cumulative

C.11 Interpreting graphed data

What the task asks: Read a graph of behavioral data and judge what it shows about the behavior and about the effect of a change in conditions, using level, trend, variability and related features.

Measurement, Data Display, and Interpretation5 min read4 practice questionsOfficial outline (opens in a new tab)

The big idea

Visual analysis is the systematic inspection of graphed data to decide whether behavior changed and whether the change can be credited to a change in conditions. Behavior analysts rely on it because single-case data are collected repeatedly for one person, and the decisions are practical ones: keep the plan, change it, or end it.

The method has two stages. First describe the data within each phase. Then compare adjacent phases.

Breaking it down

Before you start: read the graph itself

Check the axes, the measure and the condition labels. A vertical axis running from 40% to 60% makes a 5-point change look dramatic. Note missing sessions and how many data points each phase has. A common standard asks for at least three points per phase, and preferably five, before drawing conclusions.

Within-phase analysis

  • Level: how high the data sit on the vertical axis, usually summarized by the mean or median. Baseline values of 8, 10, 9, 11 and 7 have a mean of 45 ÷ 5 = 9.
  • Trend: the overall direction and steepness of the data path: increasing, decreasing or zero trend. It can be judged by eye or with a fitted line such as a split-middle line.
  • Variability: how much the points bounce around the level or trend line, often described by the range. High variability hides effects and suggests uncontrolled influences.
Level Same bounce, lower A B Trend Flat, then rising A B Variability Same level, more bounce A B
Three pairs of phases, each differing in one feature. In the first, the level drops while the bounce stays the same; in the second, a flat phase gives way to a rising trend; in the third, the level holds while variability grows.

A baseline that is stable, with low variability and no trend toward the expected change, gives the clearest prediction of what would happen without intervention (D.4).

Between-phase analysis

Compare each phase with the one next to it:

  • Change in level, trend and variability.
  • Immediacy of effect: how quickly the change appears after the condition changes. A common method compares the last three points of one phase with the first three of the next. Immediate changes are more convincing than delayed ones, which leave room for other explanations.
  • Overlap: how many points in one phase fall within the range of the adjacent phase. Less overlap means a clearer effect. The percentage of nonoverlapping data (PND) is the percentage of treatment points beyond the most extreme baseline point.
  • Consistency: whether similar phases look alike, such as both baselines in a reversal design, and whether the effect repeats each time the condition is introduced.

These six features (level, trend, variability, immediacy, overlap and consistency) follow the single-case standards of the What Works Clearinghouse (Kratochwill et al., 2010).

Seeing it in practice

Keisha’s call-outs per lesson are recorded across five baseline sessions (8, 10, 9, 11, 7) and five sessions of a hand-raising program (6, 4, 3, 2, 2).

04812BaselineHand raising12345678910SessionsCall-outs per lesson
Keisha's call-outs, with dotted lines at each phase's mean (9, then 3.4). The level drops, the intervention data fall and then level off, and no intervention point reaches the lowest baseline point.
FeatureBaselineInterventionReading
Level (mean)45 ÷ 5 = 917 ÷ 5 = 3.4Level dropped by 5.6
TrendRoughly flatDecreasing, then leveling offChange in direction
Variability (range)7 to 112 to 6Less bounce
ImmediacyLast three: 9, 11, 7 (mean 9)First three: 6, 4, 3 (mean 4.3)Fairly immediate drop
OverlapLowest point is 7No point at 7 or abovePND = 5 ÷ 5 = 100%

Every feature points the same way, so the graph shows a clear change. Whether the program caused it depends on the design: a reversal or a multiple-baseline replication would add the consistency evidence (D.7).

Now imagine the same intervention data after a baseline of 12, 10, 8, 6 and 5. Baseline was already falling steeply, and the intervention points simply continue the line. That graph gives little evidence of an effect, because the baseline trend predicted the drop.

04812BaselineHand raising12345678910SessionsCall-outs per lesson
The same intervention data after a baseline that was already falling. The intervention points simply continue the baseline trend, so the graph gives little evidence that the program changed anything.

