Design and Evaluate Preference Assessments in ABA
If you’re a BCBA or clinical supervisor, you’ve likely faced this moment: a learner isn’t responding to what should be a great reinforcer, or you’re starting a new program and need to figure out which items will actually motivate them. That’s where preference assessments come in—one of the most practical, and most misunderstood, tools in behavior analysis.
This post walks you through what preference assessments are, why they matter, how to run them, and the step most clinicians skip: verifying that your preference results actually predict reinforcement.
What Is a Preference Assessment?
A preference assessment is a structured way to find out what a person likes—and to predict which items or activities might work as reinforcers. You present options, observe which ones the person chooses or engages with, and rank them from most to least preferred.
When you identify what someone genuinely prefers, you’re far more likely to find items that increase desired behavior. This speeds up learning, reduces frustration for everyone, and makes intervention feel less like guesswork.
That said, preference is not the same as reinforcement. A learner might reach for a favorite toy every time you offer it, but that doesn’t mean the toy will increase compliance with a math instruction. That verification step—testing whether the preferred item actually changes behavior—comes next. Think of preference assessment as the screening phase that narrows your options before reinforcement testing.
Why This Matters in Real Practice
Choosing effective reinforcers improves engagement and learning efficiency. A motivated learner works harder, stays on task longer, and progresses faster. Without thoughtful assessment, you might default to reinforcers that don’t move the needle—or worse, items the learner finds neutral or aversive.
The human cost of getting this wrong is real. Weak or irrelevant reinforcers mean fewer successes, less engagement, and more frustration or problem behavior. You spend more sessions with fewer results.
There’s also an ethical dimension. Preference assessments are part of person-centered, respectful practice. Rather than assuming you know what someone likes, you actually look and listen. This honors the learner’s autonomy and gives you more reliable data to build on.
The Main Methods at a Glance
Preference assessments come in several formats. Each has strengths and fits different situations.
Paired-stimulus (forced choice): Present two items side by side, the learner picks one, and you repeat with all possible pairs. Count how often each item was chosen and rank them. This method is highly accurate because it directly observes choice behavior.
Multiple stimulus without replacement (MSWO): Spread several items in a line. The learner picks one, gets brief access, then that item is removed and the remaining ones are presented again. After a few rounds, you have a rank order based on selection sequence. It’s faster than paired-stimulus and works well with many items.
Free operant observation: Give the learner free access to items and measure how much time they spend with each one. Engagement duration becomes your data. This is useful when reaching or pointing isn’t reliable, or when you want to see what naturally holds attention.
Other methods include single-stimulus presentation, multiple stimulus with replacement, and indirect assessments like caregiver interviews. Each has trade-offs in time, accuracy, and fit with the learner’s abilities.
Key Features of a Good Preference Assessment
A solid preference assessment shares several hallmarks.
It’s systematic. You follow a clear procedure, present options in a controlled way, and collect data consistently.
It produces measurable results. Whether you’re counting choices, calculating percentages, or timing engagement, you end up with numbers that rank items—not just impressions.
It’s matched to the learner’s abilities. If a child can’t reach but can look, you track gaze or pointing. If someone has limited stamina, you keep the assessment brief.
It accounts for context. Preferences shift based on time of day, recent access to items, and setting. A child who just finished snack time may not prefer food. One who’s been playing with blocks all morning may want something different. These motivating operations change what’s reinforcing right now. Good assessments are sensitive to these shifts and may need repeating.
When to Use Preference Assessments
Run a preference assessment early in program planning—before you select your primary reinforcer, design a token economy, or launch a behavior plan. This is your baseline.
Reassess when a previously effective reinforcer stops working. A toy that was golden three months ago suddenly gets ignored. That’s satiation or a genuine preference shift. Another assessment tells you what’s preferred now.
During transition periods, reassess too. New classroom, new staff member, new routine—these are moments when preferences often shift.
When a learner isn’t progressing, reassess. Low engagement, repeated unsuccessful trials, or sudden behavior changes can signal your reinforcer isn’t working. Rather than overhauling the program, start by checking whether the learner still prefers the item.
How to Run a Paired-Stimulus Assessment
Here’s what a paired-stimulus assessment looks like in practice.
Select your items. Choose 5 to 8 things based on informal observation, caregiver input, or what catches the learner’s eye. Include variety: edibles, tactile items, things that create movement or sound.
Sample each item briefly. Let the learner touch, taste, or interact with each option for a few seconds. This ensures familiarity and rules out novelty driving early choices.
Create a trial map. List all possible pairs so every item is paired with every other item exactly once. With 6 items, you’ll have 15 pairs—manageable in one or two sessions.
Present pairs one at a time. Place two items equidistant in front of the learner. Give a neutral instruction: “Pick one” or “What do you want?” If they pick one, give brief access (10–30 seconds). Alternate which item goes on the left or right to prevent side bias.
Track every choice. Record which item was picked in each pair. Count how many times each item was chosen and calculate the percentage. You now have a rank from highest to lowest preference.
The beauty of paired-stimulus is its clarity. The learner makes a direct, observable choice with little room for interpretation.
Multiple Stimulus Without Replacement: A Faster Option
If you have many items or limited time, MSWO works well.
Arrange items in a line—5 to 7 options. Offer a choice: “Pick one.” The learner selects and gets 20–30 seconds of access. Then remove that item, shuffle the remaining ones, and present the array again. Repeat until all items have been picked once.
