Last semester, I interviewed Albert Lee ‘24 to get a glimpse of what conducting qualitative research for sociology Junior Papers can look like at Princeton. (If you’re interested in reading that piece, click here!). Discussing qualitative research with Albert was exciting because his words of advice were quite applicable to the qualitative Computer Science research I was conducting in COS 436: Human-Computer Interaction (HCI).
An example user prompt and ChatGPT response.
Prior to taking COS 436, I had little idea of what qualitative research looked like in Computer Science, particularly because many of the CS courses I had taken were quantitative, involving systems, mathematical models, or theory. Taking the course opened my eyes to a whole side of research: interview-based qualitative research. For my semester-long research project, my team and I aimed to dive deeper into educators’ opinions on the use of ChatGPT in CS education.
Though I learned many research skills during this project, a challenge both Albert and I experienced was recruiting interview participants. Because I was working on a semester-long project, the recruitment process for interviewees needed to happen fast (in about two weeks)! However, we quickly found that people tended to be more willing to fill out a 2-minute screening questionnaire than agree to dedicate 30-45 minutes to a face-to-face interview. Given these factors, our advisers said aiming to secure six to eight interviewees was reasonable—in just three weeks, we were able to not only recruit eleven interested interviewees, but successfully completed seven interviews.
An important note before we continue: at this stage, my team’s research project already received Institute of Review Board (IRB) approval. For all students working on research, please work closely with your professor, OUR’s Survey Research Center, and/or the Princeton Research Integrity & Assurance (RIA) to ensure that research and recruitment is conducted ethically. Moreover, start preparing for these approval processes early, as it often takes a number of weeks. Students conducting research, for example, may need IRB approval before they may conduct any qualitative research.
Without further ado, here are my key takeaways and advice specifically on interview recruitment:
1. Start working on your recruitment materials early.
While this may seem self-explanatory, there are several parts that go into conducting interviews. Though recommendations about “what works” may vary by project and field, our graduate TA adviser suggested we draft both social media posts and emails. We also needed to prepare our screening questionnaire, which we used to gauge general information about our prospective interviewees and ask if they would be willing to participate in a follow-up interview. We iterated through several versions of this questionnaire to ensure that our questions (such as, “What is your outlook on the student use of ChatGPT in Computer Science education?”) weren’t leading. Though we did not use these responses in our final research paper, these responses allowed us to select a wider variety of educators (by geographical location, self-reported experience with ChatGPT, and whether they were high school teachers, graduate student TAs, lecturers, or University professors). If possible, review your recruitment materials with an adviser.
2. Be careful when recruiting on social media.
Some advice we received was to try posting recruiting messages to social media platforms like Twitter/X and Mastodon (we posted to hci.social). While many people were interested in our project and were curious about the findings of our project, unfortunately we did not receive any questionnaire respondents through either of these platforms. This was possibly in part because our study was limited to educators in the United States, but this seemed to not be the most effective study participant recruitment method.
A note of caution: I would highly recommend scanning through questionnaire responses as they come in; especially when we posted on Twitter, we received several bot-like responses that were suspicious due to strange email addresses or incorrect basic information (such as writing that they were an educator at Princeton University, and then marking the institution as a high school). However, since each server or social media platform may differ depending on the number of people and the culture of those platforms, I would still recommend trying to recruit via social media.
Our team’s hci.social Mastodon post
3. Cold emailing can work wonders!
Taking into consideration who your intended interview participants are can help you adapt your recruitment approach. One helpful aspect of using Mastodon was that we were able to identify CS and HCI educators and send them a thoughtful, personalized email instead. Though they may not have seen or interacted with our recruitment post, we were able to meaningfully connect with educators via email. We also looked at websites of CS, HCI, and Natural Language Processing research labs at universities to broaden our recruitment reach. After sending around 40-50 recruitment emails, we received 14 questionnaire responses and 11 educators who were interested in conducting interviews. Looking back on my experience, I think it would have been incredibly helpful to have started cold emailing earlier, and would highly recommend this!
If you’re not sure how to cold email someone, check out this PCUR post!
4. Follow-up with participants after they express interest.
Getting people who are interested in participating in your research study is only half of the recruitment process. To help secure interview participants, send a follow-up email asking them to schedule an interview date and time. It’s also a great opportunity to thank your prospective interviewees for their time and their interest in your research. It can be quite convenient to use a website like Calendly to help you organize your availability and schedule interviews (especially if you have several to conduct)!
—
I hope these tips were helpful, especially for those of you who might not have experience with qualitative research. You get to develop so many wonderful skills while conducting research, and interview recruiting may be one that is more overlooked than other skills such as generating research questions, conducting interviews, and running data analysis. However, it’s still a crucial part of the research process! If you have further advice, feel free to email me at alexiswu@princeton.edu for a chance to be featured in a future PCUR post. Best of luck with your qualitative research!
— Alexis Wu ‘25, Engineering Correspondent
