are they risk factors? If so, provide theoretical or empirical justification in the introduction.

o   Data analysis – This could include the steps of your process of statistical inference and more detail on how you analysed the data.

·       Page 3, lines 76-80: Why are you calculating an estimation of the victims of gender-based violence based on the national prevalence rate if you have data on the prevalence of gender-based violence in Spain from 2019? Can you not use the individual- or municipality-level data from that sample of 10,000 women from the CIS? There are many limitations to proposed approach.

·       Page 3 point 7: The typical denominator for calculating a prevalence rate of violence against women is to use the total population that would have been asked the question.  So in this case, it should be the total number of women over 64 years of age who report experiencing GBV / total population of women over 64 years of age. This would give you the rate of elderly women who have experienced GBV.

·       Page 4 point 8: What is the interquartile interval? What does this represent?

·       Page 4 point 9: Why are you calculating the population under 16 years of age if you are examining violence against older women?

Results:

·       The quartiles are not explained elsewhere until mentioned for the first time on page 10. This should be up front in the methods section. Much of the data presented in your text and your tables are not sufficiently or clearly labelled or described. What is the interquartile analysis telling us? What are these different municipalities? This is important for the reader to understand the tables that are organised by quartile (tables 4-12).

·       The tables and figures are not presented in a logical manner in line with the explanation of results in the text. Why is population density (Figure 5) presented so late, after the figure on rate of women experiencing GBV by municipality (Figure 4), when this is described earlier in the results? I suggest ordering the tables and figures in the order in which the results are described in the text.

·       Please include p-values or other stats to support your results and conclusions about higher gender-based violence rates among different sub-populations. For example on page 6, line 170, where you state “the data show that gender violence against women has a comparatively higher incidence in small rural municipalities than in urban ones…” What is the statistical validity of this statement? This was not presented in the text or the tables. Do this throughout your presentation of results and consider including in the tables where relevant.

Discussion:

·       The discussion section would benefit from further interpretation of the results before delving into supporting findings in other studies. What do the results of your study mean? Then support this interpretation with existing evidence.

Author Response

Thank you very much for your valuable assessment. We have revised the text based on all your suggestions and we believe that it has helped us in our reflection and in-depth analysis. We have improved the theoretical and empirical evidence, and we have tried to facilitate the understanding of the statistical methods.
All changes have been introduced in the text and marked in red.
1. We have tried to improve the language and prevent it from being stigmatizing; however, we have reviewed the literature and official documents of the Spanish government, as well as studies and institutional reports in our country, and it turns out that "victim" is the concept used at the legal level in Spain. We have asked about this and they have told us that we should not decontextualize the concepts.
2. We have corrected the general sentence that “older women don't know they are suffering…” which is not true.
3. English will be checked by a native speaker.
4. The citation links were not working. It is fixed.
5. We have removed the summary phrase repeated in the introduction that you suggested.
6. We have strengthened the summary.
7. We have clearly defined the purpose of the investigation in the introduction section, as you have suggested.
8. We have reviewed and introduced the suggested bibliography. Thank you very much for the suggestions. We believe that it helps to better frame the investigation in the known background.
9. In materials and method, we have transferred paragraphs to the introduction as suggested, as well as improved the expression in order to favor understanding by the reader.
10. Regarding your question about the calculations at the municipal level using the national prevalence of the Ministry of Equality of the Government of Spain; I want to tell you that methodologically the CIS macro-survey with 10,000 women is carried out every year in a stratified manner and proportional to the number of inhabitants of the municipalities, following a statistical method endorsed by the CIS itself.
11. We have explained the interquartile range.
12. We have improved the presentation of the results and the inclusion of p-values.
13. We have revised the discussion based on the new suggested bibliographical references.

Thank you very much for your time and attention

Author Response File: Author Response.docx