Interviews Are More Than Conversations

Getting started with interviewing can be stress-inducing and overwhelming. It’s not too surprising then that one of the common pieces of advice for newbies to UX/design research is to “stop doing interviews, start having conversations,” “just start talking with people,” or stop worrying because “it’s just a conversation.”

You’ll find similar advice all over the internet that aims to help people feel more comfortable and be more successful in job interviews. That’s understandable — job interviews can be awkward and stressful for everyone involved, so de-stressing the process can make it all less painful.

But when you’re conducting research, it’s really not that simple. An interview is not a conversation, and a conversation is a poor substitute for an interview.

Conversations and interviews are distinct modes of interaction, with distinct purposes and (if they’re done well) distinct outcomes.

When it comes from experienced researchers (like Nicola Rushton, whose otherwise solid article I’ve linked to above), this advice to make your interview a conversation is usually followed up with really sound guidance on how to effectively run a research interview. What people seem to be trying to say is “make interviews feel more conversational,” as a response to an approach to research that calls for very standardised, formulaic “interviews” in pursuit of rigour. I am absolutely on board with making human research less cold and clinical, for lots of reasons.

But there’s a very important difference between “having conversations” and “making interviews feel more conversational.” Conversations and interviews are distinct modes of interaction, with distinct purposes and (if they’re done well) distinct outcomes.

Making “interview” interchangeable with “conversation” is a bit like conflating “running a marathon” with “going for a walk.” Pretty much everyone can “go for a walk” (with apologies for the ableist language in this metaphor). But to run a marathon safely and successfully requires extensive preparation and practice toward developing a capability that goes well beyond what the ordinary person can achieve by default.

If you actually have a conversation when you should have been conducting an interview, the best result is that you’re going to have a nice time with someone and maybe feel like you’ve learned something about them. If you’ve had a conversation when you needed to have an interview, you’re unlikely to get the outcomes you needed, and you’re going to be wasting your time, your team or client’s time, and your participant’s time.

In the worst cases, you’re going to perpetuate bad “research” practices which makes it harder for your team and/or clients to recognise good research when it happens, and you’re going to contribute to really bad outcomes for people and the planet.

In order to do — and recognise — good research, it’s important to know the differences between these modes of interaction and to know when to use them.

So, what’s a conversation?

Conversations and interviews are both modes of communication that involve the transfer of information between participants. But there are fundamental differences in the way information moves between participants in these two modes of communication.

Conversations are ubiquitous — they’re probably the most common mode of verbal communication between two or more people. We all have them multiple times during a day. But how does a “conversation” work?

One of the best explanations I’ve seen of how conversations do (and don’t) work is from Ursula K Leguin’s essay “Telling is Listening,” from her non-fiction book, The wave in the mind.

Leguin writes that the dominant model of communication in the age of technology is mechanical. Two (or more) participants in a conversation transfer bits of information between one another, so that at the end of the exchange, bits of information from box A are transferred whole into box B, and vice versa.

Imagine, if you will, a sketch by Ursula LeGuin illustrating the 'mechanical model' of conversation with two boxes transferring information between their unaltered selves.

By alternating speech acts and switching roles from speaker to listener, A and B transmit, receive, and respond to bits of information. The information itself is objective and concrete, needs no (or little) interpretation, and does not get transformed or corrupted in the process. If that’s accurate, there’s no problem treating interviews like conversations — and in fact, any mode of communication would be fully interchangeable in terms of information transfer.

In fact, conversations are intersubjective experiences, where the information flowing between parties is actively transformed throughout transmission, interpretation, processing, responding. And not only is the information held and communicated by each participant transformed, the consciousness and constitution of the participants themselves can be transformed by the act of engaging in a conversation.

Leguin’s metaphor for this conceptual model of conversation is amoebas having sex.

Imagine, here, a different sketch by Ursula LeGuin, this time illustrating her own concept of how conversation works - like 'amoebas having sex,' in which partners physically connect and transform one another by exchanging bits of themselves.

While they mainly reproduce asexually, when they do reproduce sexually, they “reach out to each other a meld their pseudopodia into a little tube or channel connecting them. … Then amoeba A and amoeba B exchange genetic ‘information,’ that is, they literally given each other inner bits of their bodies, via a channel or bridge which is made of outer bits of their bodies. They hang out for quite a while sending bits of themselves back and forth, mutually responding each to the other.”

Like amoebas having sex, humans having conversation “unite themselves and give each other parts of themselves” in communities of from two to many members.

You can recognise that a conversation is happening when participants share bits of themselves, whether to establish common ground — “oh wow, that exact thing happened to me last week!” — or to establish status through one-upmanship — “yeah, I know what you mean, but … (listen to my more impressive thing) … so, yeah.”

You can recognise a really great conversation by the way it feels — effortless, timeless, buzzing, warm, exhilarating. Those are reactions to the bits of themselves being shared among participants building on one another and turning into a kind of magic.

