Analysis of Data

Mean: Average rate of performance.

The "mean" is the best indicator of functionality of your design and plan. If using an ABA design, the mean in phase A represents your average in baseline data. The mean in Phase B represents the average performance of the student during intervention. If there is a large spread between the means of Phase A and Phase B, it would communicate the effectiveness of your intervention. Upon returning to baseline (second phase A) we would wish to see a mean that somewhat reflects the mean you achieved in the first baseline phase. This would indicate it was your independent variable (intervention) that caused the behavior to change rather than some extraneous variable.

Step One: Calculation

1. Add the numbers represented by the data point within the phase.

2. Divide the number obtained in Step One by the number of data points.

Example: Data Points 3, 5, 10. Adding these together equals 18. Divide by the total number of data points (3). M=6.

Step Two:

Calculate the mean for each phase

Step Three:

Draw a horizontal dashed-line on the graph representing the mean for each phase

 

Level of Performance: We are looking at the change in performance from the last data point in one condition to the first data point in the next condition. We refer to the "change in level of performance" as "abrupt" if the change in behavior occurs immediately when starting the next phase. (for example: last data point in baseline is 5, first data point in intervention is 10; there is an abrupt change in level of performance (5)).
1. Identify ordinate value of last data point in first condition and the first data point in second condition. 2. Subtract the smallest from the largest. 3. Note if the change in level is improving (therapeutic) or decaying (opposite direction of what you want to go).

 

Percentage of Overlap: Looks at how the data from one phase differs from the data in the following phase. The lower the percentage of overlap, the greater the impact the intervention has had on the target behavior.

1. Determine the range of data point values of the first condition.

For Example: 10, 8, 15, 14, 7; Range equals 7-14.

2. Count the number of data points in the second condition

For Example: 13, 15, 12, 8, 16, 19, 20, 15; N=8)

3. Count the number of data points in the second condition that fall within the range of the first condition.

For Example: In this example 13, 12, & 8 fall between the range of 7-14, so the number is 3.

4. Divide the number of data points that fall within the range of the first condition by the total number of data points of the second condition and multiple by 100.

Example: In the example we would divide 3 (from step 3) by 8 (from step 2) and take the answer by 100. The "percentage of overlap for the data would be 38%. This means that 38% of data point from the intervention phase overlap with the data points in the baseline phase.