Societal Perceptions and Understanding of Voyeurism & Upskirting in Young Adult Singaporean Nationals: A Reflexive Thematic Analysis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript provides a qualitative analysis of Singaporean adults’ perceptions of voyeurism and upskirting (VU). In this manuscript, authors note two main themes: the unaccountability of perpetrators and the burden of victimization. This article has several strengths, including a grounded background and rich descriptions and contextualization of participants’ narratives. However, I have several concerns that, if addressed, would markedly improve the manuscript. My three main comments are as follows:

1. The manuscript positions itself to understand “Singaporean nationals” perceptions of VU. However, “Singaporean nationals” are a heterogeneous population. By the authors’ own admission, their sample size is likely unrepresentative of this population, by virtue of age (mid 20s to early 30s), technological access/use (recruitment was done through social media), and possibly other demographic characteristics – all of which may have affected the themes identified in the study. The sample size for this study is also quite small, even for qualitative studies; 10 interviews may be enough to reach saturation for homogenous populations (https://doi.org/10.1016/j.socscimed.2021.114523), but again, the authors don’t purport to be studying a homogenous population. Authors should provide additional details about their participant demographics (e.g., exact age range, and descriptions of participants’ race, religion, education level, employment status). In addition, authors should either clarify in their title and manuscript the population they can reasonably generalize to (e.g., “young adult Singaporean nationals”) or conduct additional interviews with Singaporean adults who are not currently represented in their sample.
2. The manuscript should more clearly describe its unique contribution to the literature. The authors identify numerous points where their findings concur with past literature, and their conclusion states their findings are “consistent with Lister and Gannon’s (2024) Descriptive Model of Voyeuristic Behavior and Harper et al.’s (2021) conceptualisation of IBSA” – both frameworks developed in Western contexts (and specifically the UK). The only connection to the Singapore context in the concluding paragraph is about “Singapore’s conservative disposition surrounding sexual topics”, which is a generic characterization that also describes many Western countries. Are the authors suggesting that our understanding of VU perceptions in Western contexts is generalizable to Singapore? If not, then authors should better highlight where their study provides nuance or new information to these conceptual frameworks.
3. Although the manuscript is well-written overall, there are multiple grammatical errors. In the specific comments I note several sentence fragments, but these comments are not exhaustive. Authors should undertake a thorough edit of the manuscript.

In addition to these overall comments, I have noted several specific comments below, with line numbers for ease of reference. I thank the authors for the opportunity to review their manuscript.

• Lines 44-45: The statement after the semi-colon is a bit confusing. Consider rephrasing.
• Lines 61-64: These sentences are grammatically incorrect; rephrase.
• Lines 68-69: Is Singapore one such country with “already low levels of support services for women and girls”?
• Lines 81-82: This is a sentence fragment; please rephrase.
• Participants: Please provide more detail about participants’ demographics.
• Procedure and Materials: were interviews conducted in English, another local language, or both?
• Limitations: The recruitment strategy may also have contributed to the lack of generalizability in the sample. By recruiting through social media, the authors minimized participation of adults with less technology access.
• Lines 481-483: This implication for educators is vague and doesn’t seem to be particularly grounded in the current study.
• Line 486: What is meant by messaging “should be one of belief and trust”?

Author Response

Reviewer comments and revisions

Thank you for the opportunity to submit a revised draft of this manuscript for publication. We appreciate the time and effort that you and the reviewers dedicated to providing useful feedback on our manuscript and are grateful for the insightful comments made for improving our paper. We have incorporated most of the suggestions made by the reviewers, and have explained in detail the few suggestions that we have not actioned in this round. Please see our responses below, in blue, to each of the comments made by the reviewers’.

 

Reviewer 1.

