Trial-by-trial data means recording the child’s response after every single teaching trial, usually as correct, incorrect or prompted. It gives the fullest picture of what happens in a discrete trial session. It also takes the most effort to collect, which is why good teams use it on purpose and not on every skill.

This guide explains how it works, what the research says about it compared with sampling only the first trial, and how parents, school staff and new clinicians can judge whether a data system is serving the child.

TL;DR

• Trial-by-trial data records the outcome of every teaching trial, so nothing about a session is hidden by an average or a sample.
• Research on children with autism has found that continuous measurement can lead to stricter mastery decisions and better detection of skill maintenance, but it is harder to sustain.
• Many teams use it for new, high-priority skills and switch to lighter methods, such as first-trial-only recording, for maintenance.
• The best method is the one your team can collect accurately and actually use to change the program.
• Ethics standards for behavior analysts require ongoing monitoring and data-based decisions, so any provider should be able to show you the data.

Table of Contents

• What trial-by-trial data actually is
• What it looks like in a real session
• What the research says
• Choosing the method: a quick comparison
• Making trial-by-trial data practical
• How parents can read the data
• Why providers are expected to collect data
• For RBTs and new BCBAs: why this skill matters
• Our perspective
• Where Building Block Resolutions fits
• FAQ
• Sources

What trial-by-trial data actually is

Discrete trial training (DTT) breaks a skill into small, repeatable teaching opportunities. Each one follows the same pattern: a cue or instruction, the child’s response, and a consequence such as praise, a reward or error correction. The AFIRM brief on DTT describes this three-part structure and uses imitation, receptive language and expressive language as common teaching targets.

Trial-by-trial data collection adds one step: after each trial, the therapist marks what happened. If a session has 10 trials on a target, you end up with 10 data points for that target. Researchers call this continuous measurement. In contrast, discontinuous methods record only a sample, most often the first trial of a session.

In published studies of DTT, therapists score responding trial by trial in the moment, and that is how many programs run day to day.

The other options you will hear about

• First-trial (cold probe) data: only the first trial of the session is scored. Because no prompts or earlier trials have just shaped the answer, it shows what the child can do cold.
• Probe data: a set of trials is scored at intervals, for example one probe block per day.
• Trial-by-trial data: every trial is scored.

The comparison of data collection methods in DTT notes that discontinuous recording can show performance without the influence of immediate prompts or earlier trials. That is a real advantage, not just a shortcut.

What it looks like in a real session

Imagine a therapist teaching a four-year-old to point to a picture of a dog when asked, “Touch dog.” This is an illustration of the method, not a case from a real child.

The therapist runs 10 trials. On the data sheet, each trial gets a mark: a plus for an independent correct response, a minus for an incorrect one, and a P for a prompted response. At the end of the block the sheet might show 3 independent correct, 4 prompted and 3 incorrect.

That pattern tells the team things a single score would not:

1. The child is not yet guessing randomly, because several answers were independent.
2. Errors may be clustered. If all three misses came in trials 7 to 10, fatigue or loss of interest may be the issue, not the skill.
3. Prompts are still doing much of the work, so the next step is probably a plan for fading them, not moving to a harder target.

If only the first trial had been scored, the team would see a single minus or plus and miss that story. If the first trial was correct by luck, the program might even advance too fast.

Over days, the same sheet turns into a graph. A BCBA looks at the trend: climbing, flat or bouncing. A flat line over several sessions is the signal to change something, such as the prompt, the reinforcer or how the target is taught.

What the research says

Several studies have compared scoring every trial with scoring only the first trial. Results are not identical across studies, which is why the honest answer is “it depends.”

One study of six children with autism compared two frequencies of data collection on skill acquisition and maintenance. The authors, and later reviews, point to continuous recording leading to more stringent mastery criteria and better identification of whether skills were maintained, while noting that it is more intensive to carry out.

A replication in a behavioral intervention program for children with pervasive developmental disorders compared collecting data on all trials with first-trial-only data, asking whether the decisions made from each method differ. A separate study of tact teaching with three children compared the two systems on acquisition, generalization, maintenance and on-task behavior. A thesis from Cal State notes that continuous collection is generally considered most accurate but less preferred for efficiency and ease.

Taken together, the practical lessons are consistent:

• Trial-by-trial data is the more sensitive measure.
• It costs more time and attention during teaching.
• Lighter methods can be sound when the situation calls for them, as long as they are applied consistently.

None of this means a sampled method is careless. It means the choice is a clinical decision that should be made and explained.

Choosing the method: a quick comparison

Experienced practitioners often use trial-by-trial measurement for acquisition goals and lighter methods for skills already mastered so that data collection stays consistent across the team. We agree with that approach. The goal is data that gets collected every day, not a perfect method that gets abandoned by Thursday.

Making trial-by-trial data practical

The common failure is not a bad method. It is a data system that overloads the person holding the pen. When a therapist is running dozens of trials an hour, a cluttered sheet means skipped entries, guesses written in later and numbers nobody trusts.

The AFIRM brief and practitioner experience point to the same fixes:

• Use a sheet built for the target. A simple grid with one box per trial, with room for plus, minus and prompt codes, beats a generic form. Simple, skill-specific sheets reduce the load during intense teaching.
• Define the codes before the session. Everyone on the team should score a prompted response the same way.
• Score right after the response. Waiting until the end of the block is where accuracy slips.
• Limit the number of active targets. Ten targets scored trial by trial is a lot. Rotate lighter methods onto mastered skills.
• Graph it. Data that stays in a binder does not change programs. Review graphs on a regular schedule.
• Train and check. Staff need practice and feedback, not just a handout. Our guide to behavior skills training covers how that works.

