Trend vs level
Trend is the overall direction a data path is moving, up, down or flat, while level is where the data sit on the y-axis, usually summarized as their average value. A change in level is a jump between phases; a change in trend is a change in direction.
What is the difference between trend and level?
Trend describes the slope of the data over time: increasing, decreasing or zero. Level describes the height of the data on the vertical axis, often described as high, low or as a mean value. Two phases can have the same trend but different levels, such as both flat but one around 10 and the other around 3.
When a new intervention starts, analysts look at both. An immediate drop in level right after the phase change line suggests a quick effect. A downward trend that develops over several sessions suggests a gradual effect.
Variability is the third feature: how much the data bounce around the trend or level. High variability makes both trend and level harder to read.
Trend and level side by side
| Trend | Level | |
|---|---|---|
| Describes | Direction of change over time | Height of data on the y-axis |
| Words used | Increasing, decreasing, flat | High, low, mean value |
| Read by | Looking at the slope across points | Looking at where points cluster |
| Change between phases | Data start heading a new direction | Data jump up or down at the phase change |
| Example | Correct responses rising session by session | Hits averaging 12 in baseline and 3 in intervention |
| Related term | Variability | Variability |
How do you tell trend and level apart on the exam?
If the question is about where the data are going, it is asking about trend. If it is about how high or low the data are, or about a jump at the phase change line, it is asking about level. A useful check: trend needs several points to see a direction, while a level change can be seen by comparing groups of points.
- Words like increasing, decreasing or slope: trend.
- Words like average, higher, lower or jump: level.
- Scatter around the line: variability.
Worked scenarios
Scenario 1. In baseline, a learner's requests per session go 2, 3, 4, 5, 6. Answer: an increasing trend.
Scenario 2. Aggression averages about 15 per session in baseline. On the first day of the new plan it drops to 5 and stays around 5. Answer: a change in level, with a flat trend in both phases.
Scenario 3. Problem behavior is steady around 10 in baseline, then falls by about one each session after the plan starts. Answer: a change in trend, from flat to decreasing.
Scenario 4. Data jump between 2 and 14 every other session with no clear direction. Answer: high variability, which makes trend and level hard to judge.
What are the common traps with trend and level?
A common trap is calling a single high point a trend. Trend needs a pattern over several points. Another trap is assuming an increasing trend is good. It is good for a skill being taught but bad for a problem behavior. A third trap is missing a level change because both phases have the same flat trend.
- Judging trend from one or two points.
- Ignoring whether the target should go up or down.
- Overlooking the jump at the phase change line.
When would you see each in a session?
You will meet trend and level whenever you graph session data or look at a learner's progress with your supervisor. An unexpected trend, such as problem behavior rising, is something to report promptly. Interpreting graphs to change the program is the supervisor's job, but entering data accurately and noticing worrying patterns is part of yours.
Enter data on the graph soon after each session. Late or batch entry can hide a developing trend that the team needed to see sooner.
Practice questions
In baseline, Nina's off-task data hover between 19 and 21 across sessions. Right after the phase line, intervention data hover between 4 and 6, with no upward or downward slope. Which describes the change?
During intervention, the graph of Andre's daily self-injury shows 14, 11, 9, 6 and 4 instances. How is this pattern best described?
On a line graph of correct responses per session, the data points rise steadily across the last six sessions after a new teaching procedure started. How is this best described?
After prompts began to be faded, Hana's number of independent toothbrushing steps across five sessions was 8, 7, 7, 5 and 4. Which description is accurate?
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?
Questions people ask
Is level the same as the mean?
Level is often summarized with the mean of a phase, though some analysts also describe it by the median or by the range of the data.
What is a phase change line?
It is a vertical line on a graph marking when a condition, such as a new intervention, started.
How ready are you?
Ten questions, about 3 minutes, no account. Get your readiness score and your weakest domains.
Check my readiness