Research Synthesis

A Mentor Tutorial

Image for post
Image for post
Mentor Designers Engaged in a Research Synthesis Activity

This tutorial breaks down the typical qualitative research clustering activity used by Mentor Creative Group. In this article we will explain some of the hows and whys around our process for running this important design activity.

Before we get into specifics, it’s important to recognize that there are many different methods for qualitative research synthesis. This document outlines a clustering method similar to an affinity analysis. Regardless of your data source(s), the goal for any synthesis activity is to understand what patterns exist in your data and produce actionable next steps for designing solutions. The examples we use in this tutorial come directly from Erika Hall’s seminal text, Just Enough Research.

Who should be involved?

Anyone who is invested in the work to come should be considered a potential participant in this activity. However, moderating a synthesis activity will get exponentially harder the more folks that are involved. Be mindful of this when organizing your event and try to invite people who will have unique perspectives. You don’t have to be a user researcher to participate and it’s actually counter intuitive to only include your immediate research team. Here’s an example list of some people might want to consider:

  • Stakeholders (C level or otherwise)

Setting Some Ground Rules

Now that we’ve covered who should be included, let’s take a look at some participation ground rules. These guidelines should be communicated prior to the activity, usually as part of an event invite AND at the beginning of your session.

Save solutions until the end.

The primary goal of this activity is to pool our research observations in order to identify problems that need solving. That being said, we are not trying to solve problems during this activity. If someone starts talking about a solution, then redirect the conversation in a way that helps identify the underlying problem their solution would solve.

Prioritize action over discussion.

There’s a lot to cover in this session and discussing problems as a team is definitely part of the exercise. However, don’t spend 10+ minutes talking about any one thing. If you feel it’s taking a while to define a problem, stop where you are and come back to the topic after working through some more research data. That way you’ll return with a broader perspective and find the task much easier.

Join at the beginning and stay as long as you can.

These sessions tend to last multiple hours which we understand is a tough commitment. That being said, those who want to be involved should be able to commit at least 30 minutes to the beginning of the session for an initial overview. If folks need to dip in and out of the activity, that’s perfectly fine. However, do not take calls or have sidebar conversations not related to the task at hand. If you need to step out, please do so then come back when you are ready to continue collaborating.

As the moderator, go into this synthesis activity with the mindset that the rules will be broken and it’s your job to steer the conversation back to a positive place. We find it helps to have your guidelines written down and posted in the activity space so that participants have a visual reminder.

Activity Steps

Now let’s take a look at the clustering method itself. While the steps are written in a linear fashion, a synthesis activity will move back and forth between constructing and deconstructing your data. This is a good thing because it helps people question their predispositions and really think about what the data is saying, which is not immediately apparent. This is where diversity in your participants can create a shared source of empathy for future phases of work. Team composition aside, your goal is to end this activity with a series of data clusters.

Data clusters are comprised of the following elements:

  1. Research Quotes / Observations: The things you heard or observed during your research sessions.
Image for post
Image for post
Structured Data Cluster

Making Data Clusters

In the beginning, it’s all about getting quotes and observations written down then posted on a whiteboard. From there, your team can start to make sense of the data. Here’s a general outline:

  1. Closely review your notes, transcriptions, recordings etc.

Coding your sticky notes can be useful if you want to quickly see the origin of your data. It’s not necessary, but can help the team better understand trends as they emerge.

For sake of example, let’s say your team is working on a local community project. The community leaders want to explore ways to increase neighborhood participation for public events like fairs, art walks, and outdoor music concerts. After preparing a participant screener, you conduct 1:1 interviews with eight different community members. Using steps 1–4 above, you’re activity could start out with a couple post-it quotes:

Image for post
Image for post
Initial Observations & Quotes

Labeling Groups with Insights

Continue steps 1–4 for as long as is needed, but as soon as patterns start to emerge your team should label them immediately. These labels will change as new insights are added, so don’t spend a lot of time wordsmithing during the activity. Again, this method is about constructing and deconstructing simultaneously. Most of your original groups will be broad to begin with anyways and should be broken down to the smallest actionable insight by the end of your session.

  1. Watch for emerging patterns and label them

This is dependent on what other activities or documentation you will be generating for your project. It’s not uncommon for our sessions to end with a data cluster that’s labeled “persona bio”. This cluster doesn’t really have a problem or insight and can be pushed off to the side.

Image for post
Image for post
Group Labeled with an Insight / Problem Statement

Creating Design Mandates

Continue iterating on your groups and their labels until you get everything recorded on a post-it. Once you have all the data in a reviewable state, your participants can start writing design mandates. There could be one or more mandates per insight.

  1. Finalize your group labels with succinct insights
Image for post
Image for post
Complete Data Cluster with Design Mandate

Closing thoughts

While the method is pretty straightforward, organizing and managing this kind of activity is a lot of work. You might be wondering why this is worth doing. The answer is universal team buy-in.

Collaborating as a team to find the problems then taking time to outline some solutions means you don’t have to sell anyone on the importance of conducting research. You also don’t have to waste time drafting a comprehensive findings document that ends up being skimmed or even ignored.

If you’re working with someone you suspect has an archaic definition for what it means to conduct user research, this activity is a good entry point that will hopefully change the way your organization thinks about the value of user research. Happy clustering!

Continued Reading

If you found this article useful, you can find a list of our other tutorials and templates on the Mentor website.

Written by

We design and execute digital experiences that make a difference.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store