The big idea
Reinforcement only works if the consequence is something the person values at that moment. A preference assessment is a systematic way to identify items or activities a person is likely to choose and to rank them. It replaces guessing (“they like stickers”) with data.
Preference is not the same as reinforcement. A preference assessment tells you what someone selects or engages with. Whether that item increases behavior when it is delivered contingently is a separate question, answered by a reinforcer assessment. Preferences also shift with motivating operations (B.16), so assessments are repeated, sometimes daily or before each session.
Breaking it down
Indirect methods
Ask the client, caregivers or staff. Interviews and surveys, such as the Reinforcer Assessment for Individuals with Severe Disabilities (RAISD), produce a list of candidate items quickly. Their weakness is accuracy: what people report often differs from what the client actually chooses. Use them to decide what to test, not as the final word.
Direct methods
| Format | How it works | Strengths | Limitations |
|---|---|---|---|
| Single-stimulus (Pace et al., 1985) | One item at a time; record approach or engagement | Works for clients who cannot scan or choose among items | No ranking; may show interest in nearly everything |
| Paired-stimulus (Fisher et al., 1992) | Two items per trial; every item is paired with every other; rank by percentage chosen | Clear ranking; predicted reinforcers better than single-stimulus | Slow as items are added; removing items can occasion problem behavior |
| Multiple-stimulus with replacement (MSW) | An array of items; the chosen item is returned and the array is presented again | Quick; shows the top item | Often identifies only one or two favorites |
| Multiple-stimulus without replacement (MSWO; DeLeon & Iwata, 1996) | An array; the chosen item is removed; repeat until all are chosen | Full ranking in less time than paired-stimulus | Requires scanning and choosing among several items; awkward with large activities |
| Free-operant (Roane et al., 1998) | Items are freely available for a set time; record duration of engagement with each | Nothing is removed, so problem behavior is less likely; suits activities | May not rank items if the client stays with one |
Designing a good assessment
- Start with about five to eight candidates from interviews and observation, including items from the client’s culture and home (F.2).
- Let the client briefly sample each item first so they know what it is.
- Rotate positions in arrays so choices reflect the item, not where it sits.
- Consider separate assessments for foods and leisure items, since foods are often chosen over leisure items when the two are mixed.
- Fit the format to the client: scanning skills, problem behavior when items are removed, and the time available.
- Reassess often. Satiation and deprivation change what works.
Reinforcer assessment
To test whether a preferred item works as a reinforcer, make it contingent on a simple response and compare responding with a baseline or with other items. Common arrangements include concurrent schedules (two responses, each producing a different item), single-operant arrangements, and progressive-ratio schedules, which show how much work a client will do for an item before stopping. A highly preferred item that does not increase responding is not functioning as a reinforcer for that response under those conditions.
Seeing it in practice
- Ivy, 4, clinic. Her therapist needs a ranked list of toys for a new teaching program. Ivy can scan a row of items and shows no problem behavior when items are taken away. An MSWO with six toys, run three times, gives a ranking in about 10 minutes.
- Santiago, 11, school. Santiago screams when a chosen item is removed. His teacher runs a free-operant assessment during recess and records how long he engages with each activity. Nothing is taken from him.
- Gwen, 52, day program. Gwen has low vision and does not reach for items in an array. Staff present one item at a time and record whether, and for how long, she engages with it.
- Checking the result. The MSWO ranks bubbles first for Ivy, but her correct responding does not rise when bubbles follow it. A reinforcer assessment shows that a short video, ranked third, produces more responding, so the program uses the video.
Common mix-ups
- Preferred is not proven. Selection shows preference. Only a change in behavior when the item is delivered contingently shows reinforcement.
- MSW vs. MSWO. With replacement, the chosen item goes back, so the same favorite can win every trial. Without replacement, it is removed, which forces choices among the rest and produces a ranking.
- Paired vs. single. Paired-stimulus means two items per trial, with every possible pair presented. Single-stimulus means one item per trial and no choice between items.
- One assessment is not forever. Preferences change with time, satiation and setting. Brief, frequent assessments keep reinforcers effective.
