By Matt Harrington, BCBA · Behaviorist Book Club · April 2026 · 12 min read

The Cool vs. Not Cool Procedure: Empirical Foundations and Clinical Applications

Overview & Clinical Significance

Social skill deficits are among the most functionally impactful challenges associated with autism spectrum disorder. While many social skills programs teach individual behaviors in structured formats, the generalization of those behaviors to naturalistic social contexts remains elusive in a large proportion of cases. A core problem is that many interventions focus on skill acquisition without building the discrimination necessary for a learner to judge whether a behavior is contextually appropriate.

The Cool vs. Not Cool procedure, first empirically evaluated in the study described by Dr. Justin Leaf, addresses this discrimination deficit directly. Rather than teaching a learner simply to perform a target behavior, Cool vs. Not Cool teaches the learner to evaluate whether a behavior is socially acceptable — contextually good (cool) or contextually inappropriate (not cool) — before engaging in it. This metacognitive or self-regulatory component is what distinguishes the procedure from traditional social skills instruction.

The practical significance is substantial. For individuals with autism, the ability to discriminate socially appropriate from inappropriate behavior is foundational to peer acceptance, community participation, workplace success, and personal safety. A learner who can perform a polite greeting in a structured role-play but cannot tell the difference between an appropriate and inappropriate approach to a peer in the playground has not acquired functional social competence. The Cool vs. Not Cool procedure is designed to build that discrimination, and this first empirical evaluation gives practitioners a foundation for using it with confidence and understanding its parameters.

Background & Context

The Cool vs. Not Cool procedure was developed within the progressive ABA tradition as a naturalistic and socially meaningful way to teach contextual discrimination of social behavior. The procedure involves presenting examples of target behaviors in either a socially appropriate or socially inappropriate context — the cool and not cool conditions — and using differential reinforcement to teach the learner to discriminate between them and label the exemplar accordingly. Video examples are commonly used in clinical implementation, allowing the learner to observe the behavior from a third-person perspective before practicing in vivo.

The procedure draws on several established behavioral principles. Stimulus control is central: the learner is being taught to bring a verbal label (cool/not cool) under the control of contextually relevant stimulus features rather than simply the presence or absence of a target response. Generalization is addressed through the use of multiple exemplars across varied social contexts, which supports transfer of the discrimination to novel situations not included in training. Video modeling provides a controlled way to present exemplars consistently without requiring live social interactions during the teaching phase.

From a verbal behavior perspective, the procedure develops a specific type of verbal operant — the learner is being taught to tact behavioral exemplars using evaluative labels. This tact-like response under contextual stimulus control is a prerequisite for more complex social reasoning. The historical context for this procedure exists within a growing recognition that rule-governed behavior and perspective-taking skills are integral to social competence and should be explicitly taught rather than assumed to emerge naturally from behavioral fluency.

Clinical Implications

The clinical implications of the Cool vs. Not Cool procedure extend across the full range of social skills targets commonly addressed in ABA programs. The procedure is applicable to greeting behaviors, conversational turn-taking, personal space management, appropriate touch, emotional regulation responses, and a wide range of other targets that have a contextual component.

One important implication is that social skills instruction should include a discrimination component for virtually every target. Teaching a learner to initiate a conversation is not sufficient if the learner cannot tell when initiation is appropriate versus when it is likely to be unwelcome. Incorporating a Cool vs. Not Cool component into existing social skills targets can extend their functional utility without requiring entirely new program development.

The video modeling component of the procedure has implications for resource development. Creating high-quality video exemplars for cool and not cool scenarios requires collaboration between the clinical team, caregivers, and often peers or staff who can serve as actors. This investment pays dividends across multiple learners if a library of video examples is maintained and organized by skill domain. For agencies serving multiple autistic clients, a shared video resource library for Cool vs. Not Cool implementation is a practical efficiency.

For learners who are beginning to participate in less structured or inclusive settings — general education classrooms, community recreation programs, workplace environments — the Cool vs. Not Cool discrimination is particularly urgent. The natural environment will not provide the clear prompts and reinforcement contingencies of a structured ABA session, making self-regulatory skills all the more necessary. BCBAs working on transition planning should consider explicitly programming Cool vs. Not Cool as a preparation for naturalistic settings.

