Measuring Trump’s effect requires dating the post-period from election, not campaign.

Evidence

To examine the effect of President Trump on religious content and coverage, we set two time periods of interest. The first served as a control period, encompassing the 18 months before President Trump announced his candidacy (December 16, 2013 to June 16, 2015). The experimental time period included the 18 months directly after President Trump was elected to office (November 9, 2016 to May 9, 2018). We chose to examine the period of time after President Trump won the general presidential election, rather than the months when he was campaigning, to ensure that any pre-post effects could be contributed to President Trump’s Administration, as opposed to the campaign rhetoric of his opponents during the election season. — Emily Greene-Colozzi, Amy Adamczyk, and Senahan Kiyal Keles, Differences in Religious Framing of…, 43

Categorization

• XREF: Natural-experiment window design echoes interrupted time-series logic used across political communication research.