JOURNAL OF APPLIED BEHAVIOR ANALYSIS

SOME CURRENT DIMENSIONS OF APPLIEDBEHAVIOR ANALYSIS'

DONALD M. BAER, MONTROSE M. WOLF, AND TODD R. RISLEY

THE UNIVERSITY OF KANSAS

The analysis of individual behavior is aproblem in scientific demonstration, reason-ably well understood (Skinner, 1953, Sec. 1),comprehensively described (Sidman, 1960),and quite thoroughly practised (Journal ofthe Experimental Analysis of Behavior, 1957-). That analysis has been pursued in manysettings over many years. Despite variableprecision, elegance, and power, it has resultedin general descriptive statements of mecha-nisms that can produce many of the formsthat individual behavior may take.The statement of these mechanisms estab-

lishes the possibility of their application toproblem behavior. A society willing to con-sider a technology of its own behavior appar-ently is likely to support that applicationwhen it deals with socially important behav-iors, such as retardation, crime, mental illness,or education. Such applications have ap-peared in recent years. Their current num-ber and the interest which they create appar-ently suffice to generate a journal for theirdisplay. That display may well lead to thewidespread examination of these applica-tions, their refinement, and eventually theirreplacement by better applications. Betterapplications, it is hoped, will lead to a betterstate of society, to whatever extent the behav-ior of its members can contribute to the good-ness of a society. Since the evaluation of whatis a "good" society is in itself a behavior ofits members, this hope turns on itself in aphilosophically interesting manner. However,it is at least a fair presumption that behav-ioral applications, when effective, can some-times lead to social approval and adoption.

Behavioral applications are hardly a newphenomenon. Analytic behavioral applica-

'Reprints may be obtained from Donald M. Baer,Dept. of Human Development, University of Kansas,Lawrence, Kansas 66044.

91

tions, it seems, are. Analytic behavioral ap-plication is the process of applying sometimestentative principles of behavior to the im-provement2 of specific behaviors, and simul-taneously evaluating whether or not anychanges noted are indeed attributable to theprocess of application-and if so, to whatparts of that process. In short, analytic be-havioral application is a self-examining, self-evaluating, discovery-oriented research pro-cedure for studying behavior. So is allexperimental behavioral research (at least,according to the usual strictures of moderngraduate training). The differences are mat-ters of emphasis and of selection.The differences between applied and basic

research are not differences between thatwhich "discovers" and that which merely "ap-plies" what is already known. Both endeavorsask what controls the behavior under study.Non-applied research is likely to look at anybehavior, and at any variable which may con-ceivably relate to it. Applied research is con-strained to look at variables which can beeffective in improving the behavior understudy. Thus it is equally a matter of researchto discover that the behaviors typical of re-tardates can be related to oddities of their

2If a behavior is socially important, the usual be-havior analysis will aim at its improvement. The so-cial value dictating this choice is obvious. However,it can be just as illuminating to demonstrate how abehavior may be worsened, and there will arise occa-sions when it will be socially important to do so. Dis-ruptive classroom behavior may serve as an example.Certainly it is a frequent plague of the educationalsystem. A demonstration of what teacher proceduresproduce more of this behavior is not necessarily thereverse of a demonstration of how to promote posi-tive study behaviors. There may be classroom situa-tions in which the teacher cannot readily establishhigh rates of study, yet still could avoid high rates ofdisruption, if she knew what in her own proceduresleads to this disruption. The demonstration whichshowed her that would thus have its value.

1968, 1, 91-97 NUMBER I (SPRING, 1968)

DONALD M. BAER et al.

chromosome structure and to oddities of theirreinforcement history. But (currently) thechromosome structure of the retardate doesnot lend itself to experimental manipulationin the interests of bettering that behavior,whereas his reinforcement input is alwaysopen to current re-design.

Similarly, applied research is constrainedto examining behaviors which are socially im-portant, rather than convenient for study. Italso implies, very frequently, the study ofthose behaviors in their usual social settings,rather than in a "laboratory" setting. But alaboratory is simply a place so designed thatexperimental control of relevant variables isas easy as possible. Unfortunately, the usualsocial setting for important behaviors israrely such a place. Consequently, the analy-sis of socially important behaviors becomesexperimental only with difficulty. As theterms are used here, a non-experimental anal-ysis is a contradiction in terms. Thus, ana-lytic behavioral applications by definitionachieve experimental control of the processesthey contain, but since they strive for this con-trol against formidable difficulties, theyachieve it less often per study than would alaboratory-based attempt. Consequently, therate of displaying experimental control re-quired of behavioral applications has becomecorrespondingly less than the standards typi-cal of laboratory research. This is not becausethe applier is an easy-going, liberal, or gen-erous fellow, but because society rarely willallow its important behaviors, in their cor-respondingly important settings, to be manip-ulated repeatedly for the merely logical com-fort of a scientifically sceptical audience.Thus, the evaluation of a study which pur-

ports to be an applied behavior analysis issomewhat different than the evaluation of asimilar laboratory analysis. Obviously, thestudy must be applied, behavioral, and ana-lytic; in addition, it should be technological,conceptually systematic, and effective, and itshould display some generality. These termsare explored below and compared to the cri-teria often stated for the evaluation of behav-ioral research which, though analytic, is notapplied.

AppliedThe label applied is not determined by the

research procedures used but by the interest

which society shows in the problems beingstudied. In behavioral application, the behav-ior, stimuli, and/or organism under study arechosen because of their importance to manand society, rather than their importance totheory. The non-applied researcher may studyeating behavior, for example, because it re-lates directly to metabolism, and there arehypotheses about the interaction between be-havior and metabolism. The non-applied re-searcher also may study bar-pressing becauseit is a convenient response for study; easy forthe subject, and simple to record and inte-grate with theoretically significant environ-mental events. By contrast, the applied re-searcher is likely to study eating because thereare children who eat too little and adults whoeat too much, and he will study eating inexactly those individuals rather than in moreconvenient ones. The applied researcher mayalso study bar-pressing if it is integrated withsocially important stimuli. A program for ateaching machine may use bar-pressing be-havior to indicate mastery of an arithmeticskill. It is the arithmetic stimuli which areimportant. (However, some future appliedstudy could show that bar-pressing is morepractical in the process of education than apencil-writing response.3)

In applied research, there is typically aclose relationship between the behavior andstimuli under study and the subject in whomthey are studied. Just as there seem to be fewbehaviors that are intrinsically the target ofapplication, there are few subjects who auto-matically confer on their study the status ofapplication. An investigation of visual signaldetection in the retardate may have little im-mediate importance, but a similar study inradar-scope watchers has considerable. Astudy of language development in the re-tardate may be aimed directly at an immedi-

"Research may use the most convenient behaviorsand stimuli available, and yet exemplify an ambitionin the researcher eventually to achieve application tosocially important settings. For example, a study mayseek ways to give a light flash a durable conditionedreinforcing function, because the experimenter wishesto know how to enhance school children's responsive-ness to approval. Nevertheless, durable bar-pressingfor that light flash is no guarantee that the obviousclassroom analogue will produce durable reading be-havior for teacher statements of "Good!" Until theanalogue has been proven sound, application has notbeen achieved.

92

SOME DIMENSIONS OF APPLIED BEHAVIOR ANALYSIS

ate social problem, while a similar study inthe MIT sophomore may not. Enhancementof the reinforcing value of praise for the re-tardate alleviates an immediate deficit in hiscurrent environment, but enhancement of thereinforcing value of 400 Hz (cps) tone for thesame subject probably does not. Thus, a pri-mary question in the evaluation of appliedresearch is: how immediately important isthis behavior or these stimuli to this subject?

