In fsolt/dcpo_gayrights: Dynamic Comparative Public Opinion

We first want to thank the Editor and the reviewers for their helpful comments and suggestions. At the Editor's suggestion, we revised our original manuscript with a particular eye towards Reviewer 2's comments. We were sure, however, to address Reviewer 1's request that we emphasize that the SGR index measures public opinion throughout the manuscript and that questions of the relationship between such attitudes and the adoption of laws are among those for which we hope the SGR will provide the basis for future research. Below we summarize the additional revisions we made in response to the reviewers' helpful comments:

• Defining the purpose of the manuscript. Reviewer 2 noted that we had described the paper's purpose, in the abstract and introduction, as to 'present' the SGR dataset and questioned what that meant in this context. She or he suggested that describing 'the process of building' the dataset better defines what the paper does. We agree. We revised the abstract and introduction accordingly and took care throughout the paper to emphasize our goal of overcoming the limitations of previously available data.

• Clarifying our focus on "support for gay rights" rather than others. Reviewer 2 asked "Why are gays rights used and not another concept?" We are grateful to the reviewer for pointing out the need to elaborate on this. We added a paragraph spelling out the theoretical and practical reasons for our selection of this narrower concept at footnote 1 on page 3. On the theoretical side, we concur with Worthen's (2013) skepticism that attitudes regarding sexual orientation and those regarding gender identity have the same roots. And practically---even if a more inclusive concept was theoretically justifiable---there are many, many more observations questions bearing on support for gay rights than there are examining attitudes toward gender nonconforming or transsexual people. If included, these few latter questions would have little influence on the resulting estimates, and so claiming the result to be a measure of attitudes toward the broader LGBTQ+ concept would be misleading. Our focus on attitudes toward gay rights therefore maintains alignment among the available original survey questions, the resulting estimates, and the advertised concept.

• Detailed explanation of how we select indicators used in the Support for Gay Rights latent variable model. Reviewer 2 also asked for more detailed explanations on the selection of indicators used in our latent variable model. We added a paragraph (spanning pages 7 and 8 in the revised manuscript) that describes our process for selecting these indicators. To build the SGR dataset, we collected a comprehensive collection of survey questionnaires according to the topic. As mentioned in the revised manuscript, we collected four groups of questions capturing the public latent attitude toward gay rights. First, we collect questions about public approval and acceptance of homosexuals and homosexuality. Second, questions exploring public support for the legalization of same-sex marriage are also considered. Third, we take into account of questions regarding public opinion on the rights of gay, lesbian, and bisexual individuals, including equal rights to heterosexual individuals. Fourth, we also consider questions about attitudes for homosexual individuals’ social and political roles, for instance, running for public office. We also moved the sentence pointing those interested in the complete list of survey items to the online appendix from a footnote to the text.

• Differences of our dataset compared to GAI and limitations of our dataset. Reviewer 2 also recommends that we differentiate our dataset from the already existing dataset, the GAI. Moreover, the reviewer suggests that explanations on potential limitations of our dataset should be provided. We found these suggestions to be very helpful.

  To address them, we first added to the text at page 7 a paragraph outlining the advantages of the SGR dataset over the GAI data. The SGR data are based on a model that better fits ordinal survey data, they avoid conflating attitudes regarding sexual orientation with those of gender identity, and in many countries they draw on surveys conducted over longer time spans. Further, the GAI data do not include a measure of the estimates' uncertainty, which is crucial to include when working with latent variable estimates. And most importantly for the purposes of our goal of providing data to allow researchers to examine many countries over many years, the GAI estimates are only available for a single cross-section, making analysis of change over time impossible.

  We also revised our manuscript to mention, at pages 21-22, two limitations of our dataset. We note there that the coverage of our SGR dataset depends on the availability of the source of survey data; this means that for some countries over-time coverage can be very limited. We further point out that, as the SGR consists of latent variable estimates, researchers using the data must be mindful of the uncertainty in these data and should take it into account in their analyses.

We appreciate for all comments and advice from both reviewers. We believe that our revised manuscript is now much more nuanced compared to the previous version. Again, we thank the reviewers for the valuable and constructive comments and the opportunity to revise our manuscript.

fsolt/dcpo_gayrights documentation built on Dec. 11, 2024, 4:17 p.m.