Block title
Block content


January 26, 2017
Sean Connell

Observational research isn’t a perfect science. One of the biggest issues is reflected in the Hawthorne Effect, which states that those being observed tend to alter their behavior in a study. To balance the value of this type of research with the desire to reduce bias, we’ve developed some approaches that can enhance learning while not unduly influencing the interview. We find that the key to success in this space is to work with the environmental factors, not against them.

“Com-” prefix meaning with, together


Let the subject direct the flow.
I frequently do research with surgeons, often in different countries with different surgical approaches and access to different products. Recently, I was observing them conduct surgery on live pigs, which is necessary to simulate blood flow. For this research, I needed to be stationed closely to the surgeon (with an interpreter in non-English speaking countries) and watch him/her use a surgical device. Clearly, my ability to interfere and introduce bias was strong. To complicate matters, there was a desire to probe every minute detail in the moment it was happening. But when we sat back and gave the surgeon time to explore on his/her own, we discovered that 1) they would often talk to us about what they were seeing/experiencing without our prompting; and 2) they would use the product in their own, unique way. Had we interrupted the surgeon earlier in the process, they may never have completed the task in the way that was most natural to them.


Be discreet – it’s about them, not you.
We all know the answer we want the subject to give; but the best findings come when they give you the unexpected answer. The greatness of observational research is that it puts the subject in the moment, but your very presence can disrupt that moment. In my case, the surgeons had a non-clinician looking over their shoulder (me) speaking a foreign language, and a translator whispering in my ear. This only exacerbates the Hawthorne Effect. By making yourself small, you give the subject ample time and space to shine while not making them feel like they’re in a fishbowl.


Make the interview conversational, not rote.
In this example, there were several clients in the operatory and all of them wanted to ask questions. I want my clients to have their questions answered, but I also want to put the surgeon at ease. The key is to demonstrate to the subject and the client early on that this is a conversation between the moderator and the subject. This allows us to stay on-task, avoid questions that aren’t germane to the objectives, and establish rapport and trust. My goal as a moderator is to anticipate questions the client may have as well as filter those questions so as to meet our research objectives. I made sure that my client knew that they would have time at the end for ancillary questions or items that I may have missed. Moderators are human, after all, and sometimes we forget an occasional probe. Combining these factors in observational research helps make the experience as natural as possible.

Commit » Compromise » Communicate