Using Social-Media Tools to Connect with Users: A Case Study

Beware of Virtual Frauds

For any research, it’s important to ensure that you have the right participants to get the highest-quality feedback. This need is sometimes exacerbated in qualitative, longitudinal studies, when you want a smaller group of people to provide continuous feedback over time. One rotten egg can really throw off your data. In traditional, face-to-face research, it’s usually easy to identify someone who has fallen through the recruiting screener’s cracks. A quick conversation can be a good test to see whether they are who they said they are. When using social-media or other online tools, identifying frauds can be trickier.

In our case, we relied on old-fashioned face time to ensure our participants were what we were looking for and conducted group Skype video calls with all participants. This virtual face time gave us more insights into who our participants were—beyond their Facebook posts and survey answers—and it allowed us to tease out the online frauds.

While it’s much easier for participants to use fake personas while answering a survey, it’s much more difficult to pretend to be an 18-year-old college student during a video group chat on Skype. The group discussions made us realize that two of the participants were obviously older than they had represented themselves to be—and clearly didn’t fit our requirements. Conducting such calls early in the research process can save you from having to deal with the headache of retroactively removing data from the pool once you’ve outed individuals as virtual frauds.

Pitfalls and Advantages of Sharing with the Group

While Facebook provided a natural way of engaging with our user group, we found that the more anonymous forms of feedback were most effective at getting meaningful responses. Perhaps the public nature of the Facebook group—even though technically a private group—made it feel like we’d brought together a bunch of strangers in an unnatural way.

For our future research, establishing what would naturally bring everyone together from the start might be helpful in fostering more of a sense of community. For example, we could be sure to highlight the common age range and general interests of the group to build a cohesive feeling of inclusion among people with similar interests.

However, as with any group research, we also needed to be wary of a few different forms of bias that might be at play. In this particular case, there were two main biases that we should be concerned about: social desirability bias and conformity bias.

Social desirability bias is the tendency for people to want to put themselves in the best light possible—giving what they think is the best answer rather than giving an honest answer. This is often a concern for surveys. Social desirability bias could be even more pronounced when survey questions are visible—as in the built-in Facebook polls.

Conformity bias is the tendency for people to behave in similar ways to other people in a group. When we asked participants to post pictures or content to the Facebook group, we’d see instances of similar responses from participants. For example, one person might post about wanting a certain type of computer, then the next three posts would be about a similar topic or a similar brand. Because of some of the biases that we feared we’d observed, we ended up using these tools mainly as a way of keeping the group engaged rather than as a way to gather concrete answers.

Conclusion

While Facebook may not be an ideal platform for all research purposes, it gave us an opportunity to connect with the specific age group of our target audience—in a way that was easy for our participants, too. These people already checked Facebook often, so asking them to extend their use of Facebook for our research purposes was consistent with their natural behavior.