Comment: The manuscript positions itself to understand “Singaporean nationals” perceptions of VU. However, “Singaporean nationals” are a heterogeneous population. By the authors’ own admission, their sample size is likely unrepresentative of this population, by virtue of age (mid 20s to early 30s), technological access/use (recruitment was done through social media), and possibly other demographic characteristics – all of which may have affected the themes identified in the study. The sample size for this study is also quite small, even for qualitative studies; 10 interviews may be enough to reach saturation for homogenous populations (https://doi.org/10.1016/j.socscimed.2021.114523), but again, the authors don’t purport to be studying a homogenous population. Authors should provide additional details about their participant demographics (e.g., exact age range, and descriptions of participants’ race, religion, education level, employment status). In addition, authors should either clarify in their title and manuscript the population they can reasonably generalize to (e.g., “young adult Singaporean nationals”) or conduct additional interviews with Singaporean adults who are not currently represented in their sample.

Response: Thank you for the constructive feedback regarding the sample.

We agree that the term ‘Singaporean nationals’ are a heterogenous population, and that our sample only reflects a narrow group within this. As such, we have adjusted the title to reflect the focus on young adults in Singapore, as suggested, and have made this focus more apparent throughout the manuscript. We have also elaborated on the limitations section to further acknowledge that the sample is not a representative one. This addition reads:

First, the study recruited young adult Singaporeans who were fluent in English via social media. As such, the sample does not represent older adults, non-English-speaking Singaporeans, or those who do not engage in social media. Subsequently, the perspectives shared are from a narrow group of young Singaporean adults, which may not reflect the wider Singaporean culture. For example, the focus on social media discourse surrounding gender norms may be disproportionately salient due to the sample, which might have led to a more Western-aligned view, under-capturing more traditional or conservative values that may influence attitudes in older or less digitally engaged groups. As such, future research could further explore this by accessing a more diverse sample of Singaporean nationals, or by focusing on older Singaporean adults' perceptions, thereby enabling a more substantial understanding of victim-blaming attitudes and their interplay with societal norms.

To address the request for greater demographic detail, we have also provided a table detailing each participants’ demographic details and included the full age range. The table has been added within the Participants section.

In response to the concerns about sample size, we have used Braun and Clarke’s Reflexive Thematic Analysis (RTA), which prioritise depth, richness and reflexivity over representativeness and numerical frequency. Within RTA, small samples are common and appropriate because analytical depth is prioritised over numerical representativeness, wherein studies with some heterogeneity, smaller samples can be sufficient when interviews generate substantial information power. Whilst data saturation is commonly used in qualitative research as the ‘gold standard’ for justifying sample sizes, it is increasingly criticised for being poorly defined, inconsistently applied and for being rooted in positivist assumptions that are incompatible with several forms of qualitative analysis (Braun & Clarke, 2021; O’Reilly & Parker, 2012; Younas et al., 2025). Thus, data saturation is incompatible with RTA, with Braun and Clarke explicitly stating it should not be used to determine sample sizes or as a quality marker in RTA as it’s positivist assumptions that themes ‘emerge’ from the data contradict the assumptions of RTA where themes are subjective and actively generated by the researcher through their interpretations (Braun & Clarke, 2016; Braun & Clarke, 2021). Braun and Clarke describe data saturation as a poor-quality practice that wrongly merges incompatible methodological assumptions (Braun & Clarke, 2020, for more information: https://doi.org/10.1080/14780887.2020.1769238), and they particularly caution against including mention to data saturation within papers during the peer review process (Braun & Clarke, 2021; for more information: https://doi.org/10.1080/2159676X.2019.1704846) Given this, and the changes we have made within this round of reviews, we consider the 10 interviews documented to have provided sufficient depth for producing coherent, well-evidenced themes aligned with our research aim. We have now added a clearer rationale for the sample size to the method section (see page ...). This reads:

In line with Reflexive Thematic Analysis (RTA), sample adequacy is not determined by numerical thresholds or by achieving data saturation, which is conceptually incompatible with RTA’s epistemological stance (Braun & Clarke, 2020; 2021). Instead, adequacy is evaluated through the richness and depth of the data and the degree of information power present within the sample (Malterud et al., 2016). Given the study aims, the moderately focused population (young Singaporean adults who use social media and are fluent in English), and the richness of the interviews, the sample of ten participants provided sufficient information power to develop coherent, meaningful themes within this analytic framework.

