Opinions are not behaviors

Observing rat behaviour

Observing rat behaviour (Photo credit: jepoirrier)

One of the most important things to look at when conducting user research is whether the technique is asking for participant opinions (subjective) or whether it is measuring real user behavior (objective).

What people say they do is often not what they actually do.

Most techniques and analysis don’t fall squarely into one category or another. Surveys, for example, are directly asking for opinions, but if you use textual analysis, you’ll start to see behavioral patterns in the language people choose to describe various tasks and events. Usability testing involves watching people perform a task directly, but will often include the participant vocally interjecting their opinion into the process.

Both opinions and behaviors are valuable to understand. From a user experience research perspective, behaviors are far more valuable. A critically overlooked proposition, however, is where behaviors and opinions differ.

The most common example I can think of is choice. Users will always demand more choices, more options, more. Psychologists and user experience researchers can tell you, however, that having more choices available does not lead to higher satisfaction, even while people always tell you they want more. In fact, most people can only make a choice between about three or four things easily. After that point, you’ll see a pretty dramatic drop-off in cognitive ability and satisfaction. Filtering options become critical for this reason once you have more than 4 or so choices available.

Another major difference between behavior and opinion is we tend to way overestimate the effects of change, both in positive and negative directions. We believe adding feature X will make our lives 100 times better, while losing capability Y will completely devastate us. Time doesn’t show this to be true in the majority of cases; as humans we are quite capable of adapting to new circumstances.

The last type of difference between opinions and behaviors I’ll mention in this article is we tend not to believe much of anything influences us, even though we are heavily influenced. For example, marketers know that adding quotes from other customers on a page or adding a little picture of a lock next to a button can have a huge impact on sales. No one wants to believe these little elements will sway our decision making, but advertisers can tell you from experience, no one is above being influenced.

For more examples, I’d recommend 100 Things Every Designer Needs to Know About People.

For more on this subject, check this blog post from UserVoice.

Thoughts, questions? Other examples of behaviors and opinions differing? Reply to this post.