BehavioralBehaviorism and pragmatism seem often to

go hand in hand. Applied research is emi-nently pragmatic; it asks how it is possibleto get an individual to do something effec-tively. Thus it usually studies what subjectscan be brought to do rather than what theycan be brought to say; unless, of course, averbal response is the behavior of interest.Accordingly a subject's verbal description ofhis own non-verbal behavior usually wouldnot be accepted as a measure of his actual be-havior unless it were independently substan-tiated. Hence there is little applied value inthe demonstration that an impotent man canbe made to say that he no longer is impotent.The relevant question is not what he can say,but what he can do. Application has not beenachieved until this question has been an-swered satisfactorily. (This assumes, of course,that the total goal of the applied researcheris not simply to get his patient-subjects tostop complaining to him. Unless societyagrees that this researcher should not bebothered, it will be difficult to defend thatgoal as socially important.)

Since the behavior of an individual is com-posed of physical events, its scientific studyrequires their precise measurement. As a re-sult, the problem of reliable quantificationarises immediately. The problem is the samefor applied research as it is for non-appliedresearch. However, non-applied research typi-cally will choose a response easily quantifiedin a reliable manner, whereas applied re-search rarely will have that option. As a re-sult, the applied researcher must try harder,rather than ignore this criterion of all trust-worthy research. Current applied researchoften shows that thoroughly reliable quantifi-cation of behavior can be achieved, even inthoroughly difficult settings. However, it alsosuggests that instrumented recording with its

typical reliability will not always be possible.The reliable use of human beings to quantifythe behavior of other human beings is anarea of psychological technology long sincewell developed, thoroughly relevant, and veryoften necessary to applied behavior analysis.A useful tactic in evaluating the behavioral

attributes of a study is to ask not merely, wasbehavior changed? but also, whose behavior?Ordinarily it would be assumed that it wasthe subject's behavior which was altered; yetcareful reflection may suggest that this wasnot necessarily the case. If humans are ob-serving and recording the behavior understudy, then any change may represent achange only in their observing and record-ing responses, rather than in the subject's be-havior. Explicit measurement of the reliabil-ity of human observers thus becomes notmerely good technique, but a prime criterionof whether the study was appropriately be-havioral. (A study merely of the behavior ofobservers is behavioral, of course, but prob-ably irrelevant to the researcher's goal.) Alter-natively, it may be that only the experimen-ter's behavior has changed. It may be reported,for example, that a certain patient rarelydressed himself upon awakening, and conse-quently would be dressed by his attendant.The experimental technique to be appliedmight consist of some penalty imposed unlessthe patient were dressed within half an hourafter awakening. Recording of an increasedprobability of self-dressing under these condi-tions might testify to the effectiveness of thepenalty in changing the behavior; however, itmight also testify to the fact that the patientwould in fact probably dress himself withinhalf an hour of arising, but previously wasrarely left that long undressed before beingclothed by his efficient attendant. (The at-tendant now is the penalty-imposing experi-menter and therefore always gives the patienthis full half-hour, in the interests of preciseexperimental technique, of course.) This erroris an elementary one, perhaps. But it suggeststhat in general, when an experiment proceedsfrom its baseline to its first experimental phase,changes in what is measured need not alwaysreflect the behavior of the subject.

AnalyticThe analysis of a behavior, as the term is

used here, requires a believable demonstra-

93

DONALD M. BAER et al.

tion of the events that can be responsible forthe occurrence or non-occurrence of that be-havior. An experimenter has achieved ananalysis of a behavior when he can exercisecontrol over it. By common laboratory stan-dards, that has meant an ability of the ex-perimenter to turn the behavior on and off,or up and down, at will. Laboratory standardshave usually made this control clear by dem-onstrating it repeatedly, even redundantly,over time. Applied research, as noted before,cannot often approach this arrogantly fre-quent clarity of being in control of importantbehaviors. Consequently, application, to beanalytic, demonstrates control when it can,and thereby presents its audience with a prob-lem of judgment. The problem, of course, iswhether the experimenter has shown enoughcontrol, and often enough, for believability.Laboratory demonstrations, either by over-replication or an acceptable probability levelderived from statistical tests of grouped data,make this judgment more implicit than ex-plicit. As Sidman points out (1960), there isstill a problem of judgment in any event, andit is probably better when explicit.There are at least two designs commonly