 

 

Comment: The manuscript should more clearly describe its unique contribution to the literature. The authors identify numerous points where their findings concur with past literature, and their conclusion states their findings are “consistent with Lister and Gannon’s (2024) Descriptive Model of Voyeuristic Behavior and Harper et al.’s (2021) conceptualisation of IBSA” – both frameworks developed in Western contexts (and specifically the UK). The only connection to the Singapore context in the concluding paragraph is about “Singapore’s conservative disposition surrounding sexual topics”, which is a generic characterization that also describes many Western countries. Are the authors suggesting that our understanding of VU perceptions in Western contexts is generalizable to Singapore? If not, then authors should better highlight where their study provides nuance or new information to these conceptual frameworks.

Response: This is an EXCELLENT point, we have adapted the opening of our Conclusion section (see below) to address this more comprehensively and have also made substantial tweaks to our limitations to pick up on some of these elements. We hope that this helps to absolve this issue but would be willing to tweak further should you have any further ideas/thoughts.

 

“This study delineated the understanding and perceptions of VUs, motivations to perpetrate VUs, and the impacts thereof within a sample of young adults from Singapore. Consistent with Harper et al.’s (2021) conceptualisation of upskirting, participants attributed both sexual and non-sexual motivations to engaging in VUs, including social (e.g., technological affordance) and individual (e.g., sexual deviancy) influences. Moreover, our data – to some extent - aligned with Lister and Gannon’s (2024) Descriptive Model of Voyeuristic Behavior, in that an array of affective, behavioural, cognitive, and contextual factors were considered, by our participants, to play a key role in ones’ propensity to engage in voyeuristic behaviour. Whilst important to note that our data is not derived from those who have committed VUs themselves, we offer unique insight to these extant models through discussion of the role of societal expectations (e.g., those pertaining to modesty and sexual taboos) and how this might facilitate harm minimisation.”

 

Comment: Although the manuscript is well-written overall, there are multiple grammatical errors. In the specific comments, I note several sentence fragments, but these comments are not exhaustive. Authors should undertake a thorough edit of the manuscript.

Response: Thank you for this comment. Following our substantial edits to the manuscript, we have read through the final submission multiple times and are now happy with the readability and accessibility of it.

 

Comment: In addition to these overall comments, I have noted several specific comments below, with line numbers for ease of reference. I thank the authors for the opportunity to review their manuscript.

1. Lines 44-45: The statement after the semi-colon is a bit confusing. Consider rephrasing.
2. Lines 61-64: These sentences are grammatically incorrect; rephrase.
3. Lines 68-69: Is Singapore one such country with “already low levels of support services for women and girls”?
4. Lines 81-82: This is a sentence fragment; please rephrase.
5. Participants: Please provide more detail about participants’ demographics.
6. Procedure and Materials: were interviews conducted in English, another local language, or both?
7. Limitations: The recruitment strategy may also have contributed to the lack of generalizability in the sample. By recruiting through social media, the authors minimized participation of adults with less technology access.
8. Lines 481-483: This implication for educators is vague and doesn’t seem to be particularly grounded in the current study.
9. Line 486: What is meant by messaging “should be one of belief and trust”?

 

Response: Thank you for this feedback. We have made changes to the suggested line changes to improve readability, and have added additional participant demographic information as well as expanded on the limitations section (see response to comment 1). We have made it clear in the procedure and materials section that all interviews were conducted in English, and have expanded on the implications for educators to make this stronger. We have also made it clear by what is meant by ‘should be one of belief and trust’ which was in reference to improving the police response to VUs.

We note your comment about educators, and have both adapted the line itself, and added a caveat in the hope of future research being undertaken within this group, specifically:

“Of course, there is need to work with this group specifically to best understand what is/is not currently being done to tackle this on a micro level, and the barriers educators face.”

