No abstract

In three experiments we used control-system theory (CST) to predict the results of tracking tasks on which people held a handle to keep a cursor even with a target on a computer screen. 10 people completed a total of 104 replications of the task. In each experiment, there were two conditions: in one, only the handle affected the position of the cursor; in the other, a random disturbance also affected the cursor. From a person's performance during Condition 1, we derived constants used in the CST model to predict the results of Condition 2. In two experiments, predictions occurred a few minutes before Condition 2; in one experiment, the delay was 1 yr. During a 1-min. experimental run, the positions of handle and cursor, produced by the person, were each sampled 1800 times, once every 1/30 sec. During a modeling run, the model predicted the positions of the handle and target for each of the 1800 intervals sampled in the experimental run. In 104 replications, the mean correlation between predicted and actual positions of the handle was .996; SD = .002.

We present a typology of tactics used by probation officers to steer probationers away from day-to-day crime opportunities based on 40 qualitative officer interviews. Concerns about probationers' unstructured activities, time at crime-prone locations, and "at-risk" social interactions led officers to reduce probationers' unstructured time, set limits to locations and persons with whom they associated, and coach them to side-step such opportunities. In support of these tactics, they collaborated with community handlers, worked to enhance probationers' motivation and deter them from risky activities, leveraged probation conditions, and encouraged alternative activities.