Using Chartlytics for navigating social behavior outcomes

10/20/2015
 

Problem behavior requires the best applied science available - applied behavior analysis (ABA). ABA developed from B. F. Skinner and his experimental work with animals in tightly controlled laboratory settings. 

Picture
 

Figure 1: The man himself B. F. Skinner.

Skinner observed that animals (e.g., pigeons) responded in lawful ways to a particular set of events. For example, when a light came on, if the pigeon pecked a disk, the pigeon immediately received a food pellet. In the future, when the pigeon entered the experimental chamber it would behave in a similar manner -when the light came on, the pigeon pecked the disk and received food. Skinner called his discovery positive reinforcement. 

It turns out all the experimental findings Skinner and his colleagues made in the laboratory also applied to humans. Many people have heard of “time out.” But the time out procedure came straight out of the experimental chamber and works on animals the same the way it works on humans. Of course that assumes people actually do time out correctly! Sadly some people use distorted procedures of time out which rarely work as intended.

People also warped Skinner’s advice on using punishment; he adamantly warned against it! Skinner and his colleagues discovered what punishment procedures do. As a result of seeing the harmful effects, Skinner spent his entire career encouraging people to use positive reinforcement and other non-punitive measures to change behavior.

ABA grew from its humble laboratory beginnings as many practitioners wanted to harness the powerful science of behavior (i.e., Applied Behavior Analysis) to better humankind. Flash forward to today and an embarrassment of behavior change riches has now made its way into journals, parenting magazines, conferences, classrooms, board rooms and even popular media.

Old School Experimental Behavior Analysts used Standard Graphics

The methods of ABA rest on careful observation and experimentation. Therefore, many people call behavior analysis the science of behavior. But like any science, progress depends on the quality and precision of a measurement system. 

B. F. Skinner and his contemporaries had the advantage of a sensitive monitoring device providing standard real time data. Everyone who used a “cumulative recorder” would see data forming distinctive patterns. Just like going to a doctor who interprets standard views of heart activity on ECGs, Skinner and his crew quickly deduced what specific change patterns meant.

Skinner loved standard graphs and said the following, “… the curve revealed things in the rate of responding, and in changes in that rate, which would certainly otherwise have been missed” (Skinner, 1956, p. 225). Without a standard view and a standard data metric called frequency (or rate), ABA would not exist today.

New School Applied Behavior Analysts used Nonstandard Graphics

Along the way to developing ABA almost all of the practitioners and experimenters abandoned standard graphics like the cumulative record in favor of nonstandard graphics like the nonstandard linear graph. If that sounds like one of those slasher flicks where the guy slowly walks through a rickety old house with creepy music playing, you can guess the outcome. Once everyone started making their own graphs, everyone started making them differently (nonstandard means exactly that, no standards in graph construction).

However, one of Skinner’s graduate students (who later would go on to make amazing contributions) Ogden R. Lindsley took the lessons he learned from his dear ole mentor and created a standard visual display called the Standard Celeration Chart (SCC). What Lindsley offered the world, and especially behavior analysts, may one day completely change how people measure and view data.

Picture
 

Figure 1. A Standard Celeration Chart created by Ogden Lindsley.

Lindsley, like the other practitioners of ABA, wanted to change the world for the better. He created a standard ratio chart called the Standard Celeration Chart. The following sections describe how practicing behavior analysts can benefit from monitoring social behaviors in need of change (e.g., calling out in class, hitting others, stealing food, self injurious behaviors) on the SCC.

The Scope of Features

The following infographic displays a number of visual features offered by Standard Celeration Chart. The discernible properties of behavior appear to SCC users helping elevate the practice of ABA. Those using nonstandard linear graphics not only have none of the visible characteristics of behavior available to them, but must contend with the chaos of nonstandardization (widely variable sizes of axes, scalings, scale labels, and graphical symbols).

Picture
 

Figure 2. Twenty-six visual features of behavior provided by the SCC.

While covering all of the visual features falls beyond the scope of the present blog post, I will highlight a few essential for ABA.

Real time charting. When we visit the doctor with either an injury or illness, we ask the doctor how long the treatment will take. We all want to know how long we must suffer the pain or inconvenience caused by malady or harm.

