:: Don't try to quantify qualitative responses

Nothing is more confusing than a qualitative research study that tries to incorporate quantitative measurement. Prior to conducting a usability test, for example, my client questioned why I wasn't planning to record how fast it would take my participants to complete key tasks. There are several reasons why I feel this measurement would not yield useful information:

1. In a test environment, the speed in which a user completes a task is inherently biased due to the fact that the participant knows he/she is conducting a test.

2. The time in which it takes a user to complete a task changes based on the user's familiarity with the system or website. Therefore, "time to task completion" will decrease as the user gains familiarity with the website.

3. (most importantly) I am interested in the participant's "qualitative reaction" when completing the task. I am listening to their words and watching their facial expressions to understand how frustrating the experience is. Finally, I am looking for trends in behavior, rather than trying to quantify participants' speed to task completion.

Qualitative research is meant to uncover trends in behavior, which will yield rich insights to inform interaction design. Behavioral research is conducted with limited numbers of participants, so the data uncovered in this type of research is difficult to quantify. Quantifying behavior (e.g. - "calculating time to task completion") is less informative to design than identifying common behaviors and negative behaviors, anyway.