Housing, Instability, and Discrimination amongst Takatāpui/LGBTIQ+ Youth in Aotearoa New Zealand

Round 1

Reviewer 1 Report

The article addresses a relevant research problem and uses language in an instigating way to value local knowledge. There are, however, problems in the design of the research and in the presentation of the results and discussion:

1) When online survey data are presented, it is impossible to compare the surveyed people with the rest of the population of the same age in Aotearoa/New Zealand. Therefore, the article would benefit from comparing the obtained results with the ones from the surveys cited as the basis for part of the proposed questions: "New Zealand Census, the General Social Survey, the Household Economic Survey, and Te Kūpenga Māori Social Survey."

2) The article describes total percentages but does not make it clear whether there were statistical differences between the Maori (17% of the sample) and the full sample.

3) The article addresses the "significant association between homelessness and experience of formal or informal state care poverty," which is somewhat obvious since it is precisely the people with the highest "degree" of homelessness that require more social protection from the state.

4) The article cites results about the self-reported physical and mental health self-report, which does not allow evaluations such as those made in the discussion since there is no comparison of the answers with the Aotearoa/New Zealand youth. Are these levels high? 

5) It is also necessary to present more clearly the standards used to assess annual income (currency, minimum wage in the country, etc.) since age is associated with lower income among youth in many countries. 

Author Response

Response to reviewers - Housing, Instability, and Discrimination Amongst Takatāpui/LGBTIQ+ Youth in Aotearoa New Zealand

Thank you to the reviewers for their time, and thoughtful comments on our paper. We have made the requested changes, and feel that this has strengthened our paper. We have listed the reviewers' comments below and responded to each one (see the bullet points below each comment). We have also moved the text from our footnotes into the body of the paper as per journal guidelines.

Reviewer One:

The article addresses a relevant research problem and uses language in an instigating way to value local knowledge. There are, however, problems in the design of the research and in the presentation of the results and discussion:

1) When online survey data are presented, it is impossible to compare the surveyed people with the rest of the population of the same age in Aotearoa/New Zealand. Therefore, the article would benefit from comparing the obtained results with the ones from the surveys cited as the basis for part of the proposed questions: "New Zealand Census, the General Social Survey, the Household Economic Survey, and Te Kūpenga Māori Social Survey."

• We have added a paragraph to the introduction which provides information from these surveys in regards to young people and their housing experiences (see lines 102-117). However, these are not specific to Takatāpui/LGBTIQ+ communities. As we had already noted in our introduction (see lines 72-74), there is no whole-of-population data in Aotearoa on Takatāpui/LGBTIQ+ communities. Most of these Statistics New Zealand surveys do not collect such data, or have only begun to in recent years (and in those cases, results/outputs are scarce, but we had already included them where possible). This means that when comparing to this SNZ data we are comparing to the whole youth population in Aotearoa. We look forward to one day having whole-of-population data on Takatāpui/LGBTIQ+ communities!

2) The article describes total percentages but does not make it clear whether there were statistical differences between the Maori (17% of the sample) and the full sample.

• We have amended Table 2 to show those results for the full sample, Māori, and non-Māori, as well as the p values for differences between Māori and non-Māori. Significant differences have been emphasised in bold.

3) The article addresses the "significant association between homelessness and experience of formal or informal state care poverty," which is somewhat obvious since it is precisely the people with the highest "degree" of homelessness that require more social protection from the state.

• We agree this is a somewhat expected finding. However, since there is so little research in Aotearoa which looks at the relationship between state care and homelessness, particularly for Takatāpui/LGBTIQ+ communities, we felt it important to include this. We note that we looked at whether or not experiences of care could lead to experiences of homelessness, rather than the other way around. We have clarified this in the discussion (see line 338). We have also added reference to a report which was fortuitously released while we were completing these edits (see lines 64-69).

4) The article cites results about the self-reported physical and mental health self-report, which does not allow evaluations such as those made in the discussion since there is no comparison of the answers with the Aotearoa/New Zealand youth. Are these levels high?