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the opportunity to read this research. It is not only a much needed piece of research, but it is excellent. The theoretical framing, the depth of analysis, the linking with broader issues and wider literatue, the writing, and the structure are all very good. It was a pleasure reading it. I have only a few comments.

1. I recommend adding a thematic map to show the relationships between your themes and subthemes
2. Here “Singapore represents an under-reached context for empirically understanding VUs, 42 which are governed compositely under the Criminal Law Reform Act 2019 (s 377BB); 43 legislation that builds upon previous laws that depicted VUs as an “insult of modesty” 44 (Penal Code 2008, s 509) and targeted “obscenity” and “pornography” over consent (Films 45 Act, 1998)” – could you please clarify what you meant by depicting Vus as insult to modesty etc over consent? Do you mean to say that the previous laws dismissed the importance of framing VUs through a consent-focused lense? And does the new 2019 Act now address consent? I advise providing clarity on the legal background currently in place for VUs in Singapore, especially since the recent legal reform is mentioned again in the conclusion, and your implications also mention how legislators would benefit from your research.
3. You discuss in your limitations how the analysis would have benefitted from the voices of victim-survivors. I believe it is also worth mentioning the benefit of having heard from perpetrators, especially in trying to understand motivations. A smaller but important limitation in my view is that the interviews were conducted in English. I understand that the sample were fluent speakers, and there is an argument to be made for people finding it easier to discuss sensitive topics not in their native language, but I do wonder whether the dynamics and depth were affected by using a second language (perhaps Marschan-Piekkari, R., & Reis, C. (2004). Language and languages in cross-cultural interviewing could be of use to you)
4. Here ” how messaging around VUs specifically, and image-based offences more broadly should be one of belief and trust.” – I believe you mean to say that police should foster an environment where victims are believed, from reporting to investigation, correct? I would briefly expand on this to make it clearer how this change in policing culture (which unfortunately is present across countries) would help the victim trajectories in the system.
5. You say in your implications that legislators should “to prioritise such offences”. Prioritise in what way? How? When? I believe this section could be clarified or expanded a bit especially since the legal background is an important part of your analysis.

Author Response

Reviewer comments and revisions

Thank you for the opportunity to submit a revised draft of this manuscript for publication. We appreciate the time and effort that you and the reviewers dedicated to providing useful feedback on our manuscript and are grateful for the insightful comments made for improving our paper. We have incorporated most of the suggestions made by the reviewers, and have explained in detail the few suggestions that we have not actioned in this round. Please see our responses below, in blue, to each of the comments made by the reviewers’.

 

Reviewer 2 comments:

Thank you for the opportunity to read this research. It is not only a much needed piece of research, but it is excellent. The theoretical framing, the depth of analysis, the linking with broader issues and wider literature, the writing, and the structure are all very good. It was a pleasure reading it. I have only a few comments.

 

Very kind of you.

 

Comment: I recommend adding a thematic map to show the relationships between your themes and subthemes

Response: A thematic map has been added to show the relationship between the themes and subthemes at the beginning of the results section (see figure 1, within the Findings section). 

Comment: Here “Singapore represents an under-reached context for empirically understanding VUs, 42 which are governed compositely under the Criminal Law Reform Act 2019 (s 377BB); 43 legislation that builds upon previous laws that depicted VUs as an “insult of modesty” 44 (Penal Code 2008, s 509) and targeted “obscenity” and “pornography” over consent (Films 45 Act, 1998)” – could you please clarify what you meant by depicting Vus as insult to modesty etc over consent? Do you mean to say that the previous laws dismissed the importance of framing VUs through a consent-focused lense? And does the new 2019 Act now address consent? I advise providing clarity on the legal background currently in place for VUs in Singapore, especially since the recent legal reform is mentioned again in the conclusion, and your implications also mention how legislators would benefit from your research.

