Conditional Probabilities Observation Form (Adapted from Eckert, Martens, & Di Gennaro, 2005)1

Conditional Probabilities Observation Chart

(Adapted from Eckert, Martens, & Di Gennaro, 2005)

During a traditional Antecedent-Behavior-Consequence (‘ABC’) recording, the observer writes a running account of a student’s behaviors and the classroom events that appear to trigger, or to act as consequences of, that behavior. A drawback of ABC recordings is that they may not show clear relationships between behaviors and the consequences that reinforce those behaviors. The Conditional Probabilities Observation Form (based upon a description from Eckert, Martens, & Di Gennaro, 2005) allows the observer to compute and chart the conditional probability that a specific consequence is contingent on a target student behavior. This information can be useful in completing a Functional Behavioral Assessment and in designing behavioral interventions.

Selecting and Defining a Target Behavior. Using a combination of data-collection methods (e.g., teacher interview, direct observation of the student), determine a single target problem behavior that is to be observed using this form. Define the behavior in clear, measurable terms. Here is an example of a behavioral definition for ‘off-task’: “any passive behavior, motor activity, or audible verbalization lasting longer than 3 seconds that was not permitted and/or was not related to an assigned academic task.” (Eckert, Martens, & Di Gennaro, 2005; p. 522)

Conducting the Observation.

  • Conduct a series of brief (10-15 minute) observations in the location(s) and during the activitie(s) when the student typically displays the problem (‘target’) behavior. Three to five observations are recommended.
  • During each successive 15-second interval of an observation period, mark the ‘Behavior?’ box with an ‘X’ if the student displays the target behavior at any time during that interval. Leave the box blank if no behavior occurs.
  • During each 15-second interval, mark the appropriate code into the ‘Consequence?’ box for each consequence that occurs during the 15-second interval. Use the following codes: P = Peer Attention; T=Teacher Attention, I=Item Activity, E=Escape, N=No Consequence. Mark every consequence observed, whether or not the target behavior occurs during the interval.

To Score:

  1. Counting Up Intervals With and Without Contingent Consequences. Compute the number of intervals during the observation in which the Target Behavior AND Consequence occurred. Write these values into Table 1, Column A. Then compute the number of intervals in which the Consequence occurred WITHOUT the Target Behavior. Write these values into Table 1, Column B.
  2. Computing Conditional Probabilities of Contingent Consequences. In Table 2, enter the requested values in each column and follow the directions in the table to compute the probability that BOTH Target Behavior AND Consequence Occurred.
  3. Computing Conditional Probabilities of Non-Contingent Consequences. In Table 3, enter the requested values in each column and follow the directions in the table to compute the probability that the Consequence occurred WITHOUT the Target Behavior.
  4. Charting ‘Conditional Probabilities.’ The Y (vertical) coordinate for each consequence can be found in the appropriate row of Table 2, Column F. The X (horizontal) coordinate for each consequence can be found in the appropriate row of Table 3, Column F. Plot the ‘conditional probability’ of each consequence occurring in the presence of and in the absence of the target behavior. Next to each plotted value, write the name of the consequence plotted (e.g., ‘Peer Attention’).
  5. Interpreting the Chart. Any consequence that falls above the ‘unity diagonal’ line in the chart suggests that the consequence is contingent on (occurs in the presence of) the target behavior. The farther above or to the left of the line the plotted value falls, the greater the demonstrated relationship between consequence and target behavior. Any consequence that falls below the ‘unity diagonal’ line in the chart suggests that the consequence is not contingent on (occurs independently of) the target behavior. The farther below or to the right of the line the plotted value falls, the greater the evidence of independence between consequence and target behavior.

Eckert, T. L., Martens, B. K., & Di Gennaro, F. D. (2005). Describing antecedent-behavior-consequence relations using conditional probabilities and the general operant contingency space: A preliminary investigation. School Psychology Review, 34, 520-528.

Jim Wright