used to demonstrate reliable control of animportant behavioral change. The first canbe referred to as the "reversal" technique.Here a behavior is measured, and the measureis examined over time until its stability isclear. Then, the experimental variable is ap-plied. The behavior continues to be mea-sured, to see if the variable will produce abehavioral change. If it does, the experimen-tal variable is discontinued or altered, to seeif the behavioral change just brought aboutdepends on it. If so, the behavioral changeshould be lost or diminished (thus the term"reversal"). The experimental variable thenis applied again, to see if the behavioralchange can be recovered. If it can, it is pur-sued further, since this is applied researchand the behavioral change sought is an im-portant one. It may be reversed briefly again,and yet again, if the setting in which the be-havior takes place allows further reversals.But that setting may be a school system or afamily, and continued reversals may not beallowed. They may appear in themselves tobe detrimental to the subject if pursued toooften. (Whether they are in fact detrimentalis likely to remain an unexamined question

so long as the social setting in which the be-havior is studied dictates against using themrepeatedly. Indeed, it may be that repeatedreversals in some applications have a positiveeffect on the subject, possibly contributing tothe discrimination of relevant stimuli in-volved in the problem.)

In using the reversal technique, the experi-menter is attempting to show that an analysisof the behavior is at hand: that whenever heapplies a certain variable, the behavior is pro-duced, and whenever he removes this vari-able, the behavior is lost. Yet applied behav-ior analysis is exactly the kind of researchwhich can make this technique self-defeatingin time. Application typically means produc-ing valuable behavior; valuable behaviorusually meets extra-experimental reinforce-ment in a social setting; thus, valuable be-havior, once set up, may no longer be depen-dent upon the experimental technique whichcreated it. Consequently, the number of re-versals possible in applied studies may be lim-ited by the nature of the social setting inwhich the behavior takes place, in more waysthan one.An alternative to the reversal technique

may be called the "multiple baseline" tech-nique. This alternative may be of particularvalue when a behavior appears to be irre-versible or when reversing the behavior is un-desirable. In the multiple-baseline technique,a number of responses are identified and mea-sured over time to provide baselines againstwhich changes can be evaluated. With thesebaselines established, the experimenter thenapplies an experimental variable to one ofthe behaviors, produces a change in it, andperhaps notes little or no change in the otherbaselines. If so, rather than reversing the just-produced change, he instead applies the ex-perimental variable to one of the other, asyet unchanged, responses. If it changes at thatpoint, evidence is accruing that the experi-mental variable is indeed effective, and thatthe prior change was not simply a matter ofcoincidence. The variable then may be ap-plied to still another response, and so on. Theexperimenter is attempting to show that hehas a reliable experimental variable, in thateach behavior changes maximally only whenthe experimental variable is applied to it.How many reversals, or how many base-

lines, make for believability is a problem for

94

SOME DIMENSIONS OF APPLIED BEHAVIOR ANALYSIS

the audience. If statistical analysis is applied,the audience must then judge the suitabilityof the inferential statistic chosen and the pro-priety of these data for that test. Alternatively,the audience may inspect the data directlyand relate them to past experience with simi-lar data and similar procedures. In eithercase, the judgments required are highly quali-tative, and rules cannot always be stated prof-itably. However, either of the foregoing de-signs gathers data in ways that exemplify theconcept of replication, and replication is theessence of believability. At the least, it wouldseem that an approach to replication is betterthan no approach at all. This should be es-pecially true for so embryonic a field as be-havioral application, the very possibility ofwhich is still occasionally denied.The preceding discussion has been aimed

