When interviewing participants in a user research study, researchers have to be mindful of how feedback is treated at all stages of the interview process. There are visible and invisible pitfalls that a researcher can succumb to when gathering input from participants, which can cause ‘bad data,’ as a result.
Here are some top tips to consider when conducting your next research study, so you can make sure your data is effective. This is by no means an exhaustive list, but a good starting place to keep in mind.
Tip 1: Determine the correct methodology and mode of research for your needs
The first step is to formulate a specific hypothesis of why you are needing to conduct research in the first place. Then, ask yourself if there are any ethical ramifications to consider before moving forward. For example, if you want to learn more from individuals of a vulnerable population, refer to subject matter experts in the related industry to ensure that there are proper barriers in place to help protect the individual during the interview. Barriers could take the form of the specific type of methodology you use or how you decide to administer the interview. By being mindful of these considerations, you will be able to gather more valuable information and prevent potentially ‘bad data.’
Tip 2: Properly screen your participants
Ask yourself if you are recruiting the right kind of participants for your study, based on the type of interview you are conducting. Most researchers conduct a ‘Screener’ to help weed out the non-qualified people from joining a research study. This can take the form of a quick survey or phone call to quickly gather demographic or related information from the interested person. In organizations that have a more robust metrics & analytics platform, researchers try to get that information ahead of time, and only include unanswered questions in the screener. This is the best case scenario to help reduce drop-off because potential participants would have less questions to answer to move forward.
Tip 3: Do a pilot test
Do a test run with a colleague before speaking with participants to make sure you work out all of the kinks. This will help make sure you have enough time planned to be able to fit in all of your questions. Also, this is the perfect time to take note of potential technical problems and devise a backup plan, in case it arises during an actual interview.
Tip 4: Pace yourself
Plan out interviews with enough time in between, so you can properly debrief and prepare for the next session. Also, make sure to not exhaust yourself, so you’re able to have the stamina to give your full attention during all of the user interview sessions. This all depends on what you have the bandwidth for, but a lot of researchers tend to make a rule of limiting interviews to a certain number per day. If you’re too wiped out to give your full attention, then you might miss key insights, which could skew your results and give ‘bad data.’
Tip 5: Take note of existing biases from the participant
Biases are understandable and hard to avoid. Not only do you need to take into account certain biases from the participant, but also, biases within yourself as a researcher. One way to reduce biases that could cause ‘bad data’ is to ask non-leading questions. Sometimes participants just want to be liked or feel like they are giving a good answer. If a researcher gives too much away with a question on how they might like to hear an answer, participants pick up on this. It’s understandable when it occurs, but it skews the data to benefit no one.
Tip 6: Record the Session
When researchers don’t record a session, more memorable insights tend to be reported. Also, it’s easier to recall insights that support your original beliefs than those insights that are against it, so having a recorded session to refer to is essential to preventing ‘bad data’ from perpetuating when analyzing results.
If you keep these tips in mind when interviewing your next batch of users, then you will be more likely to catch higher quality feedback with a participant and reduce the risk of skewed or ‘bad’ data. Overall, these insights will ultimately help guide you and your team on what to do next when creating meaningful and impactful software. For further tips, regularly check this website for new posts from people who care about user research and design.
Amanda is currently a Senior Product Designer with Ncontracts. Her craft is mainly geared towards building and refining enterprise SaaS systems, but also has a passion for streamlining processes for tech startups. Currently residing in Raleigh, NC, she is a keen traveler. Think of her as Carmen San Diego but responsive to email. However, you’ll most likely find her searching for an outlet to plug in her computer.