“As the editorial term ends, I’m both looking back and looking forward . . . so, as promised, here’s a post by Allison Carnegie and Austin Carson describing their recent experience with qualitative verification at AJPS . . . and within the next week I’ll be posting an important update to the AJPS “Replication/Verification Policy,” one that will endure past the end of the term on June 1.”
– Jan Leighley, AJPS Interim Editor
Our Experience with the AJPS Transparency and Verification Process for Qualitative Research
By Allison Carnegie of Columbia University and Austin Carson of the University of Chicago
The need for increased transparency for qualitative data has been recognized by political scientists for some time, sparking a lively debate about different ways to accomplish this goal (e.g., Elman, Kapiszewski and Lupia 2018; Moravcsik 2014. As a result of the Data Access and Research Transparency (DA-RT) initiative and the final report of the Qualitative Transparency Deliberations, many leading journals including the AJPS adopted such policies. (Follow this link for a critical view of DA-RT.) While the AJPS has had such a policy in place since 2016, ours was the first article to undergo the formal qualitative verification process. We had a very positive experience with this procedure, and want to share how it worked with other scholars who may by considering using qualitative methods as well.
In our paper, “The Disclosure Dilemma: Nuclear Intelligence and International Organizations (https://doi.org/10.1111/ajps.12426),” we argue that states often wish to disclose intelligence about other states’ violations of international rules and laws, but are deterred by concerns about revealing the sources and methods used to collect it. However, we theorize that properly equipped international organizations can mitigate these dilemmas by analyzing and acting on sensitive information while protecting it from wide dissemination. We focus on the case of nuclear proliferation and the IAEA in particular. To evaluate our claims, we couple a formal model with a qualitative analysis using each case of nuclear proliferation, finding that strengthening the IAEA’s intelligence protection capabilities led to greater intelligence sharing and fewer suspected nuclear facilities. This analysis required a variety of qualitative materials including archival documents, expert interviews, and other primary and secondary sources.
To facilitate the verification of the claims we made using these qualitative methods, we first gathered the raw archival material that we used, along with the relevant excerpts from our inter- views, and posted them to a dataverse location. The AJPS next sent our materials to the Qualitative Data Repository (QDR) at Syracuse University, which reviewed our Readme file, verified the frequency counts in our tables, and reviewed each of our evidence-based arguments related to our theory’s mechanisms (though it did not review the cases in our Supplemental Appendix). (More details for this process can be found in the AJPS Verification and Replication policy, along with its Qualitative Checklist.) QDR then generated a report which identified statements that it deemed were “supported,” “partially supported,” or “not documented/referenced.” For the third type of statement, we were asked to do one of the following: provide a different source, revise the statement, or clarify whether we felt that QDR misunderstood our claim. We were free to address the other two types of statements as we saw fit. While some have questioned the feasibility of this process, in our case it took roughly the same amount of time that verification processes of quantitative data typically do, so it did not delay the publication of our article.
We found the report to be thorough, accurate, and helpful. While we had endeavored to support our claims fully in the original manuscript, we fell short of this goal on several counts, and fol- lowed each of QDR’s excellent recommendations. Occasionally, this involved a bit more research, but typically this resulted in us clarifying statements, adding details, or otherwise improving our descriptions of, say, our coding decisions. For example, QDR noted instances in which we made a compound claim but the referenced source only supported one of the claims. In such a case, we added a citation for the other claim as well. We then drafted a memo detailing each change that we made, which QDR then reviewed and responded to within a few days.
Overall, we were very pleased with this process. This was in no small part due to the AJPS editorial team, whose patience and guidance in shepherding us through this procedure were greatly appreciated. As a result, we believe that the verification both improved the quality of evidence and better aligned our claims with our evidence. Moreover, it increased our confidence that we had clearly and accurately communicated with readers. Finally, archiving our data will allow other scholars to access our sources and evaluate our claims for themselves, as well as potentially use these materials for future research. We thus came away with the view that qualitative transparency is achievable in a way that is friendly to researchers and can improve the quality of the work.
About the Authors: Allison Carnegie is Assistant Professor of Columbia University and Austin Carson is Assistant Professor at the University of Chicago. Their research, “The Disclosure Dilemma: Nuclear Intelligence and International Organizations (https://doi.org/10.1111/ajps.12426),” is now available in Early View and will appear in a forthcoming issue of the American Journal of Political Science. Carnegie can be found on Twitter at
@alliecarnegie and Carson at @carsonaust.
References
Elman, Colin, Diana Kapiszewski and Arthur Lupia. 2018. “Transparent Social Inquiry: Implica- tions for Political Science.” Annual Review of Political Science 21:29–47.
Moravcsik, Andrew. 2014. “Transparency: The Revolution in Qualitative Research.” PS: Political Science & Politics 47(1):48–53.
Speak Your Mind