<p>When and how to come out are difficult choices. In this research project, we examine one form of disclosure: the addition of an LGBTQ keyword to one's online social media profile. We construct daily time series of the prevalence of LGBTQ keywords within American Twitter users' self-descriptions. Further, we construct daily time series of inferred add and drop events. These we compare to relevant annual and one-time events. We confirm or disconfirm our pre-registered hypotheses.</p>
<p>On October 16, 2019, we finalized a list of LGBTQ-relevant events and a set of pre-registered hypotheses. We completed this before constructing the time-series datasets for this project.</p>
<li><a href="https://osf.io/zc7qd/" rel="nofollow">LGBTQ-relevant events</a></li>
<li><a href="https://osf.io/xrck2/" rel="nofollow">Pre-Registered Hypotheses</a></li>
1. Construct daily prevalence time series (cross-sectional).
2. Construct daily prevalence time series (longitudinal).
3. Construct daily add-event time series (inferred).
4. Construct daily delete-event time series (inferred). </p>