Data & Coding framework for article

We used the Factiva news database to collect the full-text articles from both countries searching for articles from January 1, 2015 through January 31, 2021. We mitigated issues of “completeness” (Lacy et al., 2015, p. 795) by comparing our search results on Factiva, LexisNexis, and the NewsAPI. The Factiva database not only aggregated the results from LexisNexis and the NewsAPI, but also included additional publications. To eliminate differences in search results, we collected our sample of articles from a single Factiva account. In addition, since selecting search terms can be a “subjective decision” (Lacy et al., 2015, p. 794), to increase “content validity” (Lacy et al., 2015, p. 794), we collected articles with at least three mentions of the term “tampon tax” or at least three mentions of the term “menstrual equity,” an umbrella term used to describe equal and comprehensive access to menstrual products. Since these terms are used in similar contexts (Weiss-Wolf, 2017), we included both to increase our chances of drawing a representative sample size. Our Factiva query returned 159 articles across both countries. Of those, 65 articles were from U.S. publications and 94 articles were from U.K. publications. Using Stryker et al.’s (2006) validation metrics for calculating the external validity of a sample of articles, our query terms retrieved 98% (± 5) of all tampon tax stories and 92% of the retrieved stories were relevant (see Appendix, Table A). Thus, our correction coefficient across our entire sample is .94 bringing the sample size closer to 149 when correcting for the sampling error associated with the search strings (Stryker et al., 2006). As such, our sampling methods meet the requirements for external validity. After excluding PR and University Wires, including just U.S.-based stories (i.e., excluding stories published by U.S. publications about the tampon tax abroad), and selecting only relevant results, our final U.S. sample included 47 articles. As Table 1 demonstrates, a total of 32 articles were published by national publications (68% of U.S. sample, 30% of total sample) and 15 articles were published by local publications (32% of U.S. sample, 14% of total sample). The U.S. articles represent 44% of the total sample. For our U.K. sample, we followed a parallel process with one exception: Since a general search on Factiva returned hundreds of articles, we selected publications based on national newspapers listed in the 2021 “Who Owns the Media Report” by Media Reform (Coalition, 2021). The Sun on Sunday was the only publication from this list that was not available on Factiva. After excluding any irrelevant results from the 94 results, including 6 articles covering a controversy surrounding a pro-life organization being the recipient of a grant from the “Tampon Tax Fund”—a fund of monies levied from taxes on menstrual products distributed to women-focused organizations who applied for the funds—our final U.K. sample included 61 articles.