Best Available Evidence in ABA, Explained

Best available evidence is the research leg of evidence-based practice. It is the body of studies that tells us what tends to work. When you pick a treatment, this is the science you lean on. It sits next to two other legs, clinical expertise and client values, but this page stays on the research part.

Why does it matter? Because not all research is equal. Some studies are strong. Some are weak. Some fit your client well, and some do not. You will rarely find one perfect paper. Best available evidence means weighing what exists and using the strongest of it right now.

What counts as best available evidence#

The word "evidence" is bigger than most people think. In her talk on the three-legged stool, Dr. Jessica Osos lays out four main types. There are group designs, like studies that split people into groups. There are single-case designs, which track one person over time. There are systematic reviews, which pool many studies together. And there are practice guidelines built by expert panels.

All four count. None of them is the only answer. A single strong study does not settle a question. And a weak study still adds a small piece. A board certified behavior analyst, or BCBA, is a clinician certified in applied behavior analysis (ABA), the science of behavior change. The BCBA's job is to look across these types and judge them together.

Relevance and certainty: the two tests#

Osos gives a simple frame for judging evidence. Every study gets weighed on two things. The first is relevance. That means how well the study matches your actual client. The second is certainty. That means how solid the study's methods are.

A study can be high on one and low on the other. A tight lab study might be very certain but not relevant to your client. A loose real-world study might be very relevant but less certain. The best evidence passes both tests at once.

In the talk, Osos puts it simply. The best evidence sits at "the intersection of both of these things."

That word "intersection" is the key. You are looking for the overlap. High relevance and high certainty, together, is the goal.

The hierarchy of evidence#

People often draw evidence as a pyramid. Expert opinion sits near the bottom. Single studies sit in the middle. Systematic reviews and meta-analyses, which pool results from many studies, sit near the top. This ranking is about certainty, meaning how much we can trust the method.

But a hierarchy is a guide, not a rule. A high-ranking review can still miss your client type. A lower-ranking study can still fit your case well. So you use the hierarchy to start, then you check relevance. Rank tells you how sure to be. Relevance tells you if it even applies.

This is why "best available" is honest wording. Sometimes the top of the pyramid has nothing on your exact question. Then you move down and use the strongest thing that does exist. You work with what is there.

Single-case design is real evidence#

In medicine, the randomized controlled trial gets treated as the gold standard. A randomized controlled trial is a study that splits people into groups by chance. ABA does not always work that way, and that is fine. Our field grew up on single-case methods, and those methods are strong.

Single-case design tracks one learner closely across time. It changes one thing, watches the effect, and repeats to show the pattern is real. When done well, it gives clear proof that the treatment, and not luck, caused the change. Osos is direct on this point. She says the field has a long history of strong single-case methods. That body of work is a real part of the evidence base.

So do not discount a single-case study just because it has few participants. The rigor lives in the design, not the head count. A well-run single-case study can carry real weight.

Systematic reviews and practice guidelines#

At the higher end of certainty sit systematic reviews and meta-analyses. A systematic review gathers all the studies on a question and reads them by a fixed method. A meta-analysis goes further and pools the numbers into one estimate. These tools cut down on cherry-picking. They show the whole picture, not one lucky result.

Practice guidelines are the next step. Expert panels read the reviews and turn them into plain guidance. They tell you what to try first and what to avoid. Guidelines save you time, but they are still built on the studies underneath. When the studies are thin, the guidance is softer. Good clinicians read the guideline and the evidence behind it.

Weighing evidence for your actual client#

Here is the part that ties it together. Evidence is one input. You still make the final choice for each client. This is where the research leg meets your judgment and the client's goals.

Ask two questions for each study. Does this fit a client like mine? And how sure can I be in the result? A study on adults may not fit a young child. A study in a clinic may not fit a home. When the match is weak, you lower your confidence and watch your own data closely. Best available evidence starts the plan. Your ongoing measurement keeps it honest.

What the research says#

Three published articles inform this page. None of them had a full citation on file, so they are named here by title and finding.

One source is a systematic review and meta-analysis on exercise training and obesity-related measures in people with intellectual disability. It shows how the top of the evidence pyramid works in practice. The authors searched many databases, applied fixed rules, and scored each study's methods with a quality scale. That is certainty done well, and it is the kind of synthesis best available evidence is built on.

A second source is an introduction to a special issue on community living and participation. It ties the phrase "state-of-the-science" to rigorous review methods like systematic reviews and meta-analysis. A third source, an editorial on autism and global mental health, shows the relevance side of the coin. It warns that most autism knowledge comes from a small slice of the world's population. That is an external validity problem, meaning how well findings apply beyond the studied group. Both sources remind us that strong methods still need to fit the people in front of us.

Frequently asked questions#

Is best available evidence the same as evidence-based practice?

No. Evidence-based practice has three legs. Best available evidence is only the research leg. The other two are clinical expertise and client values. You need all three working together. This page focuses on the research part alone.

Does a small study ever count as good evidence?

Yes. Study size is not the only measure of quality. A well-run single-case study can give strong proof for one learner. It controls the design so the treatment, not chance, explains the change. Judge the method first, then the fit to your client.

What do I do when there is no research on my exact case?

You move down the hierarchy and use the strongest evidence that does apply. That is what "best available" means. Pick the closest solid study, lower your confidence a bit, and lean harder on your own data. Watch the client's progress and adjust as you go.

Want the full framework from the source talk? Watch From Research to the Applied Setting: Breaking Down the Three-Legged Stool of "Evidence Based Practice" with Dr. Jessica Osos and earn your CEU (continuing education unit).