Response: Thank you for this thoughtful query. Upon re-reading, we very much agree that despite the wording being factually correct, its accessibility and understanding is quite vague in nature, we have therefore simplified it to enhance engagement, it now reads as follows:

 

“Singapore represents an under-reached context for empirically understanding VUs, which are governed compositely under the Criminal Law Reform Act 2019 (s 377BB). This legislation employs a consent-focused lens, which was absent in previous governance of VUs, which centred around whether a victim’s modesty was deemed to have been “insulted” (Penal Code 2008, s 509), and which focused on public morals and societal protection from explicit material (Films Act, 1998). However, the Criminal Law Reform Act 2019 still fails to address the multi-faceted nature and manifestations of VUs, as well as the underpinning motivations to capture, distribute, and/or threaten to distribute intimate media (Vitis, 2021).”

 

 

Comment: You discuss in your limitations how the analysis would have benefitted from the voices of victim-survivors. I believe it is also worth mentioning the benefit of having heard from perpetrators, especially in trying to understand motivations. A smaller but important limitation in my view is that the interviews were conducted in English. I understand that the sample were fluent speakers, and there is an argument to be made for people finding it easier to discuss sensitive topics not in their native language, but I do wonder whether the dynamics and depth were affected by using a second language (perhaps Marschan-Piekkari, R., & Reis, C. (2004). Language and languages in cross-cultural interviewing could be of use to you)

Response: Thank you for this suggestion. We had now mentioned the recognition of perpetrators and a paragraph on and a paragraph to the limitations reflecting on how speaking in one’s non-native language may have impacted the data. See below:

Thirdly, the interviews were conducted in English, as required by the University’s ethical procedures. While English is widely used in Singapore, conducting interviews in the participant’s non-native language may influence rapport building and the interpretative quality of the data. Marschan-Piekkari & Reis (2004) argue that cross-cultural interviews conducted in a non-native language may result in more simplified expressions or misunderstandings, even when fluent. Thus, whilst the interviews were conducted by a Singaporean national to reduce power asymmetry and foster cultural alignment, the use of English may still have influenced how participants articulated their views, which may have meant certain cultural nuances were lost. Future research may therefore benefit from conducting interviews in the participants’ preferred language to capture richer linguistic and cultural meanings.   

 

Comment: Here ” how messaging around VUs specifically, and image-based offences more broadly should be one of belief and trust.” – I believe you mean to say that police should foster an environment where victims are believed, from reporting to investigation, correct? I would briefly expand on this to make it clearer how this change in policing culture (which unfortunately is present across countries) would help the victim trajectories in the system.

Response: We have now corrected and expanded on this to make it clearer that we are referring to changing the policing culture. This now reads:

“For law enforcement agencies, this study has implications for how means of reporting need to be clear and easy-to-use, and how messaging around VUs specifically, and image-based offences more broadly, should be more victim-focused. As research shows that victims’ who have negative experiences disclosing their experiences of sexual violence have worse psychological outcomes (Ullman, 2023), adopting a legal environment where victims feel they can trust police and be taken seriously is vital for the long-term recovery of victims. Whilst the need for legal and police reform for VUs is not unique to Singapore, the sociocultural context in Singapore has been found to reduce the effectiveness of legal practice for other forms of violence against women and girls (Ganapathy, 2006).”  

 

Comment: You say in your implications that legislators should “to prioritise such offences”. Prioritise in what way? How? When? I believe this section could be clarified or expanded a bit especially since the legal background is an important part of your analysis.

Response: Thank you for pointing this out. Taking a step back after the comprehensive review of the article, we are cautious about our writing implying that other (non)sexual offences should not be treated with equal rigour. As such, we have removed this wording, with the passage now reading:

 

“For legislators, there is a clear need to improve the communication of laws and consequences pertaining to VUs, especially given the high (known) rates of perpetration in this region.”

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I thank the authors for their careful consideration of my comments, and I find the manuscript is strengthened by the added details. My only subsequent comment is that the addition of the new Figure 1 (which I find to be a nice addition to the paper) makes Table 2 fully redundant, so authors should consider removing Table 2. Otherwise, I have no additional comments.

Author Response

Thank you for your kind comments. We have now removed Table 2.