Acquire information using interviews

Information acquisition directly from others most commonly takes the form of an interview. Interview makes the process sound fairly formal, but it need not be. Keep discussions with others simple, conversational, and relaxed. But, to ensure you're not wasting your time or others':

• prepare for conversation by thinking about goals and having and approach to recording information
• perform the interview keeping some basic rules in mind about acquiring information
• consolidate and distill the information so that you can leverage it during design.

Prepare

Preparation can take the form of pulling interviewers together for a meeting to discuss goals and process to simply pausing for 30 seconds before a discussion to clarify in your own mind the goals you have and the approach you intend to take. Prepare in a way that best fits your project context, but prepare.

When preparing think of these things.

Set Goals

When you sit down to have a conversation with someone, have a goal in mind. Quickly ask yourself these questions:

1. What information will support your design efforts?
2. What topics of discussion will best get to that information?
3. What shouldn't I ask about? What would details be best observed or gathered another way?

Write the goals in your notes. Pay attention to them as you're interviewing. When you feel you've met a goal, check it off. Try to meet all your goals before the interview is over.

Keep Notes

Be ready to record notes from your interview. A notebook, deck of index cards, or a pad of post-its work well. Placing one idea per notebook line, index card, or post-it will allow you to move more quickly into a modeling phase to distill the key points from your notes.

Perform

Perform the interview in a conference room, at an interviewee's desk or with them where they perform their work, or at the water cooler during the day. Interviews can be formal or informal.

Ask Questions

You may already be familiar with two common types of questions: open and closed questions. Become familiar with both types. Know which you're asking and why.

Open Questions

Open questions initiate or continue a conversation.

Use open questions to begin conversation on a new topic, or keep conversation moving on a topic.

• Can you tell about the last time you sent an email message?
• What do you like about your current software?

Avoid asking open questions that are too general. You may confuse the interviewee or set the conversation on a wide trajectory that may be difficult to recover from.

Closed Questions

Closed questions obtain a single piece of information.

Use closed questions to obtain simple facts. Also use closed questions to slow down or change the course of long running discussions.

• What applications do you use to send email?
• Have you used this application in the last week?

Avoid using closed questions to ask for opinions. These are best handled as open ended questions. Avoid asking rapid-fire lists of closed questions. These can make interviewees feel as if they're being interrogated.

Ask Interviewee to Recall a Specific Event

When you're looking for information on a particular activity that your interviewee has experience in, ask them to recall the last time they participated in that activity. Start with questions that help them imagine where and when that event occurred. Proceed with questions that ask about specific things they did during that activity. You'll find this approach helps you get more and higher quality information.

What you're avoiding by doing this is asking your interviewee to generalize about what happens during an activity. Generalization often results in the interviewee describing what should happen rather than what did happen. Generalization often means you'll likely lose important details.

Refer to Props or Artifacts

Often during an interview the interviewee will refer to something, a form they filled out, an email message they received, a piece of software they used, a report they read… Ask to see these things. Bring these artifacts into the discussion. Refer directly to them in discussion. If the exact artifact isn't available, look for something to use as a stand-in or prop. The act of holding and referring to a physical object often allows the interviewee to recall things they wouldn't have otherwise. Seeing the physical object may cause you to ask questions and obtain information you wouldn't have otherwise.

Avoid Interview Objections

This may date me, but I've watch quite a few episodes of Perry Mason the lawyer who always solved crime usually by finding information no one else could and often by getting the subject to admit to the crime on the witness stand.

Imagine yourself as Perry Mason, or if you aren't familiar with Perry, picture in your mind another lawyer who's smart, friendly, and very effective at asking questions and listening. Now picture yourself as that lawyer, we'll call him Perry, interviewing a friendly witness, someone who you like, respect, and that has valuable information that can help your case. To keep the picture in our head clear, let's think of her as Perry's trusted assistant Della. However, there are some rules to follow that you shouldn't break. You don't want the other lawyer, let's call him Mr. Burger, to object or the judge to declare a mistrial, so you learn and follow the rules.

The rules for a good interviewer are much the same as a good lawyer's. To understand the rules, let's look at the typical objections we're likely to encounter in a courtroom drama. Picture that nasty opposing attorney, Mr. Burger, shouting these objections to the judge if you make a mistake during your interview with your trusted friend Della. That's what I do, and it helps. Let's lump some common objections into three categories: 1. biasing 2. confusing and antagonizing 3. assumptions and guesses

Biasing

You can inadvertently or on purpose steer the interviewee to give the answers you want. If you really want specific answers, just write them down and avoid the interview all together. Don't waste time. But, I know you really want the information, so watch out for these common objections during an interview.

Confusing and Antagonizing

Remember your interview subject is Della, you're trusted friend. It doesn't help her, you, or the jury to confuse her. Avoid questions that might do so.

Assumptions and Guesses

If we're hoping to gather information to help us better design, we need to be aware of the quality of that information. Sometimes the information we gather can be considered fact - or at least if it's a first hand observation, it's the observer's version of a fact. But, sometimes you may think you're capturing a fact, but actually you're recording an assumption or guess. While capturing assumptions and guesses aren't always a bad thing, recording them as fact can be. If you're trying to capture fact, avoid questions that result in assumptions and guesses.

