Sharon Summerall Architect Of Innovation In Applied Behavioral Science

A Practical Look At What Sharon Summerall Actually Does With Behavioral Science

I have spent more years than I care to count wading through buzzword-heavy proposals about behavioral science being applied in organizations, schools, and clinical settings. Most of it amounts to rebranding basic ABA concepts with new slide templates. The work attributed to Sharon Summerall Architect Of Innovation In Applied Behavioral Science is one of the few things I have seen that actually moves the needle past the typical consultant-level generalities. Here is what it actually looks like when you put it into practice, along with the parts that tend to trip people up. The core of this approach is not a new theory. It is a structured method for taking established behavioral principles—reinforcement schedules, antecedent manipulation, functional assessment data—and translating them into iterative, testable interventions that can be adapted across environments. The "architect" label is used because it emphasizes building the intervention systematically rather than grabbing a technique that worked once in another context and hoping it sticks. At a practical level, the process typically involves: defining the target behavior with operational precision, gathering baseline data through direct observation or existing records, identifying the function of the behavior, designing an intervention tiered to that function, implementing with fidelity tracking, and then cycling through data review to adjust. The distinguishing factor Summerall emphasizes is that each step is documented with enough detail that the intervention can be replicated or modified by someone who was not there at the start.

One thing most people miss when they first encounter this framework is how much weight it places on the measurement plan before any intervention begins. You can have a beautiful functional hypothesis and still waste weeks if your measurement does not capture the actual rate, duration, or intensity of the behavior you are trying to change. I learned that the hard way. About three years ago I was working with a program that had chronic attendance issues in a transition-age youth population. The intuitive answer was reinforcement for showing up. Everyone in the room agreed on that. We had done the functional assessment. We had a clear hypothesis that the behavior was maintained by escape from non-preferred tasks rather than access to something positive. But our initial measurement was a simple daily attendance count, which told us nothing about the relationship between absences and the specific types of classes being skipped. We spent six weeks chasing a reinforcement contingency that moved the needle by 4 percent and were about to pull the plug when I realized we were measuring the wrong thing. I switched to a time-of-day and course-type breakdown combined with a concurrent recording of presence during the first fifteen minutes of each period. That shifted the data immediately and revealed the pattern nobody had seen: absences clustered around two particular afternoon blocks, not across the whole day. The intervention got redesigned around those specific windows and we saw a 23 percent increase over the next eight weeks. If we had stayed with the original measurement plan, we would have written off the entire program as ineffective. That is the kind of precision the Summerall framework pushes for, even though it is slower upfront. The friction is real. Most teams want to start intervening because they feel the urgency of the problem. Delaying that impulse to nail down measurement details feels like doing nothing. It is not. It is the difference between guessing and knowing whether something actually works.

How To Implement This Framework Step By Step

Start with the behavior definition. I do not mean writing something like "reduce aggressive behavior." I mean specifying exactly what counts as the behavior, what does not, under what conditions it occurs, and how you will record it. A common failure point here is vague operational definitions that lead to interobserver disagreement rates above 10 percent, which invalidates any data you collect afterward. Run an interobserver agreement check before you ever design an intervention. It takes about twenty minutes and saves you from building on flawed data. Next, gather baseline. This means collecting data on the current level, frequency, and pattern of the behavior across at least five to ten data points before introducing any changes. In school or organizational settings this often means pulling existing records if they are detailed enough, or conducting direct observations if they are not. The key is consistency in conditions. Baseline data collected during a crisis period or after a staffing change is not representative and should be flagged or discarded. Functional assessment comes third. This is where the field sometimes gets sloppy. A questionnaire alone does not constitute a functional assessment. Summerall's work stresses the distinction between indirect methods, functional analysis, and descriptive assessment, and recommends using them in combination rather than relying on any single one. If you are doing this in a non-clinical setting where full functional analysis is not feasible, at minimum combine an interview-based assessment with systematic descriptives across different environments. The goal is to identify whether the behavior functions as escape, attention-seeking, access to tangibles, or automatic reinforcement.

Intervention design should flow directly from the functional assessment. This is where the "architect" part matters. You are not selecting a strategy from a menu. You are constructing one that addresses the identified function with specific contingencies, schedules, and environmental modifications. The intervention should be written out in enough detail that another qualified person could implement it without asking you clarifying questions. I keep a standard intervention template that includes target behavior, hypothesized function, antecedent strategies, consequence strategies, reinforcement schedule, data collection method, and fidelity criteria. It takes about an hour to fill out properly but prevents the drift that kills most implementations. Data collection during implementation needs to track both the target behavior and treatment fidelity. Fidelity is non-negotiable. An intervention that is not implemented as designed cannot be evaluated. I recommend a simple checklist for each session covering whether antecedent strategies were delivered, whether reinforcement was contingent on the correct response, and whether any protocol deviations occurred. If fidelity drops below 80 percent, the data are inconclusive and you need to address implementation quality before drawing conclusions.

Where This Approach Falls Apart

I should be honest about the limitations because most people presenting this framework will not. The Sharon Summerall model of behavioral science innovation requires a level of documentation and measurement rigor that most organizations are not set up to sustain. If your team is already stretched thin, adding systematic data collection on top of regular duties will either get cut first or become perfunctory, which renders the whole process useless. The framework also assumes access to people trained in behavioral methodology. That is not always available in smaller schools, community programs, or workplaces that want to apply behavioral principles without having a BCBA or equivalent on staff. In those situations, the structured approach can still help, but you will need external consultation support, which adds cost and scheduling complexity. There is also a boundary issue that does not get discussed enough. This framework is built around voluntary, ethical application in educational and organizational settings. When applied in correctional or involuntary clinical contexts, the line between behavioral innovation and behavioral control becomes morally complicated. I have seen programs use the language of evidence-based practice to justify restrictive interventions that would not survive a functional analysis review. The framework itself does not prevent that misuse. Anyone applying this work needs to be clear about consent, participant rights, and whether the behavior being targeted is genuinely harmful or merely noncompliant with institutional expectations.

If you are in a setting where full functional assessment and systematic implementation are not possible, consider starting with a simpler behavioral audit: identify one high-impact behavior, collect a week of baseline data, implement one clear antecedent strategy, and measure the result. It will not give you the same rigor, but it will teach you whether behavioral methods are worth investing more into.

What The Framework Gets Right That Other Approaches Miss

Most behavioral training programs stop at teaching techniques. The Summerall framework insists that technique selection is secondary to process discipline. You can apply reinforcement perfectly and still fail if you are reinforcing the wrong thing or measuring the wrong outcome. That emphasis on process over product is counterintuitive for people who come from fields where quick wins are valued over systematic validation. Another nuance that beginners overlook is the role of stimulus control. Functional interventions often work initially and then plateau because the behavior generalizes poorly or loses sensitivity to the intended antecedents. Building in explicit stimulus discrimination training from the start, rather than treating it as an afterthought, tends to produce more durable results. I use a simple rule: if the intervention relies on a cue or prompt to work, plan for fading that cue from day one, not after the behavior has changed. The emphasis on replication also matters more than it gets credit for. Behavioral science as a field has a replication problem, and much of the applied literature rests on single-case studies that are never tested in different settings. The architectural approach treats each intervention as a prototype, not a final product. That mindset shift helps practitioners stay open to modification rather than defending a plan that stopped working.

There is no download link or software package associated with this framework because it is not a tool you install. It is a set of practices and documentation standards. The closest thing to a resource I would point you toward is the published work and training materials associated with Summerall's presentations and papers on applied behavioral science innovation, which tend to be available through academic and professional behavioral psychology channels. The bottom line is that this approach will slow you down at first. It will also make whatever you build harder to defend as anything other than luck. That tradeoff is worth it if you intend to keep doing this work rather than running one-off interventions and moving on.