Social psychologists suggest that before people engage in an action, they cognitively create behavioral plans (i.e., intentions).  You probably intend to drive to work before you get in the car, and you plan to bake cookies before warming up the oven.  Hundreds of studies have examined behavioral intentions, and these studies have found that behavioral intentions alone reliably predict 20% to 30% of human behavior. These percentages might sound insufficient to make strategic data-driven decisions, but compared to other factors, behavioral intentions are relatively strong predictors of human behavior.

We can take advantage of this reasonably reliable relationship between behavioral intention and behavior. When research funds are tight, project time is limited, or when we need to make predictions (e.g., will this ballot measure pass), measures of behavioral intention can substitute for measures of actual behavior. Behavioral intention can be measured by one or a few questions (e.g., How likely is it that you will drive to work tomorrow; if the election were held today who would you vote for), which can easily be incorporated into most surveys. This approach is typically much faster than measuring actual behavior in a follow-up survey; unfortunately it also has some drawbacks.

The most obvious disadvantage for using behavioral intention as a proxy for behavior is that people don’t always follow through with their intentions. How many Americans intend to hit the gym after the New Year? Once I told my wife that I intended to buy her jewelry for a special occasion.  She told me, “It is not the thought that counts.”  She made a good point, intentions alone aren’t worth much. Considering the countless reasons that we fail to fulfill intentions (e.g., we run out of time, we change our priorities), substituting measures of actual behavior with measures of behavioral intentions has its shortcomings. Still, this approach does provide a cost effective way to gain deeper insights and may be the only way to make certain kinds of projections.

The last blog in this series will talk about measuring actual behavior.  Stay tuned.

Other  blogs in this three part series: