IDEA: Functional Behavioral Analysis & The Law: Proof

The GRAPH

The DATA

These are three critical questions when focusing on a Functional Analysis of behavior. Consider the above graph and the following example…...

The question: "Given John’s low frequency of ‘in-seat’ behavior, will the introduction of computer use as a reward when sitting in his seat prove an effective intervention?"

• There are three "phases" to proving (yes, computer time works!) or disproving (no, John still refuses to return to his seat when asked!) whether or not the "computer intervention" will work -- whether or not the computer will work in getting John to return to his seat.
• Phase A: For a predetermined number of opportunities ("John, please sit down"), the data-collector (teacher, aide, parent, student-teacher) marks down on a piece of paper whether or not Johnny returns to his seat when asked by the teacher or aide. The data can track, say, 50 task requests ("Johnny, please sit down"), or maybe even more trials over the course of perhaps a week. For the purpose of this discussion and looking at the graph above, notice that Johnny sat down only 10 of the 50 times asked.
• Phase B: An Apple IIe computer is set up on Johnny’s desk. He is allowed to engage in computer-tasks related to his functioning level while in his seat. Aha! Notice that Johnny sits down 40 of the 50 times!
• Now, that seems like a great improvement …. Using appropriate computer-based tasks as the "lure" or "reward" to entice Johnny to return to his seat.
• (A return to) Phase A: What will happen if the teacher suddenly removes the computer? Will Johnny return to his out-of-seat behavior? Well, we need to collect the data one more time to find out! Such is the purpose of the "return-to-baseline" phase.

The ANSWER

Conclusion: "Let the data do the talking" – Yes, computer-use is an effective intervention for getting Johnny to "do what he otherwise resists" … returning to his seat when requested.

• Now, an ABA design is a cruel design. If an intervention works, it is by definition reinforcing if it increases the rate or frequency of the target behavior (e.g., returning to seat). So why would a moral, kind, teacher (see Highet, 1951; Lortie, 1973) even think of stopping something that is good and that works? No teacher would. So in fact, an ABA design really finishes up as an ABAB design: The last/4^th Phase should always involve the re-introduction of that which worked (in this case, periodic computer use) for the student.
• Will the use of the computer always work with Johnny? Probably not. We all get bored with most things we like, if what we get that we like is only one thing! So the teacher would be wise to use the same ABAB design to ferret out other interventions that can also be used to persuade Johnny to return to his seat.
• Is it practical to lug a computer over to Johnny’s desk time and time again. Nope! Not if it is a desktop computer! As stressed below, there are now dozens of used notebook-size computers available from Dell, from Gateway, from Toshiba and other leading notebook vendors. Many are under $500 and dropping in price almost by the month. And good teachers learn how to write mini-grants to get such hardware. Not a problem.
• Is there another measure one might use other than "frequency counts or tallies"? Indeed. Duration is a very practical measure as well: Measure how long Johnny remains in his seat once he sits down. Or, measure how long it takes him to wander over to his seat when asked. Or, measure … well, as we know from elsewhere on this website, it’s all a function of context.
• Are there so-called "single subject" research designs (also referred to as "within-subject" or "intrasubject") other than or in addition to the ABAB Design discussed here? Indeed! Withdrawal design; Multiple baseline designs (there are several); Changing Criterion design; Alternating Treatments design; Multi-element designs; and more. All such single subject designs have their strengths and shortcomings, so to understand the ABAB design is to whet your thirst for more! Hence, the references below. The Sidman book is a classic. The Campbell and Stanley reference below offers a classic discussion and succinct distinction between single-subject and multiple-subject research designs. And remember: the purpose and point of all research designs is to establish, if possible, a functional relationship between the behavior ("dependent variable") and intervention ("independent variable").
  

• Why an Apple IIe computer? Isn't the IIe, ah, rather outdated, to say the least? Yes and no (surprise)! Yes, the IIe is ancient, but it is certainly not outdated. The IIe remains without question the most appropriate computer for teachers of low functioning students in particular. The IIe is rugged, solid, one-piece, easy to learn, easier to use, happily inexpensive ($500) and there are hundreds of software titles still available. (Again, download, try and buy any one of our inexpensive software titles and we will direct you to the page on this website where you will find a unique, one-of-a-kind Educational Matrix (screen shot). The matrix lists hundreds of software titles, websites, phone numbers, etc for the IIe, Mac, DOS, and Windows platforms.) In this day and age, lightweight, powerful, easily mobile notebooks may be a practical hardware alternative in some circumstances, to be sure. (My vote stays with the IIe for the severely handicapped, however.)
• Can we draw any "cause-effect" conclusions from this data we have collected? No. There are simply too many other factors that might come into play at any time that might affect the frequency (or duration) of Johnny’s behavior. Some possible factors: medication, fatigue, hunger, ambient noise, language cues, what might have gone on in his life just prior to entering the classroom ... and on and on. However, yes, we can state a "functional relationship:" We can state that if you do this, the chances are that such and such will happen …. IF you do "this" (whatever "this" may be).

The REFERENCES

NOTE: Some computer-savy practitioners may want to consult the Carr & Burkholder reference in particular, cited below. The authors provide an excellent recipe for creating and customizing ABAB design graphs using Excel. Moreover, the authors also set forth the steps for creating Multiple Baseline graphs, also valuable for serious data-conscious teachers.

* Barlow, D.H., & Hersen, M. (1984). Single case experimental designs: Strategies for studying behavior change (2nd ed.). New York: Pergamon Press.

Birnbrauer, J.S., Peterson, C.P., & Solnick, J.V. (1974). The design and interpretation of studies of single subjects. American Journal of Mental Deficiency, 79, 191-203.

* Campbell, D.T., & Stanley, J.C. (1963). Experimental and quasi-experimental designs for research on teaching. In N.L. Gage (Ed.), Handbook of Research on Teaching (pp. 171-246). Chicago: Rand McNally. (Also published in 1966 as Experimental and Quasi-experimental designs for research. Chicago: Rand McNally.)

* Carr, J.E., & Burkholder, E.O. (1998). Creating single-subject design graphs with Microsoft Excel. Journal of Applied Behavior Analysis, 31, 245-251.

Edgar, E., & Billingsley, F. (1974). Believability when N = 1. The Psychological Record, 24, 147-160.

* Hartmann. D.P., & Atkinson, C. (1973). Having your cake and eating it too: A note on some apparent contradictions between therapeutic achievements and design requirements in N=1 studies. Behavior Therapy, 4, 589-591.

Highet, G. (1951). The art of teaching. New York: Vintage Books.

* Huck, S.W., Cormier, W.H., & Bounds, W.G. (1974). Pseudo-experimental designs. In S.W. Huck, W.H. Cormier, & W.G. Bounds, Reading statistics and research (pp. 226-240). New York: Harper & Row.

Kazdin, A.E. (1982). Single case research design: Methods for clinical and applied settings. New York: Oxford University Press.

Lortie, D.C. (1975). Schoolteacher. Chicago: University of Chicago Press.

Miller, D.C. (1983). Handbook of research design and social measurement (4th ed.). New York: Longman.

* Sidman, M. (1960). Tactics of scientific research: Evaluating experimental data in psychology. New York: Basic Books.

Don't forget: Biblio-Refs offers dozens more references in this area.
Download and enjoy!