How parents can read the data

You do not need to chart anything yourself, but you should be able to follow the story. When you meet with your child’s BCBA, ask to see the graph for one or two current targets and these questions:

1. Which skills are scored on every trial, and which are sampled? Why?
2. What is the mastery criterion, and how many sessions must a child meet it?
3. Is the line moving up, flat, or bouncing? What are we changing as a result?
4. How are maintenance skills checked after they are mastered?
5. Do the numbers match what I see at home?

The last question matters most. Data from the clinic is only useful if skills show up in your kitchen, at the park and with grandparents. Parents who track a few everyday milestones at home, with a simple log or a milestone tracking tool built for first-time parents, give the team a second view of progress that session sheets cannot capture.

For more on how analysts record and summarize information, see our parent-focused pieces on how behavior analysts collect data and telling reliable progress from session noise. If you want to know what outcomes to expect over time, read what parents can expect from ABA program outcomes.

Why providers are expected to collect data

Data is not a courtesy. The BACB Ethics Code, in its section on collecting and using data, expects behavior analysts to select and implement data collection procedures correctly, display the data graphically and use it to decide whether to continue, modify or end an intervention.

For parents, that gives you a fair standard. A provider should be able to show you current data, explain what it means and describe the change it prompted. If the answer is vague, ask again. Our guide to choosing an ABA provider lists more questions worth asking.

For RBTs and new BCBAs: why this skill matters

If you are building a career in behavior analysis, data collection is a core competency, not an administrative chore. Practitioners often describe designing, running and interpreting data systems as foundational to what makes ABA evidence-based.

Practical habits worth building early:

• Learn to read a graph for level, trend and variability before you learn to design one.
• Practice scoring the same video as a colleague and compare, so you notice where your definitions drift.
• Ask your supervisor why a skill is on trial-by-trial and another on first-trial. Good supervisors enjoy that question.
• Notice when your data sheet is slowing teaching, and say so. A system that does not fit the setting is a design problem, not a personal failing.

Our perspective

We have spent more than 20 years watching what happens when data is treated as paperwork and when it is treated as a conversation with the family. The second version wins every time. Trial-by-trial data is the most detailed view of a child’s learning that we have, so we use it where detail changes decisions, especially while a child is acquiring a new skill.

We also respect the limits of a therapist’s attention. If scoring every trial pulls eyes off the child, teaching suffers and the data gets worse, not better. That is why we match the method to the skill and the setting, and why we involve parents early: your observations at home tell us whether the numbers on our sheets are meaning something in real life.

Data, to us, is a beacon of clarity. It lets a worried parent see progress that is easy to miss day to day, and it lets a clinician change course before weeks are wasted. When families ask us what they should demand from any provider, our answer is simple: show me the graph, and tell me what you changed because of it.

— Jennifer

Where Building Block Resolutions fits

Our ABA programs are individualized, with a treatment plan built around each child’s challenges and parent coaching so progress carries into home routines. If you want to understand how your child’s data will be collected and reviewed, we are happy to walk you through it. Learn more about ABA autism therapy, explore parent coaching, or see our full services list, including behavioral staff training for schools and organizations.

FAQ

What is a key advantage of trial-by-trial data?

It captures every learning opportunity, so you can see patterns within a session, such as errors clustering late in a block. Studies suggest it is a more sensitive measure of change than sampled methods. That sensitivity helps teams set stricter mastery criteria and notice when a skill is not holding.

What are the three parts of discrete trial training?

A cue or instruction, the child’s response, and a consequence. The consequence is reinforcement for a correct answer or error correction for a miss. Trial-by-trial data records the outcome of this sequence each time it is run.

Do ABA plans require data collection?

Yes. The BACB Ethics Code expects behavior analysts to collect and use data to decide whether to continue, modify or end an intervention. If you cannot see how your child’s program is being measured, ask your BCBA to show you.

Is first-trial data worse than trial-by-trial data?

Not necessarily. It is faster, and it shows how the child performs without recent prompts or practice. It can miss within-session detail, so many teams save it for maintenance or mastered skills and use trial-by-trial for new targets.

How many trials should be run in an hour?

There is no single number, because it depends on the skill, the child and the teaching format. Teams usually set a target rate for each program and adjust if data collection slows teaching. Ask your BCBA what rate they plan for your child and why.

Sources

• Evaluating Progress in Behavioral Programs for Children with Autism Spectrum Disorders Via Continuous and Discontinuous Measurement
• A Comparison of Methods for Collecting Data on Performance During Discrete Trial Teaching
• Comparison of Data-Collection Methods in a Behavioral Intervention Program: A Replication
• Comparing Continuous and Discontinuous Data Collection during Discrete Trial Teaching of Tacting
• Discrete Trial Training Brief Packet (AFIRM)
• Ethics Code for Behavior Analysts (BACB)
• Varied Treatment Response in Young Children with Autism
• Continuous and discontinuous data collection thesis, Cal State

Recommended

• The Role of Data Collection in ABA Therapy Explained
• How Behavior Analysts Collect Data: A Parent’s Guide
• 5 Steps for Clinicians: Discrete Trial Training With BCBA Oversight
• ABA Progress Tracking: Spot Reliable Data, Not Session Noise