GoalTallyQuick answer: A trial-by-trial data sheet records the outcome of every learning trial in a session, usually + for a correct, independent response and – for an incorrect, prompted or missing one. At the end of the session, % correct = number of + ÷ number of trials run × 100. That percentage is the number you graph.
How to fill in a trial-by-trial data sheet
- Write the date at the top of the column before the session starts.
- Present the instruction (the discriminative stimulus, or SD) and wait for the response.
- Circle + or –. + if the student responded correctly without help; – if they responded incorrectly, didn't respond, or needed a prompt. To track prompted trials separately, switch the codes to + / P / –.
- Total the column. After the last trial, write the number of + marks and the % correct in the rows at the bottom.
- Graph the percentage. With on-screen scoring, the totals and the graph point are filled in for you.
Worked example
Ten trials of “Touch red” across three sessions:
| Session | Correct (+) | Trials run | % correct |
|---|---|---|---|
| 1 | 6 | 10 | 60% |
| 2 | 7 | 10 | 70% |
| 3 | 6 | 8 (session ended early) | 75% |
Session 3 shows why the percentage uses trials actually run: 6 of 8 is 75%, not 60%.
Every trial, or first trial only?
Recording every trial gives the most detail and shows learning within a session. Many programs record only the first trial of each target per session instead, because it's fast and a strict test of what the learner can do without help. For that approach, use the cold probe data sheet. Whichever you choose, keep it consistent so sessions can be compared.
Choosing a mastery criterion
The BCBA or IEP team sets the criterion. Common examples are 80% or 90% correct across 2 or 3 consecutive sessions, sometimes with a different staff member or setting to check that the skill generalizes. Write the criterion on the sheet so everyone knows when to move on, and use the mastery check in the ABA graph maker to see when it's been met.
Tips
- Score straight after each trial rather than from memory at the end.
- Keep the number of trials similar across sessions where you can. If it varies, % correct still compares fairly.
- Follow your program's rules for error-correction trials. Most don't count them as new trials.
- Use the notes box for anything that might explain the data: a new reinforcer, a substitute teacher, a short session.
Frequently asked questions
How do you calculate percent correct on a trial-by-trial data sheet?
Divide the number of correct trials by the number of trials actually run and multiply by 100. 7 correct out of 10 trials is 70%. If only 8 trials were run and 6 were correct, it's 6 ÷ 8 = 75%.
Should a prompted response count as correct?
Usually not, when the measure is independent correct responding. Score it as – or use the + / P / – codes so prompted trials are visible but not counted as correct.
How many trials should a session have?
It depends on the learner and the program. Sessions of 5 to 20 trials per target are common. The sheet supports 1 to 40 trials.
Can I fit more sessions on one sheet?
Yes. Each column is a session. Choose up to 31 columns; the sheet switches to landscape automatically when the columns need more room.
More free graphs and data sheets
Last updated · Written by the MiniTools team.