Pair interviewers

It's useful for two interviewers to work together. This will help the interview move a little quicker making better use of your interviewees time. You're likely to capture more information than an individual interviewer would. You'll also have the benefit of evaluating the interview notes together which generally results in much higher quality information.

Follow the general approach to pairing.

The driver will primarily ask the question.

The navigator will primarily record notes.

Passing control happens when the interviewer asking questions runs aground and can't think of a next good question. In this case the navigator, the one writing notes may ask a next question. Then control has been passed and the other interviewer assumes the role of navigator and begins writing notes.

At times the navigator may notice that the questions are pulling the interview far from the original goals. In this situation the navigator may also choose ask a question that pulls the conversation back towards the original goals of the interview.

At the end of the interview paired interviewers to need to reserve time to compare and consolidate notes. In discussion they may alter or clarify notes, or write down information that wasn't recorded by either interviewer.

Interviewers should reflect on their approach to pairing with the intent of possibly changing their approach in the next interview.

Modeling in an interview

It's often effective to work collaboratively with the interviewee to construct a model that represents both your understandings. Keep post-it notes, index cards, or paper handy. Or use a close by whiteboard. But if you do, don't forget to photograph the results to take back.

An ad hoc model can be helpful for making sure interviewer and interviewee understanding each other. You'll find that if you construct a model you can easily refer back to it during the discussion to help set context. For example if I'm asking you to tell me about your morning email routine we might use post-it notes to construct a simple workflow model. It's easy for me to point back to a post-it note and say "back here at the time when you look for urgent messages, what do you do with the urgent messages you identify?" Pointing and saying "back here" immediately allows the interviewee to reestablish context to a different time in their process.

When modeling follow general approaches to modeling.

Also, keep in mind that the models are used for ideas or information you wish to share and collaborate over. This may not be true of the observations in your notes. You get to decide what information to put into the model, and what to leave in your notes. Sometimes putting observations out on the table in a model causes discussion that you may not want to occur during the interview.

Parking Lots and Feed Forward Bins

At times during the interview process ideas or information come up that while valuable may not be important to the goals at hand. Create a parking lot or use named feed forward bins to keep this information.

This is a helpful technique to use when an interviewee has a particular subject that's important to or interesting to them, but isn't information you necessarily need just now. If you're the interviewer consider saying: "This is important, but there's a few more questions I want to get to first. Let me write this down here so I don't forget it, and we can come back to it later."

Write the information in a parking lot. Come back to it later if time permits.

Apprenticeship style interview

When trying to better understand how someone does their work it helps to take on the role of an apprentice. Ask the person you're working with to treat you as a new employee that needs to learn their job. Their goal is to teach you. You're goal is to ask enough questions that you understand the work.

Carry out an apprenticeship interview in the context where your interviewee does her work. Work with the same software, handle the same forms, listen in on the same phone calls.

When asking your interviewee to step through their work, ask them to start with a tangible example. For example, if ask them to pick up and complete work they were occupied with just before you came. Or, to think back to work they'd done previously that same day.

Snap a couple photos while you're there to commemorate the moment, and to get a pictures of your interviewee's workspace, and physical artifacts they handle. It's all potentially valuable information.

This approach can take some time, but you'll get a wealth of real data, and a strong appreciation and empathy for the person you may be building software to support.

Focus Group Interview

Interviews are best conducted with one interviewee. When you have multiple interviewees in the room, the interview starts to take on the dynamic of a focus group.

A focus group gathers a group of people together to obtain information or feedback on a particular subject or product, typically in a structure way. While a focus group may result in a lot of information, placing people in a group often results in unexpected and inappropriate side affects that bias or skew your information. For example some group members may uncomfortable discussing some subjects with each other. Often groups begin to fall into a group-think rut where they all tend to agree with each other. Behaviors like these can interfere with collecting information.

However, in some circumstances a focus group style interview with two or more people can be valuable. Sometimes the discussion between group members can yield valuable information. If you have a similar questions to ask each group member, watching one member answer questions allows another member to begin constructing answers in their mind to have ready whey they're asked. This can result in more quickly collecting information. But, the interviewers must always consider how the act of obtaining the information in a group setting may have affected the outcome. Interviews should consider this when evaluating interview results.

Consolidate

At the end of an interview, set aside time to consolidate your interview notes and observations.

• Make sure your notes are legible, and that you wrote what you meant.
• Look for anything you might recall but forget to write down in your notes.
• Consider transcribing your notes into an electronic format if you'll need to share them or merge them with other's notes.
• If you've performed the interview as a pair consolidate notes.
• For consolidating notes or simply distilling themes from your notes, consider constructing and affinity diagram from your notes. Consider each note a separate idea to model in your affinity diagram. A collaborative work session works well to help speed up the distillation process across multiple interviews.

Evaluation is an important part of interviewing. Even if you only have a few moments, take that time to look over and correct your notes. If you have the opportunity to plan ahead, plan a small block of time to evaluate and distill your notes to identify important themes and ideas in them.

Ideally evaluate your notes pairing. Read your notes out loud to your pair. You'll catch errors or issues as you read out loud. Your pair will ask questions that may alert you to questions you might ask next time or questions you did in fact ask but neglected to write complete notes on.

If the interview you've just conducted is part of a battery of interviews, use a collaborative data consolidation worksession to consolidate notes across interviews.