General: The survey was conducted in Qualtrics, and it consisted of ~60 questions (some questions were contingent on participants' response to an earlier question). All questions were optional, and they consisted of a mix of multiple choice and open-ended questions. Participants could complete the sections in any order they desired, and they had two weeks to complete the survey, after which their completed questions would be automatically submitted. The survey was completely anonymous; not even the IP address was visible. Participants could give out their e-mail address if they wished to be contacted a follow-up interview.

Participants and recruitment: Our goal was to reach as many people as we could who in the past or currently worked on language research, whether or not they self-identified as being a "linguist". Because language research happens within and outside of the academy, and within and outside of linguistics departments, it was important to us that as many experiences be included as possible. In addition, experience of negative climate and harassment can influence who stays in academia, and so we did not limit ourselves to those currently in the academy. Finally, academia is global, and we did not want to limit our survey just to those in the United States, though that is where our study team members are based, and our terms and perspectives were US-based.

The survey was distributed via various listservs, LinguistList, Twitter and Facebook, the personal and professional networks of the survey team, and a link in a newsletter sent to members of the Linguistic Society of America.

The listservs included those associated with the CUNY sentence processing conference, IWLP (the International Workshop on Language Production), CogSci (the Cognitive Science Society), CEDL and COSWL (The Committee for Ethnic Diversity in Linguistics and the Committee on the Status of Women in Linguistcs, both organizations associated with the Linguistics Society of America), LingAnth (Linguistic Anthropologists), HISON (Historical Sociolinguistics Network), FunkNet (), The Women in Generative Linguistics Facebook Group, African Diaspora Listserv, ISGS (the International Society for Gesture Studies), SLLS (the Sign Language Linguistics Society), and more.

Several professional societies were contacted in an attempt to reach a broader and more international audience, but we received no response or action from them.

As such, our participants skew young, white, female and are predominantly from North America/Europe, and were more likely to be affected by harassment and bias incidents. It is likely that more people who were not affected by harassment and bias incidents opted to not take the survey than those who were affected. Thus the sampling of participants is not entirely random, nor necessarily representative of the fields of linguistics and language sciences.

Follow-up interviews: We received 67 responses for follow-up interviews. The participants had the option remaining anonymous or using a name (real or assumed), choosing any of the survey team members to be the initial point of contact and to conduct the interview, choosing the medium of communication for contact and interview (over e-mail, phone, Skype/Zoom, other videoconference programs, face-to-face).

We followed a scripted template for initiating contact and scheduling interviews, and adopted a semi-structured interview model. This model allowed us to introduce ourselves, explain our motives for the survey, and explain how we are maintaining their anonymity. We asked the participants to share a few general things about themselves so they could situate themselves and whether they had a story to share.

To date, we have conducted 25 follow-up interviews. We initiated contact with most of the remaining participants, but they did not reply. We did not record the interviews, but rather we summarized them in writing for analysis and reviewed the summaries to ensure that they did not divulge potentially revealing information that would risk exposing affected individuals.