As Browder (2001) outlined three steps for data-based decision, this module will go one step further. The first step deals with summarizing the data. The second step addresses analyzing the data. The third step is to apply data-based decision rules. The final step is to design a plan.

Graphing is one of the easiest ways to summarize data. Data can be charted by the summary of responses that are independently correct (4 out of 5) or the percentage of mastery (80 out of 100). Dates are listed in a row along the bottom of the graph or data sheet along with the number correct/percentages. Prior to the intervention and after approximately three points of baseline data are collected, **the aim line** or expected progress during the data collection period is charted. The aim line uses the baseline average as its beginning point. If the student has no independent correct responses during the first three data points, the aim line should begin at 0. If, however, the student has some independent correct responses, the first three points should be averaged together to represent an accurate starting point for the baseline. Let’s look at the example below in Figure 1. Notice the aim line begins at 4 since the student’s data for the first three points are 0, 0, and 13.

Examine the graph closely. It includes the steps of the task analysis in the first column and includes the percentage correct in the second column. This allows the teacher to include a **progress line** on the graph that shows the percentage correct. To create a line of progress, use graph paper or the data sheet to plot independent correct responses or percentages, marking each time data is taken. The points are then connected to show the students progress. Notice the line of progress changes when the student responds correctly in trials 3 and 8.

Figure 1. (P=prompt)

Let’s look at another example in Figure 2. The skill is "Given a one dollar bill, a five dollar bill, a ten dollar bill, a twenty dollar bill, and a written total (e.g., $4.50), Student will identify the correct bill to make a purchase for 4 out of 4 amounts for 3 consecutive days."

Figure 2.

In this graph, the number of correct responses for a trial is plotted instead of the percentage correct. The aim line begins at 0 since the first three data points average to 0.

The teacher should review the data every 2 to 3 weeks. Since the progress line requires six points of data, six is the minimum number of points required before conducting a review of the student’s progress. There can be many more than six depending upon the review timeline and the opportunity to collect data. (Ten or more is a better representation of student performance over time.) The progress line will either accelerate (i.e., increase over multiple trials), decelerate (i.e., decrease over several trials), or there will be no change. Depending upon the type of skill, data should accelerate or decelerate over time. A progress line that does not change indicates the student is either not responding or responding incorrectly on most if not all trials.

Browder (2001) and Browder, Spooner, and Jimenez (2010) outlined several data rules. Let’s review each rule and look at a corresponding graph that exemplifies the rule.

1. *Mastery*. A student who reaches the criterion set in the behavioral objective is considered to have mastered the skill. The previous graph in Figure 2 above illustrates a student who mastered the skill as the last three data points reached the mastery level (4 out of 4 correct for last three data points).

Figure 3.

The graph in Figure 3 illustrates that the student is making steady progress by increasing the number of independent accurate responses over the course of the review period. While the student has not reached the mastery criterion on any occasion, the acceleration of the progress line has changed several levels during the review period and is consistent in direction (upward trend). The review points in the graph (the last three data points) average over a 50% change from the baseline points (0% baseline to 58%). Adequate progress is being made.

3.Figure 4.

The graph in Figure 4 illustrates that the student is making slow progress as only a few independent correct responses are being made after several weeks of instruction. The trend of the progress line is moving but not in a consistently upward direction to reach mastery. The review points in the graph (the last three data points) average to be under a 50% change from the baseline points (0% baseline to 41%). Slow progress is being made.

4.Figure 5.

Remember the skill “Given a one dollar bill, a five dollar bill, a ten dollar bill, a twenty dollar bill, and a written total (e.g., $4.50), Student will identify the correct bill to make a purchase for 4 out of 4 amounts for 3 consecutive days.” Using that skill and expectation for performance, the graph in Figure 5 shows that the student can do the skill - there are four days of 3 correct and one day of 4 correct. But intermixed between those days are performances of 0, 1, and 2 correct. If there are no medical or behavioral issues for the student, this performance warrants improving motivation for the student. It would be difficult to tell if the student will master this goal as there is not consistent trend in the progress line. It is important to note that while reviewing the first three data points (average of 0) and the last three (average of 58) the difference is the same as was found for our steady progress graph. However, the trend of the line is very different! There is no consistency within the trend of the line. Hence, the progress decision would be variable progress and require a different approach than when a student is making steady progress.