Common mix-ups

  • Level vs. trend. Level is how high; trend is which way and how steeply. A phase can be high and falling.
  • Baseline trending toward improvement. If baseline is already moving in the hoped-for direction, a continued change does not show an effect.
  • Deciding from one or two points. Wait for enough data to see the pattern.
  • Relying on the mean. Two phases with the same mean can differ completely in trend and variability.
  • Overlap alone. One extreme baseline point can distort PND. Weigh it with the other features.
  • Ignoring graph construction. Check the axes before judging the size of a change.

Exam tip: Work in order: within each phase (level, trend, variability), then between phases (immediacy, overlap, consistency). When a stem asks whether there is an effect, ask what the baseline predicted would have happened without the change.

Quick recap

  • Visual analysis starts within phases (level, trend, variability) and then compares phases.
  • Between phases, look at changes in level, trend and variability, plus immediacy, overlap and consistency.
  • Stable baselines give the clearest prediction; baselines trending toward improvement weaken conclusions.
  • Less overlap and more immediate change mean a clearer effect.
  • Check the axes and the number of data points before interpreting.

Further reading

  • Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.), chapter 6.
  • Kratochwill, T. R., Hitchcock, J., Horner, R. H., Levin, J. R., Odom, S. L., Rindskopf, D. M., & Shadish, W. R. (2010). Single-case designs technical documentation. What Works Clearinghouse.

Key terms

Visual analysis (Visual inspection)
The systematic inspection of graphed data to judge whether behavior changed meaningfully and whether the design shows that the independent variable caused the change. It considers level, trend and variability within phases, and immediacy, overlap and consistency across phases.
Level
The value on the vertical axis of a graph around which a series of data points clusters, often summarized by the mean or median. A change in level between phases is a common sign of an intervention effect.
Trend
The overall direction of a data path (increasing, decreasing or flat) along with its steepness. A baseline trend toward the improvement the intervention is expected to produce makes any effect hard to interpret.
Variability
How much the data points in a phase differ from one another or from the trend line. The more variable the data, the more measurements are needed before a confident decision can be made.
Immediacy of effect
How quickly behavior changes after a change in conditions, usually judged by comparing the last few data points of one phase with the first few of the next. The more immediate the change, the easier it is to attribute to the independent variable.
Overlap
In visual analysis, the extent to which data points in one phase fall within the range of data in the adjacent phase. Less overlap indicates a clearer difference between conditions.

Check yourself

4 questions on C.11. Choose an answer to see whether you're right and why.

  1. Question 1Baseline data for Mira's correct responses are 4, 6, 5, 5 and 5. Intervention data are 9, 11, 10, 12 and 13. Which description is accurate?
    Show the answer

    Answer: B. Baseline mean is (4 + 6 + 5 + 5 + 5) ÷ 5 = 5 and intervention mean is (9 + 11 + 10 + 12 + 13) ÷ 5 = 11. The highest baseline point (6) is below the lowest intervention point (9), so no points overlap. Baseline is roughly flat, and intervention values range more widely (9 to 13) than baseline values (4 to 6).

  2. Question 2Across six baseline sessions, Jonah's off-task behavior decreases steadily. An intervention then begins, and off-task behavior keeps decreasing at the same pace. What is the best interpretation?
    Show the answer

    Answer: D. A baseline trend predicts where the data would go with no change. Here the intervention data simply continue the baseline trend, so the graph does not show that the intervention did anything. A decrease alone is not evidence of an effect, and nothing in the stem describes a jump in level at the phase change.

  3. Question 3When a token system begins, Ines's completed assignments jump from about 2 per day to about 8 on the first day and stay there. Which feature of visual analysis does this illustrate most clearly?
    Show the answer

    Answer: A. A large change that appears right after the condition changes, with no delay, is a strong example of immediacy of effect. Consistency refers to similar phases looking alike across replications, which needs more than one phase change. The intervention data stay at about 8, so they show a change in level rather than an increasing trend.

  4. Question 4Baseline data for Mr. Brennan's daily outbursts are 12, 15, 10 and 14. Intervention data are 9, 11, 6, 5 and 4. What is the percentage of nonoverlapping data (PND)?
    Show the answer

    Answer: C. The goal is a decrease, so count intervention points below the lowest baseline point, which is 10. Four points (9, 6, 5 and 4) fall below it and one (11) does not, so PND = 4 ÷ 5 × 100 = 80%. The 20% figure is the share of overlapping points, not nonoverlapping ones.

Adds a step to your study record. Saved in this browser only.

Type to search all 104 guides and the glossary.