Variability
Variability is the amount of bounce in a set of data points. Scores like 2, 15, 4, 18, 1, and 16 are highly variable, while 10, 11, 10, 9, 10 are stable. High variability makes it hard to see trend or level, and it is worth telling your supervisor about.
What is variability in ABA?
Variability describes how spread out data points are around their typical value. Low variability means points stay close together, and high variability means they swing widely from session to session. It is one of the three main features read from a graph, along with level and trend.
What does variable data look like?
On a graph, variable data zigzag up and down with large gaps between neighboring points. A learner whose correct responses go 2, 15, 4, 18, 1, and 16 across six sessions shows high variability. A steady set, such as 10, 11, 10, 9, and 10, is stable and easier to interpret.
Example: Tyler's performance bounces between very low and very high scores with no clear reason. A supervisor looking at that graph cannot say much about level or trend, so the first job is finding out what is causing the swings.
What can cause variability?
Variability can come from real differences between sessions, such as sleep, illness, medication changes, different staff, schedule changes, or whether a preferred item was available. It can also come from inconsistent data collection, like unclear definitions or observers who score differently. Reporting such details helps your supervisor sort out which it is.
- Changes in the learner's health, sleep, or routine
- Different people running sessions in different ways
- Unclear definitions or inconsistent scoring
- Changes in motivation, such as no access to a preferred item
Why does variability matter when reading graphs?
Stable data let you see change clearly. Highly variable data make it hard to tell whether a shift is a real effect of the plan or just normal bounce. Supervisors often wait for steadier data in baseline before starting, or look at what is creating the swings before judging whether the plan works.
What should an RBT do about variable data?
Keep collecting data carefully using the written definition, and add objective session notes about anything unusual, such as a poor night's sleep or a new staff member. Do not average, smooth, or leave out low or high points to make the graph look cleaner. Share the pattern with your supervisor.
Exam trap: questions about which baseline is most stable test whether you can spot narrow ranges. Pick the set whose points are closest together.
Practice questions
A graph of Tyler's correct responses shows 2, 15, 4, 18, 1 and 16 across six sessions. What stands out most about these data?
Which set of baseline data points shows the most stable pattern?
Which session note entry best documents a variable that may have affected 5-year-old Jae's session?
A technician notices a learner's aggression is highest at Monday sessions. She learns the learner spends weekends at his other parent's home, where routines are different. What should she do?
Questions people ask
Is variability always a problem?
Not always, but it makes interpreting a graph harder, so your supervisor will want to know about it.
Should I remove outliers from the graph?
No. Record and graph accurately, and describe any unusual circumstances in your notes.
How ready are you?
Ten questions, about 3 minutes, no account. Get your readiness score and your weakest domains.
Check my readiness