Ethical Considerations

Teaching social discrimination involves navigating a complex ethical landscape. Defining what is cool and not cool necessarily reflects social norms, and social norms vary across cultures, communities, and contexts. BCBAs working with diverse populations must be thoughtful about whose norms are being taught and whether those norms reflect the learner's community and values rather than a narrow mainstream standard. Code 1.05 (Non-Discrimination) is relevant here, as practitioners must ensure that social skills instruction is culturally sensitive and not inadvertently pathologizing cultural practices that differ from the clinician's own cultural frame.

Code 2.01 (Providing Effective Treatment) supports the use of evidence-based procedures, and this first empirical study gives practitioners a basis for using the Cool vs. Not Cool procedure. At the same time, practitioners should be aware of the current state of the evidence and communicate honestly with caregivers about what is known and what remains to be studied, consistent with Code 2.03 (Informed Consent).

The social validity of the targets selected for Cool vs. Not Cool instruction is also an ethical concern. BCBAs are obligated to select goals that are meaningful to the client and their family. Involving caregivers and the learner themselves — to the extent possible — in defining what cool and not cool look like across relevant settings ensures that the program reflects genuine priorities rather than the clinician's assumptions about social acceptability.

Finally, Code 2.15 (Least Restrictive Procedures) applies to the selection of teaching methods for social skills. The Cool vs. Not Cool procedure is largely positive and naturalistic, which is consistent with this principle. However, practitioners should ensure that video examples and feedback during training are implemented in ways that are respectful of learner dignity and do not create shame or embarrassment around the not cool examples.

Assessment & Decision-Making

Before implementing the Cool vs. Not Cool procedure, practitioners should complete a thorough assessment to identify appropriate target behaviors and develop valid stimuli for instruction. The starting point is a social skills assessment that identifies the specific behaviors the learner needs to discriminate — either behaviors they currently exhibit in inappropriate contexts, behaviors they fail to exhibit in appropriate contexts, or behaviors they are beginning to learn and for which contextual discrimination is part of the instructional objective.

The stimulus development phase involves creating or identifying video exemplars that clearly illustrate cool versus not cool versions of the target behavior. These exemplars should vary across actors, settings, and contextual features while keeping the target behavior consistent across conditions. The ability of the videos to evoke reliable discrimination in a normative audience should be verified before use in instruction — show exemplars to staff or caregivers and confirm that there is agreement about whether each example is cool or not cool.

Baseline probes should be conducted before instruction begins to assess the learner's ability to label or discriminate target behaviors in the relevant contexts. This serves both as a pre-measure for evaluating change and as a task analysis of what the learner needs to learn. Some learners may have the discrimination for some targets but not others, allowing for efficient prioritization of instruction.

Ongoing data collection during instruction should track the percentage of correct cool or not cool judgments across stimulus examples. Probes with novel exemplars that were not part of training should be conducted regularly to assess generalization of the discrimination. Maintenance data should be collected at defined intervals after mastery to confirm that the discrimination is durable over time.

What This Means for Your Practice

The Cool vs. Not Cool procedure represents a meaningful advance in social skills instruction for autistic individuals because it addresses a dimension of social competence that behavioral approaches historically underemphasized: the evaluative component of social behavior. By teaching learners to discriminate and label contextually appropriate versus inappropriate behavior, the procedure builds a foundation for self-monitoring that can support generalization far beyond what rehearsal-only approaches produce.

For BCBAs, the practical implication is to examine existing social skills programs and identify which targets include a contextual discrimination component and which do not. For targets that lack this component, consider whether adding a Cool vs. Not Cool phase would increase the functional utility of the skill. This is not about overhauling programs but about strategically adding a component that extends the reach of instruction.

For agencies, investing in the development of a video exemplar library for Cool vs. Not Cool instruction is a high-leverage resource decision. Well-designed videos can be used across multiple learners, can be reviewed and updated over time, and can be standardized to ensure that what is being taught reflects agreed-upon community norms for the population served.

For caregivers, the Cool vs. Not Cool procedure offers a simple and intuitive framework for discussing social behavior at home and in the community. Training caregivers to use the cool and not cool language in natural settings — as a teaching prompt, not a punitive label — can extend the procedure's effects beyond clinic hours and into the contexts where generalization matters most.