The SCC has four different charts with different time scales. The daily SCC, for instance, presents data in the scale of days. Behavior analysts who use the daily SCC will see how the intervention they applied to behavior changes across each week (seven days = a “celeration period”).

If more behavior analysts used the SCC we would know generally how long procedures such as time out, positive reinforcement derived interventions, or differential reinforcement take to change behavior. 

In other words, behavior analysts could tell parents how long an intervention would likely take. Furthermore, real time charting shows the effects of illness, vacation days, and time spent away from an intervention.

Celeration line. Skinner and his colleagues used frequency or rate (the count or number of behaviors over a specific time period) to measure behavior. Lindsley and Precision Teachers (Lindsley started a measurement system called Precision Teaching) also used frequency to measure behavior. Placing successive frequencies across time gave rise to a new, powerful behavioral measure called celeration.

The celeration line shows the direction and speed of behavior change. The steeper the celeration line the faster the behavior has changed. Applied behavior analysts who use celeration lines can see how fast a pinpointed behavior such as “calls out answer in class without permission” changes as a result of an intervention. 

In Applied Behavior Analysis a hallmark of the science resides in a trait called “Effective.” According to the authors, Effective means “If the application of behavioral techniques does not produce large enough effects for practical value, then the application has failed” (Baer, Wolf & Risley, 1968, p. 96). 

A celeration line flashes a red or green light in the face of the behavior analyst. Effective interventions change behaviors in the right direction and do so quickly. Without celeration lines and real time behavior monitoring, behavior analysts have no way of knowing, or communicating to others, the precise speed of change of any intervention.

Using the celeration line also permits behavior analysts to match up individual interventions and their efficiency. On a daily Standard Celeration Chart a x1.5 celeration value (the quantification of the celeration line) means the behavior grew 50% each week. Behavior analysts could report the effects to parents, teachers, insurance companies, and other people who have a vested interest in rate of behavior change.

Projected Celeration Line

The celeration line becomes possible because of the special architecture of the Standard Celeration Chart. The vertical ratio scale shows behavior changing by equal ratios instead of equal intervals (as they do on linear graphs). 

The properties of the ratio scale result in very nice (accurate) projections of future behavior. Straight line projections on ratio charts like the SCC turn into curvilinear lines on linear graphs. The ability to accurately project the future course of behavior facilitates the analysis, evaluation, and planning of interventions.

Bounce

Lindsley used the plain English word “bounce” to communicate the concept of “variability” to teachers, parents, and students. When measured across time, behavior bounces around. The bounce always occurs because humans never perform a behavior the exact same way. Those variations or bounce may take place very slightly (low bounce) or quite significantly (high bounce).

The degree of bounce directly reflects the influence of an intervention. As an example, examine the charted data of two kindergartners in the same class. The teacher does have an intervention; classroom rules which say students cannot leave their seats without permission.

Picture
 

Figure 3. Bounce the SCC.

Which rule exerts greater influence on each student? The bounce for student 2 shows consistency and regularity. By contrast the bounce of student 1 has a much larger envelope. The larger bounce means student 1 shows much more inconsistency; one day he may get out of seat once, whereas another day he may stand up 5 times. 

The SCC not only shows bounce clearly, it also allows behavior analysts to explicitly quantify the degree of influence. Student 2 has a bounce of x2. Student 1 has a bounce of x5. The x2 bounce makes obvious the degree of influence the intervention has on out of seat behavior. Student 1’s x5 bounce has more variability indicating the classroom rules (intervention) do not have very strong control (influence) over the targeted behavior. In fact student 1 has 2.5 times more behavioral variability when compared to student 2. 

Conclusion

The previous three attributes of the the SCC markedly enhance the power of behavior analysts' ability to detect meaningful changes. Furthermore, the analysis, interpretation, and communication of applied data outcomes greatly improves.

References

Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior analysis. Journal of Applied Behavior Analysis, 1, 91–97. doi: 10.1901/jaba.1968.1-91 .

Skinner, B. F. (1956). A case study in scientific method. American Psychologist, 11, 221–233.

Subscribe to Email Updates