To navigate the social world, people need to be able to anticipate how others will act. Whether in competition or cooperation, you're better off if you can predict other people's behaviors and tailor your own actions appropriately. Fortunately, the world makes this possible: there are many features such as situations, personalities, and emotions that can help us predict other people. Here we focus on one such feature: other people's current actions. If you know what some is doing know, there's good reason to hope that you can predict what they're going to do next. The current study assesses participants' accuracy in making these action-to-action predictions.

The task you will engage in consists of rating the probabilities of transitions from one action to another. On each trial you will view the name of an action, such as "cooking" and be asked to rate how likely it is that someone currently engaged in this activity will next engage in another activity, such as "eating." Sometimes you will rate how likely an action is to "transition" back to itself. In such cases you should rate how often people repeat the same action twice in a row (e.g., if you're currently eating, how likely is it that the next thing you'll do is eat another meal?). At the end of the study, you will receive personalized feedback about the degree to which your ratings match up with estimates of the actual transitional probabilities between these actions, which we measured using separate data. You will be able to see which specific transitions you were more or less accurate about, as well as your overall accuracy. If you wish to participate in this study on action prediction, please click the 'begin' button below.