Skip to main content
Posted on March 9th, 2012 by
Bring up the idea of using Focus Groups with most User Researchers and they will look at you as if you said you’ve been abducted by aliens. The skepticism is not totally unwarranted. Undergrad Psychology books are filled with examples of how focus groups are full of social dynamics that can produce misleading results. What researchers seem to forget is that all methods have their biases. Even the bastion of user testing, the usability study, has social dynamics that can produce false or misleading results.
(Image from alienxfiles.com)
The usability study bias is rarely discussed because any credible researcher knows how to design a study to reduce these biases. So why don’t these researchers think to apply this same process to focus groups?
First, while usability and field research methods are often taught and/or conducted as part of most graduate programs, focus group methods are not. We hear about the bad aspects, but never cover how to control for them.
Second, most of us researchers are a bit on the introverted side. We do fine in 1:1 environments, but we aren’t good in groups of strangers. (how many user researchers do you know that are the life of the party?) The whole idea of conducting a focus group is a bit unnerving. Combine a lack of experience with an uncomfortable situation and you get a method that is shunned by many.
Another problem is the baggage the term “Focus Group” carries with it. For many, it conjures up an image of 10 people sitting around a large table bored out of their mind while a moderator drones on. To me, the term means nothing more than conducting a research session with more than two people who can interact with each other face to face. By that definition, Focus Groups include more readily accepted methods such as participatory design or co-discovery.
So why bother with focus groups? The fact is they provide a unique environment for exploring different viewpoints, understanding priorities and determining common ground that is very difficult to get via other methods. Done properly focus groups are a powerful research tool.
In the next series of blog posts, I’ll discuss what I’ve learned in the last 12 years regarding how to make focus groups both effective and fun for the participants. In this post, I’ll focus on recruiting.
Proper recruiting is essential for any type of research, but even more critical for focus groups because of the social dynamic mentioned earlier.
Age. Mixing young and older audiences can be problematic. Mixing teens with older adults will usually result in the teens pulling back and deferring to the adults. Although most people would rather be quiet than look foolish, this is true in the extreme for teens. They are resistant to speak up in a setting with strangers.
Subject knowledge. Mixing experts and novices in a group will create a dynamic where the experts either show off or converse among themselves, leaving the novices feeling ignorant and left out. Therefore, make sure your screener separates out the extremes into their own groups.
While keeping groups uniform in terms of subject knowledge is always a safe bet, they don’t alway generate the most interesting discussion. For that, you need to bring in people with somewhat different perspectives. Although you don’t want to mix experts and novices, mixing in a few experts with a group of knowledgeable users or a few knowledgeable users in a group of novices can generate very interesting discussion. By mixing in only a few more knowledgeable people, you control the knowledge bias by putting them in the minority.
Group size. It’s simple: The bigger the group, the harder it is to manage and keep everyone engaged. The smaller the group, the easier it is to manage, the easier it is to generate dialog and make them feel like a “team”. I’ve found 4-5 per group works the best. Some people like to do groups of 8-12 because that way they get a bigger “sample”. I don’t like that approach because you are relying on the opinion of the group, so treating every person individually is methodologically questionable.
How many groups? Well, it is a bit like asking how many users. In usability studies, the minimum number of users I like per condition is 3. It’s hard to see patterns with less than that. While focus groups are more qualitative than usability studies, you still need enough groups to feel confident your results are somewhat generalizable. My usual recommendation is a minimum of two groups per condition, but it can vary based on specifics of the topic at hand. Although it is tempting to combine people into large groups for efficiency, I find that two groups of 5 will provide much better results that one group of 10.
Next time, I’ll cover ideas for encouraging individual and group participation.