Filetype PDF | Posted on 21 Aug 2022 | 3 years ago
Authentication
digital resources
230x Filetype PDF File size 2.05 MB Source: www.ncbi.nlm.nih.gov
The Behavior Analyst 1996, 19, 35-47 No. 1 (Spring)
Distinguishing Between Applied Research and Practice
James M. Johnston
Auburn University
Behavior-analytic research is often viewed along a basic-applied continuum of research goals and
methods. The applied portion of this continuum has evolved in ways that combine applied research
and service delivery. Although these two facets of applied behavior analysis should be closely
related, more clearly distinguishing between them, particularly in how we conceptualize and conduct
applied research, may enhance the continuing development of each. This differentiation may im-
prove the recruitment and training of graduate students. analysis, graduate training
Key words: basic research, applied research, applied behavior
Evolution ofApplied Behavior Experimental Analysis of Behavior,
Analysis which had been established in 1957 to
Applied behavior analysis was for- publish the Journal of the Experimen-
malized in the 1960s with the efforts tal Analysis of Behavior (JEAB). The
of a relatively small number of re- definition of the new specialty was de-
searchers to determine if the basic prin- scribed in a seminal article in the first
ciples of operant conditioning could be issue, "Some Current Dimensions of
used to solve behavioral problems in Applied Behavior Analysis" by Baer,
everyday situations. Although these at- Wolf, and Risley (1968). Its placement
tempts were practical in focus, the in the premier issue, the authors' rep-
style of analysis was often experimen- utations, and the influence of their ac-
tal, perhaps in part because many of ademic department at the University of
the researchers were experienced in ba- Kansas insured the article substantial
sic animal research. Early reports of and enduring impact (it has been cer-
this work were scattered among pro- tified as a citation classic by the Social
fessional journals in clinical psycholo- Science Citation Index). terms
gy, psychiatry, and child development. The article defined a set of
However, edited volumes by Krasner germane to the conception of applied
and Ullman (1965), Ullman and Kras- behavior analysis. The definition dif-
ner (1965), Ulrich, Stachnik, and Ma- ferentiated the new journal from its
bry (1966), and Neuringer and Michael older sister publication, JEAB. What
(1970) helped to identify and encour- emerged was a definition of applied
age the nascent discipline. behavior analysis as a particular style
If there was an official announce- of analysis, differing in certain ways
ment of the birth of applied behavior from the style of analysis practiced in
analysis, it was the publication in 1968 animal laboratories. Baer et al. dis-
of the Journal of Applied Behavior cussed the definition of seven terms:
Analysis (JABA) by the Society for the The evaluation of a study which purports to be
an applied behavior analysis is somewhat dif-
ferent than the evaluation of a similar laboratory
Portions of this paper are based on presenta- analysis. Obviously, the study must be applied,
tions at the 1993 meeting of the Association for behavioral, and analytic; in addition, it should
Behavior Analysis (Johnston, 1993a, 1993b). I be technological, conceptually systematic, and
would like to thank Bill Hopkins and Hank Pen- effective, and it should display some generality.
nypacker for comments on an earlier draft of this These terms are explored below and compared
manuscript, to Janet Ellis for her encourage- to the criteria often stated for the evaluation of
ment, and to the reviewers and the editor for behavioral research which, though analytic, is
their considerable guidance. not applied. (1968, p. 92)
Correspondence should be addressed to J. M. The authors approached the discussion
Johnston, Department of Psychology, Auburn in terms of the challenge of evaluating
University, Auburn, Alabama 36849.
35
36 JAMES M. JOHNSTON
whether a candidate study would qual- many cases, the pervasiveness of this
ify as an applied behavior analysis, es- tendency suggests a narrower view of
pecially in contrast to a nonapplied an- research interests and methodological
alytical effort. In considering the defin- options than may be necessary. The
ing features of an applied behavior-an- priority of the applied context may
alytic study, the paper conveyed a therefore discourage studies using sub-
certain model of applied behavior anal- jects not indigenous to the applied set-
ysis as a field of research. ting, target behavior selected for ex-
Over the ensuing years, there have perimental rather than applied purpos-
been many assessments of the evolu- es, or analogue settings that offer a
tion of applied behavior-analytic re- higher degree of control than might
search as represented by JABA and oth- otherwise be available.
