“…As is common practice with question wording survey experiments, ours included two control conditions—a pure control that conveyed no text prior to our first outcome measure and a regular control that included a generic statement about U.S. military bases providing both benefits and detriments for local communities. Although we find no significant difference in attitudes toward U.S. military bases between the two control conditions, we present treatment effects with respect to both groups, in line with our preanalysis plan and standard recommendations in the methodological literature (Gaines, Kuklinski, & Quirk, 2007; Tingley, 2014, p. 447). Our treatment conditions begin with the generic statement about benefits and detriments from the regular control, followed by additional information about U.S. bases.…”