at the problem of reliability: whether or nota certain procedure was responsible for a cor-responding behavioral change. The two gen-eral procedures described hardly exhaust thepossibilities. Each of them has many varia-tions now seen in practice; and current ex-perience suggests that many more variationsare badly needed, if the technology of impor-tant behavioral change is to be consistentlybelievable. Given some approach to reliabil-ity, there are further analyses of obvious valuewhich can be built upon that base. For exam-ple, there is analysis in the sense of simplifi-cation and separation of component processes.Often enough, current behavioral proceduresare complex, even "shotgun" in their applica-tion. When they succeed, they clearly need tobe analyzed into their effective components.Thus, a teacher giving M & M's to a childmay succeed in changing his behavior asplanned. However, she has almost certainlyconfounded her attention and/or approvalwith each M & M. Further analysis may beapproached by her use of attention alone, theeffects of which can be compared to the ef-fects of attention coupled with candies.Whether she will discontinue the M Sc M's,as in the reversal technique, or apply atten-tion with M Sc M's to certain behaviors andattention alone to certain others, as in themultiple baseline method, is again the prob-lem in basic reliability discussed above. An-other form of analysis is parametric: a dem-onstration of the effectiveness of differentvalues of some variable in changing behavior.

The problem again will be to make such ananalysis reliable, and, as before, that mightbe approached by the 'repeated alternate useof different values on the same behavior (re-versal), or by the application of different val-ues to different groups of responses (multiplebaseline). At this stage in the development ofapplied behavior analysis, primary concern isusually with reliability, rather than with para-metric analysis or component analysis.

Technological"Technological" here means simply that

the techniques making up a particular behav-ioral application are completely identifiedand described. In this sense, "play therapy"is not a technological description, nor is "so-cial reinforcement". For purposes of applica-tion, all the salient ingredients of play ther-apy must be described as a set of contingen-cies between child response, therapist re-sponse, and play materials, before a statementof technique has been approached. Similarly,all the ingredients of social reinforcementmust be specified (stimuli, contingency, andschedule) to qualify as a technological pro-cedure.The best rule of thumb for evaluating a

procedure description as technological isprobably to ask whether a typically trainedreader could replicate that procedure wellenough to produce the same results, givenonly a reading of the description. This isvery much the same criterion applied to pro-cedure descriptions in non-applied research,of course. It needs emphasis, apparently, inthat there occasionally exists a less-than-pre-cise stereotype of applied research. Where ap-plication is novel, and derived from princi-ples produced through non-applied research,as in current applied behavior analysis, thereverse holds with great urgency.

Especially where the problem is applica-tion, procedural descriptions require consid-erable detail about all possible contingenciesof procedure. It is not enough to say what isto be done when the subject makes responseR1; it is essential also whenever possible tosay what is to be done if the subject makes thealternative responses, R2, R3, etc. For exam-ple, one may read that temper tantrums inchildren are often extinguished by closingthe child in his room for the duration of thetantrums plus ten minutes. Unless that pro-

95

DONALD M. BAER et al.

cedure description also states what should bedone if the child tries to leave the room early,or kicks out the window, or smears feces onthe walls, or begins to make strangling sounds,etc., it is not precise technological description.

Conceptual SystemsThe field of applied behavior analysis will

probably advance best if the published de-scriptions of its procedures are not only pre-cisely technological, but also strive for rele-vance to principle. To describe exactly howa preschool teacher will attend to jungle-gymclimbing in a child frightened of heights isgood technological description; but further tocall it a social reinforcement procedure re-lates it to basic concepts of behavioral devel-opment. Similarly, to describe the exact se-quence of color changes whereby a child ismoved from a color discrimination to a formdiscrimination is good; to refer also to "fad-ing" and "errorless discrimination" is better.In both cases, the total description is ade-quate for successful replication by the reader;and it also shows the reader how similar pro-cedures may be derived from basic principles.This can have the effect of making a body oftechnology into a discipline rather than acollection of tricks. Collections of tricks his-torically have been difficult to expand system-atically, and when they were extensive, diffi-cult to learn and teach.