- Caregiver report is a starting point. It is useful for choosing what to test, but it can miss or overestimate preferences.
Exam tip: Match the format to the constraint in the scenario. Problem behavior when items are removed points to free-operant. A client who cannot choose among several items points to single-stimulus or paired-stimulus. A need for a quick, complete ranking points to MSWO.
Quick recap
- Preference assessments identify and rank items; reinforcer assessments test whether items increase behavior.
- Indirect methods generate candidates; direct methods test them.
- Single-stimulus: one item at a time. Paired-stimulus: every pair. MSW: chosen item returned. MSWO: chosen item removed. Free-operant: free access, engagement timed.
- Choose the format by the client’s skills and behavior and the time available, and reassess often.
Further reading
- Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992). A comparison of two approaches for identifying reinforcers for persons with severe and profound disabilities. Journal of Applied Behavior Analysis, 25(2), 491-498.
- DeLeon, I. G., & Iwata, B. A. (1996). Evaluation of a multiple-stimulus presentation format for assessing reinforcer preferences. Journal of Applied Behavior Analysis, 29(4), 519-533.
- Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, B. A. (1998). Evaluation of a brief method for stimulus preference assessment. Journal of Applied Behavior Analysis, 31(4), 605-620.
Key terms
- Preference assessment (Stimulus preference assessment, SPA)
- A systematic procedure for identifying items or activities a person is likely to choose and, usually, ranking them, in order to find likely reinforcers. Methods range from asking people to directly observing the person's choices or engagement.
- Single-stimulus preference assessment (Successive-choice preference assessment)
- A direct preference assessment in which items are presented one at a time and the observer records whether the person approaches or engages with each. It does not require choosing among items, but it gives no ranking and may show interest in nearly everything.
- Paired-stimulus preference assessment (Forced-choice preference assessment)
- A direct preference assessment in which items are presented two at a time, with every item paired with every other item, and the person chooses one. Items are ranked by the percentage of trials on which each was chosen (Fisher et al., 1992).
- Multiple-stimulus with replacement preference assessment (MSW)
- A direct preference assessment in which several items are presented in an array, the person chooses one, and the chosen item is returned to the array before the next trial. It is quick but often identifies only the most preferred item.
- Multiple-stimulus without replacement preference assessment (MSWO)
- A direct preference assessment in which several items are presented in an array, the person chooses one, and that item is not returned, so each later choice is made among the remaining items. Repeating the process produces a ranking in less time than a paired-stimulus assessment (DeLeon & Iwata, 1996).
- Free-operant preference assessment
- A direct preference assessment in which several items or activities are freely available for a set period and the observer records how long the person engages with each. Because nothing is removed, it is less likely to occasion problem behavior (Roane et al., 1998).
- Reinforcer assessment
- A direct test of whether a stimulus functions as a reinforcer. The stimulus is delivered contingent on a response, and responding is compared with a condition in which it is not delivered. It confirms what a preference assessment can only predict.
Check yourself
5 questions on F.4. Choose an answer to see whether you're right and why.
Show the answer
Answer: B. In a free-operant assessment nothing is removed, so it avoids the event that sets off Lena's hitting. The paired-stimulus and multiple-stimulus formats all involve removing items after brief access, which would occasion problem behavior on many trials.
Show the answer
Answer: A. Returning the chosen item to the array before the next trial is the defining feature of MSW, which often shows only the top one or two items. In MSWO, the chosen item is removed, so later choices are made among the remaining items and a full ranking emerges.
Show the answer
Answer: A. Preference shows what a person chooses; reinforcement is shown only when contingent delivery increases behavior. Here it did not, so puzzles are not working as a reinforcer for this response. The preference result may still be accurate.
Show the answer
Answer: A. Because each trial forces a choice between two items, the paired-stimulus format ranks items and has predicted reinforcers better than single-stimulus presentation. It takes longer, not less, as items are added, since every pair must be presented.
Show the answer
Answer: A. Repeated access to the same item reduces its value for a while, which is satiation, a motivating operation. That is why preference is reassessed often. The morning assessment could have been accurate at the time it was done.
Adds a step to your study record. Saved in this browser only.