Five Upcoming Academic Books on Data Work, Big Tech, and Misinformation
Our most hotly anticipated books for fall!
Posted on in Blog Posts
Posted on January 6, 2025 in Blog Posts
Authors:
Robin Camille Davis
So you want to hear from real people who use your library! There are many user research methods to choose from, such as surveys, usability testing, and interviews. Your research goals will dictate which method you use.
In a nutshell: interviews let you get at the why and the how, not just the what.
For example, you could easily use surveys to ask a large number of your users which library services they use. But only an interview can tell you why your patrons seem to use virtual references services often but never the in-person information desk. Moreover, you can ask interviewees follow-up questions to dig into their individual answers. For instance, “I notice you said you don’t want to ‘bother’ the librarian at the information desk. What kinds of things give the impression that they would be ‘bothered’ by a question?”
Opportunities afforded by conducting interviews: they enable you to…
Challenges of conducting interviews: you will need to…
🔥 Stay up-to-date with LibTech Insights by signing up for our free newsletter. Just one weekly email with our new blog posts, top tech news stories, and other bonus content. Check out some posts from our archive:
Will a 5-minute interview suffice? Get your hands on some pastries, stickers, or other little goodies and set up a table in the lobby. Call it a pop-up interview event and take advantage of a convenience sample. On a busy day, you could easily conduct 10–20 interviews.
Do you need more time for longer interviews? Recruit interviewees ahead of time. You could use posters, social media, and/or targeted emails to find interviewees. Consider using a screening survey if your interview will focus on a specific topic. For instance, if you’re interested in learning how patrons use the children’s section, your screening survey could ask, “How often do you browse or borrow books from the children’s section? Never/Sometimes/Often.”
For longer interviews, you could hold these in person (best for building rapport) or remotely using video conferencing software (most convenient for your interviewee, probably).
How many interviewees should you recruit for longer interview sessions? It depends on how much time and how many resources you have! More interviews mean more data, but that also means more time spent scheduling the interviews, conducting them, transcribing them, and analyzing the data. If this is your first interview study, aim to recruit 7 people, and assume that 1 or 2 may cancel or be no-shows. A set of 5–7 interviews is a great place to start.
With your colleagues, you’ll figure out what exactly you want to know. Aim for 3–5 research questions that will likely result in something actionable. For example:
Note that these are questions that will probably have complex answers. Simple questions, like “Do our users like our remodeled lounge?”, could be the basis for an interview, but if that’s all you want to know, stick to a survey. On the other hand, if you expect that a simple question could result in interesting follow-up questions on your part, go ahead and include them in your list.
Now for the tough part. An interviewer should almost never pose a research question to an interviewee. Instead, turn a research question into an interview question. Here are some examples that you can adapt to your study:
For instance, if your research question is “How could we improve our event programming?,” you could start with “Tell me about the last time you attended an event at the library,” and include some follow-up questions like, “How did you find out about the event?” and “If we held that event again, what would you suggest we do to make it better?”
The Nielsen Norman Group has a great guide to compiling an interview guide (script) with plenty more examples.
I strongly recommend doing a dry run of your interview questions with a colleague or friend. (Bribe them with coffee or snacks!) Doing a practice round can reveal whether the interview script flows well and whether any questions are worded confusingly.
Transcribe the audio files. You can do this all by hand if you want to—I find that while it’s more time-consuming, you really get to know your data well. You can also save time by using AI tools to do a first pass. I’ve been known to play the audio file on my phone while in a Zoom call by myself, turning on the automated captions and saving the resulting file. Note that you will definitely have to look over the transcript. Even the best AI tools will make ridiculous errors. Plus, you’ll have to mark whether the interviewee or interviewer is talking. Don’t forget to remove any personally identifying data from the transcripts. Call participants “Participant 1” or some other anonymizing nickname.
Read every transcript, beginning to end. You will absolutely benefit by familiarizing yourself with the data (transcripts) and reading them back-to-back.
Analyze the data. Find themes and commonalities. It’s up to you whether you want to do a lightweight analysis or a really thorough one. Here’s an in-depth way to do a comprehensive analysis from the Nielsen Norman Group.
Write a report for your colleagues. The report should include a description of how the study was conducted, themes and findings from the interviews, and resulting recommendations for actions to take. Include any especially great quotes from your users—these can make a big impact!
—
Interviewing is truly a skill. Expect to stumble a bit or hit a few snags if it’s your first time. But remember, it’s just a conversation! As you continue to conduct interviews, you will build your interviewing skills and gain confidence in this versatile, powerful user research method.
Our most hotly anticipated books for fall!
Posted on in Blog Posts
What if work could feel remoralizing?
Posted on in Blog Posts
Insights and best practices for teaching AI literacy to history students
Posted on in Blog Posts
What our micro-course participants had to say about AI in libraries
Posted on in Blog Posts