Variability
Variability is how much data points differ from one another within a phase; high variability means values bounce up and down widely from session to session.
Variability describes how spread out the data are. Data with low variability are stable: the points stay close together, for example 8, 9, 8, 10, 9. Data with high variability jump around, for example 2, 14, 5, 18, 3. Stable data make it easier to see trend and level and to judge whether an intervention made a difference.
High variability is information, not just noise. It suggests that something other than the planned procedures is influencing behavior from session to session. Possible sources include changes in sleep, illness, medication, setting, people present, the time of day, the materials used, inconsistent implementation of the plan, or inconsistent data collection. Supervisors may wait for data to stabilize before changing phases.
An RBT cannot fix variability by smoothing the numbers, and data should never be adjusted. What RBTs can do is collect data consistently, follow the plan exactly, and document and report variables that might explain unusual sessions, such as "client slept 3 hours" or "substitute teacher". That information helps the supervisor understand the pattern and decide what to do.
Example and non-example
| Example | A client's daily count of elopement attempts goes 1, 9, 0, 12, 2. The data show high variability, and the RBT's notes reveal that the high days were days with a substitute teacher. |
|---|---|
| Non-example | Counts of 6, 5, 4, 3, 2 change every session but in a steady direction; that is a decreasing trend with low variability. |
Don't confuse it with
Test yourself
-
Question 1Which best describes these data?
Number of correct responses by Session: 1: 5, 2: 5, 3: 6, 4: 5, 5: 5, 6: 6, 7: 5. Phases: Intervention from session 1. Show answer
B: Stable, with no trend. The points stay within a narrow band without rising or falling overall, so the data are stable with little or no trend.
- A: The values do not rise across sessions.
- C: The values do not fall across sessions.
- D: The points differ by only one unit at most, which is low variability.
Task A.7 · 40-hour curriculum section C
-
Question 2Which best describes these data?
Number of instances of throwing by Session: 1: 2, 2: 9, 3: 1, 4: 10, 5: 3, 6: 8, 7: 2, 8: 9. Phases: Intervention from session 1. Show answer
A: Highly variable, with no trend. The points jump between high and low values from session to session without moving in one overall direction. That pattern is high variability.
- B: The high and low points stay at about the same values over time, so the data do not rise overall.
- C: The data do not fall overall; the highs and lows stay about the same.
- D: Stable data stay close together; these points spread far apart.
Task A.7 · 40-hour curriculum section C
-
Question 3Which best describes these data?
Number of independent requests by Session: 1: 10, 2: 30, 3: 20, 4: 40, 5: 30, 6: 50, 7: 40, 8: 60. Phases: Intervention from session 1. Show answer
A: An increasing trend with some bounce. Although the points go up and down from one session to the next, the overall direction is upward. The data show both an increasing trend and some variability.
- B: Each low point and each high point is higher than the one before, so the direction is up.
- C: The ups and downs sit around a line that is clearly rising.
- D: The values both bounce and rise, so they are neither stable nor flat.
Task A.7 · 40-hour curriculum section C
Answer every question to see your score.
On the RBT exam
This term belongs to task A.7 Identifying trends in graphed data in content area A, Data Collection and Graphing, which is 13 of the 75 scored questions.