Record the order in which items were chosen. Sum the positions for each item. The item with the lowest total sum is highest preferred—it got picked first most often.
MSWO is faster because you’re not doing every pair. It works especially well with 6 or more items in a single sitting. The trade-off is slightly less precision than paired-stimulus, but it’s plenty reliable for most clinical purposes.
The Free Operant Approach: Measuring Engagement
Sometimes you don’t ask the learner to choose—you observe.
Set up a room or area with several items available and give unrestricted access for 5 to 15 minutes. Track how much time the learner spends engaged with each item using a stopwatch or tally system.
This approach is valuable when choice behavior is hard to observe or when you want to see what naturally captures attention. Engagement duration often predicts what will work as a reinforcer.
One caveat: time spent doesn’t always equal reinforcement value. A learner might spend 10 minutes with a preferred toy, but using that toy as a reinforcer for completing a worksheet might not land the same way. That’s why reinforcement testing is essential.
The Critical Step: Verify Reinforcement
Here’s where many clinicians stumble. You’ve identified a clear top choice and you’re ready to use it as your primary reinforcer. But wait.
Preference does not guarantee reinforcement. A preference assessment tells you what the learner likes right now. A reinforcer assessment tells you whether that item actually increases a target behavior when delivered as a consequence. These are different questions.
After your preference assessment, design a brief reinforcement test. Teach a simple target behavior and use your top-ranked item as the reinforcer. Does the learner try harder? Do they engage more, complete more trials, or show faster improvement? If yes, you’ve confirmed a reinforcer. If no, try something else or adjust your contingency.
This step takes one to three sessions and saves weeks of ineffective programming.
When Preferences Change (and Why)
Preferences aren’t permanent. Satiation is the big one: too much access makes an item less powerful. Motivating operations like hunger, fatigue, or recent exposure change what’s valuable right now. Development and learning matter too, especially in younger children whose interests evolve quickly.
Context also plays a role. A toy preferred at school might be less interesting at home. Seasonal changes, new toys, peer influence—all can shift preferences.
This is why you don’t run a preference assessment once and assume you’re done. Reassess periodically: monthly or quarterly for most learners, more often if engagement drops or behavior changes. When you transition settings, change staff, or introduce new routines, a quick preference check removes the guesswork.
Common Mistakes to Avoid
Treating preference as reinforcement without testing. A child reaches for Play-Doh every time, so you assume it’s a powerful reinforcer. Then learning doesn’t improve. The preference was real, but the reinforcement effect was weak. Test first.
Presenting too many items at once. Fifteen toys creates overwhelm and random choices. Stick with 4 to 8 items for paired-stimulus, 5 to 7 for MSWO.
Ignoring motivating operations. Running an assessment right after lunch? Edibles will rank high out of novelty, not true preference. Run the assessment at a typical time and note the context.
Never reassessing. The reinforcer that worked last quarter may not work now. Check in when progress stalls or periodically by plan.
Not matching the method to the learner. If the child can’t point, adapt by observing approach, eye gaze, or engagement time. Capture preference in a form the learner can show you.
Ethical Considerations
Preference assessments involve choices, access to preferred items, and sometimes restriction of access. Ethics matter here.
Respect the learner’s choices and autonomy. The whole point is to honor what the learner values. If they show reluctance—hesitation, withdrawal, avoidance—pause. Assent is ongoing, not a one-time check-in.
Avoid coercion. Don’t arrange situations where the learner feels forced to choose between disliked items or participate in something aversive. Keep assessments low-pressure.
Be thoughtful about deprivation. If you restrict access to increase motivation, keep it brief, proportionate, and documented. If the learner becomes distressed, re-evaluate.
Protect dignity and privacy. A learner’s preferences can reveal personal information. Maintain confidentiality, use appropriate language, and be culturally sensitive.
Obtain informed consent and assent. Explain the assessment to guardians and, in developmentally appropriate terms, to the learner. Document what the assessment is for and how results will be used.
Adapting for Different Learners
Not all learners can point, reach, or make obvious choices. A child with cerebral palsy, vision loss, or minimal motor control may need adaptation.
Eye gaze or head turning can replace reaching. Present paired items and note which one the learner looks at or turns toward.
Verbal or signing responses work if the learner can say or sign item names.
Engagement time via free operant observation often works best for learners with limited deliberate choice behavior—no request or motor response required.
Behavioral approach can substitute for choice. If presenting items causes the learner to move toward, smile, or vocalize, you have a preference signal.
Match the method to the learner’s communicative and motor abilities. The goal is honest data about preference, not adherence to a rigid protocol.
Pulling It All Together
Preference assessments are a cornerstone of effective, ethical ABA. They’re not complicated, but they are powerful. A few minutes spent systematically observing what a learner prefers can reshape an entire program.
Remember the core sequence: run a preference assessment, identify likely reinforcers, verify with a reinforcement test, and implement with monitoring. Reassess periodically and whenever context or behavior changes. Protect the learner’s dignity, autonomy, and privacy throughout.
Clinicians who master this skill see faster progress, fewer behavior problems, and better engagement. They spend less time troubleshooting weak programming and more time celebrating real wins.
If you’re running a new program this week, add a quick preference assessment before selecting your reinforcer. If you’re working with a learner whose progress has stalled, check whether the original reinforcer is still effective.