EffectiveIf the application of behavioral techniques

does not produce large enough effects forpractical value, then application has failed.Non-applied research often may be extremelyvaluable when it produces small but reliableeffects, in that these effects testify to the op-eration of some variable which in itself hasgreat theoretical importance. In application,the theoretical importance of a variable isusually not at issue. Its practical importance,specifically its power in altering behaviorenough to be socially important, is the essen-tial criterion. Thus, a study which shows thata new classroom technique can raise the gradelevel achievements of culturally deprived chil-clren from D- to D is not an obvious exam-ple of applied behavior analysis. That samestudy might conceivably revolutionize educa-tional theory, but it clearly has not yet revo-lutionized education. This is of course a mat-

ter of degree: an increase in those childrenfrom D- to C might well be judged an im-portant success by an audience which thinksthat C work is a great deal different than Dwork, especially if C students are much lesslikely to become drop-outs than D students.

In evaluating whether a given applicationhas produced enough of a behavioral changeto deserve the label, a pertinent question canbe, how much did that behavior need to bechanged? Obviously, that is not a scientificquestion, but a practical one. Its answer islikely to be supplied by people who must dealwith the behavior. For example, ward person-nel may be able to say that a hospitalizedmute schizophrenic trained to use 10 verballabels is not much better off in self-helpskills than before, but that one with 50 suchlabels is a great deal more effective. In thiscase, the opinions of ward aides may be morerelevant than the opinions of psycholinguists.

GeneralityA behavioral change may be said to have

generality if it proves durable over time, ifit appears in a wide variety of possible envi-ronments, or if it spreads to a wide varietyof related behaviors. Thus, the improvementof articulation in a clinic setting will prove tohave generality if it endures into the futureafter the clinic visits stop; if the improved ar-ticulation is heard at home, at school, and ondates; or if the articulation of all words, notjust the ones treated, improves. Applicationmeans practical improvement in importantbehaviors; thus, the more general that appli-cation, the better, in many cases. Therapistsdealing with the development of heterosexualbehavior may well point out there are sociallyappropriate limits to its generality, once de-veloped; such limitations to generality areusually obvious. That generality is a valuablecharacteristic of applied behavior analysiswhich should be examined explicitly appar-ently is not quite that obvious, and is statedhere for emphasis.That generality is not automatically accom-

plished whenever behavior is changed alsoneeds occasional emphasis, especially in theevaluation of applied behavior analysis. It issometimes assumed that application has failedwhen generalization does not take place inany widespread form. Such a conclusion hasno generality itself. A procedure which is ef-

96

SOME DIMENSIONS OF APPLIED BEHAVIOR ANALYSIS 97

fective in changing behavior in one settingmay perhaps be easily repeated in other set-tings, and thus accomplish the generalizationsought. Furthermore, it may well prove thecase that a given behavior change need beprogrammed in only a certain number ofsettings, one after another, perhaps, to ac-complish eventually widespread generaliza-tion. A child may have 15 techniques fordisrupting his parents, for example. The elim-ination of the most prevalent of these maystill leave the remaining 14 intact and inforce. The technique may still prove both val-uable and fundamental, if when applied tothe next four successfully, it also results inthe "generalized" loss of the remaining 10. Ingeneral, generalization should be pro-grammed, rather than expected or lamented.

Thus, in summary, an applied behavior

analysis will make obvious the importance ofthe behavior changed, its quantitative charac-teristics, the experimental manipulationswhich analyze with clarity what was responsi-ble for the change, the technologically exactdescription of all procedures contributing tothat change, the effectiveness of those proce-dures in making sufficient change for value,and the generality of that change.

REFERENCESJournal of the Experimental Analysis of Behavior.

Bloomington: Society for the Experimental Analy-sis of Behavior, 1957-.

Sidman, Murray. Tactics of scientific research. NewYork: Basic Books, 1960.

Skinner, B. F. Science and human behavior. NewYork: Macmillan, 1953.

Received 24 December 1967.