A preference assessment is a structured way of figuring out what a person actually likes, so those items or activities can be used as motivators during teaching. In Applied Behavior Analysis, the entire success of a program depends on finding things that genuinely motivate the learner. A reward only works if the person wants it, and preference assessments replace guesswork with direct observation of what someone reaches for, engages with, or chooses repeatedly.
This matters more than it might seem. What a parent or therapist assumes a child enjoys doesn’t always match reality, and preferences shift over time. A toy that worked as a motivator last month might bore someone today. Preference assessments give practitioners a reliable, updatable picture of what’s reinforcing right now.
Why Preferences Matter for Learning
A reinforcer is anything that, when delivered right after a behavior, makes that behavior more likely to happen again. If a child correctly labels a picture and then gets access to a favorite toy, they’re more likely to label pictures in the future. But the toy only has that effect if the child genuinely wants it in that moment. The effectiveness of both skill-building and behavior-reduction programs hinges on identifying potent reinforcers.
There’s also an ethical dimension. Choice is a central principle in ABA services. A learner should have alternatives, be able to act on those alternatives, and experience the natural result of their choice. Preference assessments formalize that principle by centering the learner’s own selections rather than imposing what someone else thinks they should enjoy.
Types of Preference Assessments
There are five main formats, each suited to different learners and situations. They range from simple observation to structured choice trials.
Free Operant Observation
This is the starting point when you don’t yet know what a person likes. The learner is given free access to an environment with various items, and an observer simply watches what they gravitate toward and how long they engage with each thing.
A naturalistic version happens in a typical, everyday setting. A teacher might set aside 15 minutes to quietly observe a new student in the free play area, noting which toys or activities the child approaches on their own. A contrived version involves intentionally arranging a set of items within reach before the session begins. This is especially useful for children with limited mobility or when potentially reinforcing items, like a tablet or a water table, aren’t normally available in the space.
One notable advantage: items are never taken away after the learner selects them. This makes free operant observation a good fit for children who engage in challenging behavior when preferred things are removed. Engagement is measured by how much time the person spends with each item, not by a forced choice between options.
Single Stimulus
Items are presented one at a time, and the observer records whether the learner approaches or engages with each one. This is the simplest structured format. It works well for learners who don’t yet consistently choose a favorite when given two options side by side, or for individuals with limited motor or communication skills who might struggle with more complex presentations.
Paired Stimulus
Every item is paired with every other item across multiple trials, and the learner picks one from each pair. This produces a clear ranking because each item is directly compared against all alternatives. It’s thorough and considered highly accurate, but it takes more time than other methods. A learner needs the skill of consistently choosing a preferred item when two options are in front of them.
Multiple Stimulus With Replacement (MSW)
Several items (often five to seven) are presented at once. The learner selects one, gets brief access to it, and then all items, including the one just chosen, go back into the array for the next trial. Because the same item keeps reappearing, a learner can pick the same thing every time, which is useful information in itself.
This format works when a learner can scan and choose from three or more options but tends to have challenging behavior when favorite items are taken away. Since everything stays in the array, nothing is “lost” between trials.
Multiple Stimulus Without Replacement (MSWO)
This follows the same setup, but once the learner selects an item, it’s removed from the array. The remaining items are rearranged, and the learner picks again. This continues until all items have been selected or the learner stops engaging. The result is a ranked hierarchy from most to least preferred.
The MSWO is one of the most commonly used formats in practice because it’s efficient. Research published in the Journal of Applied Behavior Analysis found that the time needed to administer an MSWO was less than half the time required for a paired stimulus assessment, while still producing a useful preference ranking. It’s appropriate for learners who can choose from multiple options and who don’t become upset when a selected item is removed.
Choosing the Right Assessment
The right format depends on the learner’s current skills, their communication abilities, and their age. Practitioners typically work through a decision process: if you know nothing about the person’s preferences, start with a free operant observation. If the learner can’t reliably choose between two options, use a single stimulus approach. If they can choose between two but not three or more, the paired stimulus format fits. For learners who handle multiple options, the choice between MSW and MSWO comes down to whether removing items triggers problem behavior.
This isn’t a one-time decision. Preferences change, sometimes dramatically. An item that ranked first two weeks ago might drop to third. Many practitioners re-run brief preference assessments at the start of sessions or at regular intervals to make sure they’re working with current information rather than outdated assumptions.
How Results Get Used
The output of a preference assessment is a ranked list of items or activities. The top-ranked items become candidates for reinforcers during teaching sessions. “Candidates” is the right word because preference and reinforcement aren’t the same thing. A child might consistently choose a spinning top during the assessment, but it only counts as a reinforcer if using it after a correct response actually increases that response over time. Preference assessments identify what someone wants; reinforcer assessments confirm whether those items change behavior.
In practice, the top two or three items from an assessment are typically rotated during sessions. Rotating prevents satiation, where the learner gets tired of the same reward and it stops being motivating. Lower-ranked items aren’t discarded either. They can serve as reinforcers for easier tasks, or they can be paired with social praise to build new sources of motivation over time.
Preference assessments also reveal what someone doesn’t engage with, which is equally useful. If a learner consistently ignores certain items, those can be removed from the reinforcer pool rather than wasting instructional time offering things that don’t motivate.
Preference Assessments Beyond Young Children
While much of the research and training examples focus on young children, preference assessments are used across ages and settings. For older individuals or those with strong verbal skills, a simple interview or survey asking what they enjoy can serve as an initial screen, though direct observation assessments tend to be more accurate than self-report alone. People sometimes say they like things they think they should like, or they forget about items they’d actually engage with if given access.
For nonverbal individuals or those with significant disabilities, the structured formats become essential. A person who can’t tell you what they want can still reach for it, look at it, or spend time with it. Preference assessments translate those observable behaviors into actionable information, giving practitioners a way to honor someone’s choices even when traditional communication isn’t available.