• We have added in information from the General Social Survey about self-reported mental and physical health which allows us to compare Takatāpui/LGBTIQ+ communities with non-Takatāpui/LGBTIQ+ populations (see lines 95-100). Unfortunately, this data is not currently available broken down by age and LGBTIQ+ identities vs non-LGBTIQ+ identities, so we are unable to fully determine whether or not our reported levels are congruent. We do have limited data from the General Social Survey about the entire youth population, which we have included in our new paragraph on youth data not specific to Takatāpui/LGBTIQ+ communities (see lines 115-117).

5) It is also necessary to present more clearly the standards used to assess annual income (currency, minimum wage in the country, etc.) since age is associated with lower income among youth in many countries.

• We have clarified in the methods that we use New Zealand Dollars, and provided details on minimum wage rates (see lines 169-176). We have added the median annual income for the wider youth population in Aotearoa into the discussion (see lines 402-405) to emphasise the much lower incomes of our sample.

Author Response File: Author Response.docx

Reviewer 2 Report

I'm glad to see this contribution to scholarship about Takatāpui/LGBTQ youth housing instability and discrimination in Aotearoa/New Zealand.  There is a such a lack of data about LGBTQ youth, especially outside the US, and homelessness and discrimination in these groups are extremely high and concerning.  I'm also pleased to see the attention to cultural/local forms of queerness, such as using takatāpui and the Maori name for NZ.  That said, as a foreigner, I had to look up Aotearoa to be sure it was a name for NZ, and not a region within the country--a quick definition would be helpful.  As the author names are masked, I cannot tell if that authors are from Aotearoa, but I surely hope so. 

Main issues:

The reader needs more info about definitions and analysis of the variables:

1) Please define "involuntary mobility" sooner than the discussion. The intro for a basic definition and the methods for a the full operationalization would be great.

2) Also move footnote 6 to the Methods to clearly define homelessness, especially since it's the main outcome. 

3) Please share in table form the homelessness scores, at least the mean, median, IQR, etc. I could not find them.

4) Further, there is no explanation of the analysis, other than you did multiple linear regression. Why linear? Are homelessness scores normally distributed? If not, then dichotomizing the scores and using logistic regression is appropriate. Odds ratios would be easier to interpret than Betas, as well.  Please add an Analysis section to the Methods. to answer questions such as what variables were included in the model? How did you decide on the final model? What p-value cutoff did you use to include/exclude variables from the model?  What did you control for in the model?

Minor editing: 

Abstract line 7, "significant a relationship" should be "a significant relationship"

Intro

line 43-45, replace commas with semi-colons after the words housing, conditions, and markets.

line 48 "importance community engagement" should be "importance of community engagement"

line 92, 133, 140 and 169: Do not start sentence with numerals; either spell out long form, or rephrase to avoid

Results

line 127, I suggest you remove the commentary "interestingly" in the sentence about the proportion of nonbinary folx who consider themselves transgender. It's common. 

Table 1: income $50,000-$10,000 should be $50,000-$100,000, correct?

Table 1: for LGBTQ identities: could participants tick all that apply, since the identities are not mutually exclusive? For example transgender AND queer? Bisexual AND intersex?  Please mention that in the methods and/or table.

Table 2: instead of "type of discrimination," what you have listed seems more like "source of discrimination"? People are not types of discrimination.

Discussion 

line 195: is there anything more updated than Corliss 2011? Anything by Amy Green's team from the Trevor Project's annual LGBTQ youth survey in the US?

line 211: capitalize Further at beginning of sentence

Author Response

Response to reviewers - Housing, Instability, and Discrimination Amongst Takatāpui/LGBTIQ+ Youth in Aotearoa New Zealand

Thank you to the reviewers for their time, and thoughtful comments on our paper. We have made the requested changes, and feel that this has strengthened our paper. We have listed the reviewers' comments below and responded to each one (see the bullet points below each comment). We have also moved the text from our footnotes into the body of the paper as per journal guidelines.

Reviewer Two:

I'm glad to see this contribution to scholarship about Takatāpui/LGBTQ youth housing instability and discrimination in Aotearoa/New Zealand.  There is a such a lack of data about LGBTQ youth, especially outside the US, and homelessness and discrimination in these groups are extremely high and concerning.  I'm also pleased to see the attention to cultural/local forms of queerness, such as using takatāpui and the Maori name for NZ.  That said, as a foreigner, I had to look up Aotearoa to be sure it was a name for NZ, and not a region within the country--a quick definition would be helpful.  As the author names are masked, I cannot tell if that authors are from Aotearoa, but I surely hope so.