This is the thing that can make conversations feel warm and fuzzy and conversational, and it is exactly the thing we need to avoid when we need to conduct an interview.

And what is an interview?

An “interview” is a mode of communication where the purpose is to collect information from one or more people, in order to understand their experience, thinking, principles, and/or whatever sits underneath and shapes their thinking and experience.

In popular consciousness, “interviews” are fairly obvious — it’s one person asking another person questions and that other person responding. It’s a dry, sometimes combative, form of call and response. We know them from pop culture representations of what journalists, detectives, lawyers, and even researchers do.

Whether the research question calls for a highly structured interview or a completely open-ended one, the challenge for an interviewer is to keep their own experiences, expectations, and ego out of the interaction.

In research, there are several different types of interview — from strictly structured to something more like free jazz — that are used in (and favoured by) different disciplines. The different types of interview can all be useful for specific purposes, and misleading when used inappropriately. Whether the research question calls for a highly structured interview or a completely open-ended one, the challenge for an interviewer is to keep their own experiences, expectations, and ego out of the interaction.

As an interviewer, your job is to elicit from your participant(s) the kind of information that will help to answer the questions you and your team need answered. This is impossible if you treat an interview like a conversation and insert yourself into the interaction. As an interviewer, you are an instrument, not a participant in a conversation. You must adopt the role of “interviewer” and do your best to put your subjective identity aside for the duration of the interaction.

Just as you need to avoid injecting bits of yourself into an interview with stories and statements about yourself, it’s important to keep your interpretations and emotional reactions to yourself as well. You need to practice a form of deep listening in which you don’t interrupt, you don’t push back. You notice the responses that occur when you encounter bits of your participant and you let those responses go, because the interview is not about you.

In that sense, a researcher is similar to a psychologist — their job is to listen and to guide a client to reveal deep, meaningful bits of themselves in order to effect change. If your therapist spends time talking about themselves, you’re having a conversation, and you’re not likely to get where you need to be.

What about rapport and ethical responsibility?

One of the things people who say “interviews are conversations” seem to be trying to highlight is that interviewees need to feel comfortable and safe and to trust the person they are speaking to; likewise, interviewers need to relax and feel less nervous about talking to strangers.

Given that conversation is essentially reciprocal — bits of participants being shared among them — it can feel yucky to think that as an interviewer your role is to collect bits of other people without giving back. Taking without paying back can be feel exploitative (and of course there are numerous examples of researchers actually exploiting participants… see anthropology’s colonial roots and their lasting legacy in the discipline for far too many examples). We also know that the stereotypical cold, clinical style of interviewing can be off-putting and yield very underwhelming results.

To get the best result out of an interview, it is critical to establish rapport early and maintain it throughout the session. You need to make your participant feel safe and help them get comfortable opening up and sharing those bits of themselves that we want to collect. When a participant is uncomfortable or feels unsafe, they’re more likely to provide partial truths or to shut down altogether, which makes it impossible to learn what you set out to understand.

These dynamics, and the inherent power imbalances of research interviews are why we also have an absolute obligation to protect the well-being of research participants; this is (one of the reasons) why informed consent is non-negotiable, and why we have to be constantly attentive to how a participant is feeling so that we can help mitigate any harm that might occur.

So, while making an interview feel easy, and feel like general conversation, is an important ingredient in the success of interview research, we still have to avoid the trap of mistaking “conversation” for an interview. Instead, we need to get really good at using a range of specific techniques to help build rapport and trust, such as the ones Indi Young describes in her book, Practical Empathy, all underpinned by the need to “demonstrate warmth, human support, and clarity of direction to the participant.” (For a handy cheat sheet of useful probing phrases, check out Steve Krug’s “Things a therapist would say (pdf).”)

When do I need an interview, and when can I have a conversation?

We know that confusing an interview for a conversation is likely to lead to bad outcomes — at the very least wasted time and misleading results. But sometimes the situation doesn’t require an interview.

The key is to know the purpose of the interaction.

Use conversation mode when:

You just want to have some fun or when you want to avoid confrontation or offense. Or when you want to establish a personal or professional connection based on shared experiences, opinions, perspectives, beliefs, and goals but don’t need to critically evaluate another person’s understanding of those.

Switch into interview mode when:

You need to deeply understand someone else’s experiences, perspective, thinking, or the forces and factors that shape them. Or when you want to get less biased reactions and feedback on your own ideas (e.g., in the form of concepts, prototypes, products, plans, etc.).

But even if conversation mode is an acceptable form of communication for you to achieve your goal, consider when and how to bring elements of interview mode into the mix. If you need to really understand someone else, especially if their experience or perspective is different from yours, make sure to carefully check your assumptions and biases, practice active listening, be compassionate and empathetic, and be humble.

As long as you know when it’s okay to default to conversation mode, and when you need to switch over into a more intentional, interview-like mode of communication, you’ll be more able to avoid misinterpretation and false consensus with the people you’re trying to live and work with.