Scatterplot Data for Behavior: Build One, Read One

A teacher says, "The yelling is all day." The scatterplot says otherwise: it happens mainly in the 9:30 reading block and the 1:00 transition, and almost never in math or at lunch. That difference is the whole point of a scatterplot.

This page shows how to build one, how to read it, and what it can and cannot tell you.

What a scatterplot is

Queensland's guide describes the scatter plot form this way. Write down all the regular activities or timetabled lessons the student has each week in the first column. Mark an X in the corresponding day column each time the behavior occurs in that activity. Leave blank if it does not. Tally totals after at least a week of data collection to identify patterns.

The IES practice guide describes a similar tool for a high-frequency behavior: a chart of the daily classroom schedule where the teacher makes a simple tally under the time of day and lesson activity when the target behavior occurs. Over time, it says, patterns should become apparent, showing when the behavior is more likely and less likely to occur.

When to use one

Queensland says a simple event tally or scatter plot may be useful for higher-frequency behaviors, while an ABC form may be best for behaviors that occur less often or last for several minutes. The two are not rivals. A scatterplot finds the times worth watching, and ABC entries explain what happens in them.

Build one in five steps

1. Define the behavior. One behavior, in observable terms. See writing an operational definition.
2. List the schedule. Every activity or period in the student's week, in order, one per row. Include transitions, lunch, recess, and specials.
3. Make columns for days. Five columns for a week, or ten for two weeks.
4. Decide who marks it. The person in the room. Use a clipboard or an index card, not memory.
5. Mark and total. An X or a tally for each occurrence. Total each row at the end.

Reading the grid

Read across first, then down.

• Across a row, a high total says this activity is where the behavior concentrates. Here the reading block has 16 and the transition to specials 10, out of 34.
• Down a column, a day that stands out may point to a day-specific cause, such as a different schedule or a substitute.
• The zeros matter. Lunch and math at 0 and 1 show that the student can manage those settings. That is useful for the plan, because the team can ask what is different about them.
• Do not over-read small numbers. One mark on a Tuesday is a clue, not a finding.

In the invented grid above, 26 of 34 occurrences fall in two activities. A team would write down both, then watch those two with an ABC form to see what comes before and after.

What to ask of the pattern

For each high row, ask four things: what is the task or demand, who is present, what is the setting like (noise, space, seating), and what happens just after. The IES guide lists time of day, classroom location, subject matter, type of learning activity, difficulty of the task, and the presence of particular peers or adults as key information to note about each instance. A scatterplot gives the first two. ABC entries give the rest. See ABC data: how to record.

From pattern to first steps

A scatterplot can suggest a quick change to try while the FBA continues. If the behavior clusters in the transition to specials, the team might test a warning, a job, or a visual schedule at that transition. If it clusters in the reading block, they might test shorter chunks. Keep the same scatterplot going, and compare totals before and after. That comparison is a baseline and a first check at once. See antecedent strategies in the classroom.

Before and after a change

Suppose the team tries a two-minute warning and a job at the transition to specials. They keep the same grid for the next week.

The transition row fell by six while the reading row barely moved, and the other rows stayed small. That is a useful result: the change seems to have helped where it was aimed and nowhere else. It is not proof, because one week can be unusual, so the team keeps going for another week or two before drawing a conclusion. They also turn to the reading block next, with ABC entries to see what happens there.

The habit worth copying is changing one thing at a time and keeping the grid the same. If three changes go in at once, the totals cannot say which one worked.

Limits

• It shows when, not why. The same cluster could come from different functions.
• It depends on the grid. If activities change week to week, update the rows and note the date.
• It can hide a behavior that moves. If the behavior shifts to another activity after you change one, you will see it only if every row stays on the sheet.
• It needs a steady recorder. Gaps in marking look like zeros.

Illinois's guidance says an FBA should describe the settings in which the behavior occurs most and least often, including data on its intensity, frequency, and duration, and that direct observation should occur across various settings. A scatterplot is a quick way to see both ends of that range, the busy rows and the quiet ones.

Keeping it light

Queensland's form fits on one page. Make the grid once, print a week at a time, and keep it on a clipboard where the behavior happens. If marking is hard in the middle of a lesson, mark during the next break and note that you did, so the team knows. Terms are in the glossary.

Local forms

Your district may have its own scatterplot form or a tool in its data system. The method is the same: activities by days, a mark for each occurrence, totals after a week or more. Ask your behavior specialist or school psychologist if you are not sure which to use.