• Kia ora, thank you for your kind words. Our apologies for the confusion around terminology – we have clarified that Aotearoa is indeed the Māori name for our country (see lines 23-24). We all live and work/study in Aotearoa.

Main issues:

The reader needs more info about definitions and analysis of the variables:

1) Please define "involuntary mobility" sooner than the discussion. The intro for a basic definition and the methods for a the full operationalization would be great.

• We have included this definition, alongside our definition of homelessness, in a new paragraph in the introduction (see lines 44-60). We have also added details on how we measured involuntary mobility to the methods (see lines 148-153).

2) Also move footnote 6 to the Methods to clearly define homelessness, especially since it's the main outcome.

• We have added a paragraph in the introduction (see lines 44-60) to provide our definition of homelessness, and added the further detail about our measurement in the methods (see lines 141-148).

3) Please share in table form the homelessness scores, at least the mean, median, IQR, etc. I could not find them.

• We have added these to Table 2.

4) Further, there is no explanation of the analysis, other than you did multiple linear regression. Why linear? Are homelessness scores normally distributed? If not, then dichotomizing the scores and using logistic regression is appropriate. Odds ratios would be easier to interpret than Betas, as well.  Please add an Analysis section to the Methods. to answer questions such as what variables were included in the model? How did you decide on the final model? What p-value cutoff did you use to include/exclude variables from the model?  What did you control for in the model?

• We have added more detail right throughout the methods, including an additional paragraph specifically about the multiple linear regression which answer these specific questions (see lines 224-240).

Minor editing:

• Ngā mihi – thank you for pointing out our many minor errors. We have amended all of them (with embarrassment at not having noticed them in the first place!).

Abstract line 7, "significant a relationship" should be "a significant relationship"

Intro

line 43-45, replace commas with semi-colons after the words housing, conditions, and markets.

line 48 "importance community engagement" should be "importance of community engagement"

line 92, 133, 140 and 169: Do not start sentence with numerals; either spell out long form, or rephrase to avoid

Results

line 127, I suggest you remove the commentary "interestingly" in the sentence about the proportion of nonbinary folx who consider themselves transgender. It's common.

Table 1: income $50,000-$10,000 should be $50,000-$100,000, correct?

Table 1: for LGBTQ identities: could participants tick all that apply, since the identities are not mutually exclusive? For example transgender AND queer? Bisexual AND intersex?  Please mention that in the methods and/or table.

• We have clarified this in the methods (see lines 155-167). This reporting on LGBTIQ+ identities included results from several questions for ease of reading/as to not overload the tables; as the sexual orientation, gender, transgender, and intersex questions were all separate, participants could thus select all that apply.

Table 2: instead of "type of discrimination," what you have listed seems more like "source of discrimination"? People are not types of discrimination.

Discussion

line 195: is there anything more updated than Corliss 2011? Anything by Amy Green's team from the Trevor Project's annual LGBTQ youth survey in the US?

• Thank you for the suggestion – we have included a sentence referencing their latest survey, which does indeed include relevant data (see lines 323-324). In using the Corliss reference, we were specifically looking at surveys of LGBTIQ+ youth which had also asked about homelessness, rather than surveys of youth experiencing homelessness which had also asked about LGBTIQ+ identities (of which there are many more recent studies). This was so we could easily compare to our data, and until being directed to this report from the Trevor Project we had been unable to find more recent data. Perhaps a downside to our geographic location and context – not being particularly aware of the types of data that NGOs outside of Aotearoa are likely to collect and report!

line 211: capitalize Further at beginning of sentence

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The changes made include the proposed suggestions. There is, however, a need for changes in language a passage that address the reports made in the research, as suggested:

"Furthermore, our findings show that our sample of Takatāpui/LGBTIQ+ youth is low-income earners experiencing poverty, REPORT HAVING fair to poor mental health, fair to good physical health AND difficulty remembering, and have moved an average of seven times throughout their lives."

Author Response

Kia ora, thank you for pointing out this error. We have amended the wording accordingly.