5.Figure 6.

There is no change in the trend of the progress line in Figure 6. There are very few independent correct responses.

There are a few issues that must be taken into consideration before a decision about student performance is made (i.e., if the data demonstrates slow progress or no progress, if the data is slow progress or steady progress). First, when reviewing the data, the number of items the student is working on can significantly influence the percentage of change. In the previous graphs, there were only four items. When the student produces even one more item correct in the last three data points (0% in baseline to 8% in last three points such as in the no progress graph, Figure 6), the change is enough to think that some progress is being made. But if you looked at the graph, the flat trend of the progress line and the length of the period for instruction support a decision of no progress.

A second issue is the goal for the student performance. A careful examination of the expectation for the student performance (e.g., is mastery set at 50%, 80%, or 100%?) is necessary to allow an accurate visual representation of the expectation through the use of the aim line. For example, it may be reasonable to expect a student to identify 2 out of 10 new sight words independently in one week. But a 20% mastery level is not an acceptable level to end instruction. The student may need four weeks of instruction and an increasing mastery expectation over time (i.e., week 2 may be 40%, week 3 may be 60%) to reach the acceptable rate of 80% for mastery.

Listen as Dr. Shawnee Wakeman discusses these issues further.

Additionally, there may be some instances when you would not apply data-based decisions. For example, if the student demonstrates a lack of progress that is not related to instruction such as the regression of student performance across all skills, there may be medical or behavioral interventions that are necessary before data can be collected on student academic performance. Another instance in which it would be erroneous to apply data-based decisions is when data collection is inconsistent such as when the criteria within the data collection system is not clear to instructors, there are too few data collection sessions, or there is too long of a time period between data collection sessions.

Applying data-based decisions requires graphing and summarizing the data, applying the decisions rules, and designing a plan for the next instructional sequence. Understanding what the data represents for student performance and carefully considering the student’s needs and preferences allow the teacher to better plan instructional efforts. Based upon the recommendations within the decision rules, the plan should be tailored to the individual student. The plan should have enough detail that it can be implemented with fidelity during the next benchmark review period of data collection and instruction.

There are three types of problems that correspond to the decision about student performance. The first is when no progress (i.e., no independent correct responses) is being made by the student. The recommendation is to simplify the skill. The second problem is when the student is making progress, but too slowly. The recommendations are to plan ways to fade prompts, teach the skill more times each day, and use peer modeling. The final problem is when the student is showing regression or variable performance. The recommendation, if the performance is not related to a medical or behavioral issue, is to enhance natural consequence and use reinforcers. Browder (2001) outlined several ways to address changes needed once a decision has been reached. For example, to simplify a response, the teacher can use chaining (i.e., instruction of only one part or piece of the task analysis); break the task analysis down into more discrete or smaller steps; include more distinct response choices to make the discrimination easier; and include scaffolds such as a jig or checklist. By selecting one or more of these recommendations, the teacher can then refine the plan with specific information that is unique to the student’s needs and abilities. The slides below provides more information about designing a plan for individual students that have one of these three types of progress and problems.

Transcript for Ideas to Design Plans Using Data-based Decisions

The presentation provides details about ideas for change within an instructional plan when a student is making no progress, slow progress, or variable progress. While these ideas to facilitate change are general and can be applied to different situations, the examples in the slides provide more detail about the types of situations that may occur. For example, if a student is unable to respond in the manner in which the task is designed, they will not be able to make progress. Therefore the task must be simplified by either the use of assistive technology and/or a change in the expectation of how the student is to respond within the task (e.g., use of an eye gaze instead of a verbal response).

Designing a plan may require the teacher to further analyze the student performance data. For example, when a student is performing at a variable rate, what is occurring on the days when the student is performing at his best that may not be occurring on the days when the student is not performing well? This level of analysis can help guide the plan for change.