Question: Why are prompts “evil?” Do your concerns with prompts extend to clinical work with typically-developing individuals?
Answer:
The trouble with prompts
Prompting as a strategy to establish behavior is ubiquitous in much of ABA practice. Behavior under the control of one stimulus is used to try to establish control by another stimulus. This is often implemented by pairing an existing stimulus control relation with a different stimulus. Another method is to compel responding, via a guided correction if a learner does not respond as requested. These procedures often fail or take many repeated trials to achieve the targeted outcome. Fading prompts are often exercises in sharpening control by less and less of the prompted stimulus.
In the case of the pairing procedure basic research in stimulus control suggests that success may be preempted by a phenomenon known as blocking. A stimulus that controls behavior, when combined with another that does not, has been shown to prevent, or block the new stimulus from gaining control (see Olaff, et al., 2022 for implications for autism interventions). This has been shown to occur across species. But hasn’t there been laboratory examples of the transfer of stimulus control? Yes, but these are painstakingly defined relations that involve careful manipulation of both SD and S-Delta stimuli combined with stimulus change (that disrupts current control. Terrace, 1963; Touchett, 1971) that is mostly absent in applied applications. Moving a hand or other prompts, when kinesthetic feedback from movement is not the targeted stimulus control, may produce a particular behavior under context, but extension beyond that context is unlikely, and the learner may resist further instruction.
The failure of prompting became obvious as programed instruction evolved in the 1960s. First, formal prompts were discouraged, then recency and sequence prompts, eventually followed by thematic prompts (Markle, 1969). It was shown that instead of facilitating responding prompts inhibited the stimulus control the program was trying to establish. Judy Doran and Jim Holland (JABA, 1971) using eye movement measures demonstrated that learners skipped reading text in which critical information was presented if a prompt controlled responding. Prompting was training learners to ignore other important textual stimuli. By 1990 all prompting was discouraged (Markle, 1990).
Establishing targeted stimulus control is best served by procedures that make the control more likely from the outset. Some use the term stimulus prompting, but it is the controlling stimulus control relation that should be the focus. A stimulus control relation, (AKA stimulus control topography) must first occur before it can be reinforced (Ray & Sidman, 1970). Reinforcing a behavior in the presence of a stimulus does not insure such control. Accordingly, the job is to “prime” (as one primes a pump) the controlling relation from the outset. There are a range of procedures for doing so. One involves ensuring the learner observes the actual stimulus or stimulus feature targeted. Another is to make sure the response itself can be performed fluently. A programed instruction example is presented below.
In the presence of left and right handles on a faucet, the learner will turn the right hand knob to obtain warm water. Can they discriminate left from right, warm and cold? Can they reach, grasp, and turn fluently? Those are all components of the discrimination that need to be fluently in place before teaching the relation. So the question becomes. What are the stimulus elements that need to be established and what are their components, and what are the response elements that are required and that are their components? Teach those first prior to establishing the targeted relation. We can then set up discrimination exercises using juxtaposed stimuli, and we may highlight dimensions to be used, or make responding along the response vector more likely. Here is a link to the errorless establishing of conditional chain. Notice how elements are used to establish directional control. Also notice how past history is recruited, pushing a wheeled object, which primes directional movement. Look at how earlier conditional control is established. This is all achieved without prompting a response that is controlled by a stimulus that is orthogonal (differs greatly from) to the target repertoire.
Bibliography
Markle, S. M. (1969). Good frames and bad: A grammar of frame writing: 2nd Edition. NY: Wiley.
Markle, S. M. (1990). Designs for instructional designers, Seattle, WA: Morningside Press.
Olaff, H.S., Vandbakk, M. & Holth, P. Blocking of Stimulus Control in Children with Autism. Psychol Rec 72, 305–317 (2022). https://doi.org/10.1007/s40732-020-00454-7
Ray, B. A., & Sidman, M. (1970). Reinforcement schedules and stimulus control. The theory of reinforcement schedules, 187-214.
Terrace, H. (1963). Errorless transfer of a discrimination across two continua. Journal of the Experimental Analysis of Behavior, 6, 223-232.
Touchette P. E. (1971). Transfer of stimulus control: measuring the moment of transfer. J Exp Anal Behav.,15(3): 347-54. doi: 10.1901/jeab.1971.15-347.