er applied behavioral journals (Hayes, Furthermore, the overriding criterion
Rincover, & Solnick, 1980). For ex- for the effectiveness or success of ap-
ample, some have noted that the goal plied research often seems to be the ex-
of much applied behavioral research tent to which the results ameliorate the
seems to be to demonstrate a new tech- presenting problem. Obtaining a large
nique or a new application of an exist- effect seems important (Baer, 1977;
ing technique (Birnbrauer, 1979; Hayes Parsonson & Baer, 1986). This can
et al., 1980). Studies designed primar- mean that studies reporting clear but
ily to analyze an intervention proce- modest effects are less likely to be
dure or treatment effect or to replicate published, even though such effects
a prior study are less common (Hayes may be important in understanding a
et al., 1980) and seem more likely to certain type of behavior. A desire for
be rejected or given secondary status large effects can also discourage ef-
by being categorized as a report instead forts to establish good control over ex-
of an article (Birnbrauer, 1979). traneous variables and attempts to con-
In pursuit of the interest in demon- duct analytical studies that examine the
strating a practical capability, an over- contributions of component variables
riding consideration observed in ap- whose separate effects might be rela-
plied journals is delivering a behavior- tively small.
change service that can solve an ap- Finally, some have expressed con-
plied problem, not just in general but cern that the relation of intervention
typically in the immediate circum- procedures to underlying behavioral
stance presented by each study. Dem- principles is sometimes made on a
onstrating a capability to solve a prob- nominal basis guided by topographical
lem sometimes takes precedence over similarities (Birnbrauer, 1979; Hayes et
more experimental objectives such as al., 1980; Michael, 1980). That is, pro-
learning about the etiology of the be- cedures may sometimes be described
havior, elucidating sources of control, as embodying certain basic principles
or understanding the contribution of of conditioning because the operations
component variables (Birnbrauer, 1979; take the same form as the referenced
Deitz, 1982; Hayes et al., 1980; Mi- principles.
chael, 1980; Pierce & Epling, 1980). Whether these features are problem-
The applied context seems to domi- atic may depend on one's conception
nate not only the research agenda but of what applied behavior-analytic re-
often how it is pursued. Research search literature should look like (e.g.,
methods are sometimes subservient to see Baer, 1981). However, it may be
the project's compatibility with the de- important to consider how they influ-
mands of the applied situation. Thus, ence the pursuit of still other charac-
studies typically use subjects, behavior, teristics of the discipline. For example,
and settings presented by the applied there are a number of possible costs of
problem (Hayes et al., 1980). Although emphasizing the demonstration of new
such decisions may be appropriate in techniques or applications. These in-
APPLIED RESEARCH AND PRACTICE 37
clude a failure to develop an adequate whose origins are not limited to the
understanding of the etiology and traditions of JABA. Nevertheless, it
maintenance of important behavior, a may be argued that to the degree that
shortage of analyses describing how the problematic characteristics de-
complex procedures actually work (or scribed above are present, they limit
why they do not), and the lack of a the effectiveness of the literature in ex-
thematic style in the literature. Al- plaining human behavior and offering
though there may be large areas of re- a consistently effective technology for
search that address a certain problem ameliorating problems.
or method, the studies collectively may Given that these concerns have been
fall short of providing a complete and discussed by various authors over the
integrated picture of the nature of the past two decades, it is reasonable to
problem or the mechanisms of its so- wonder about the influences that main-
lution. These tendencies have been ob- tain some of the applied research prac-
served in the applied literature (Birn- tices that might be considered prob-
brauer, 1979; Deitz, 1982; Hayes et al., lematic. Although early guidance and
1980; Michael, 1980). exemplars may have played some role,
Giving the applied priority of con- they cannot be held responsible for so
text over research methods can also be much behavior over such a long peri-
costly. Measurement procedures some- od. Speculation about the contingen-
times seem to accommodate the con- cies at work might start with the rein-
straints often associated with applied forcers gained by helping in a direct
circumstances, such as limitations on and immediate way those serving as
observing target behavior. One result is subjects in applied research. Many who
the frequent use of problematic mea- conduct applied research probably
surement procedures such as interval chose careers focusing on applied is-
recording (Kelly, 1977). (A number of sues because they enjoy working with
studies have emphasized the shortcom- certain populations or in certain set-
ings of interval recording; e.g., see tings. Indeed, many who conduct ap-
Powell, Martindale, & Kulp, 1975; plied research are employed in service
Powell & Rockinson, 1978; Repp, settings and are therefore likely to be
Roberts, Slack, Repp, & Berkler, greatly influenced by the service con-
1976.) Compromises in experimental tingencies of such settings. Conceptu-
control and design may be settled in alizing a study that requires a compar-
favor of the dictates of the applied sit- atively contrived setting or nonindi-
uation. For instance, the modal number genous subjects is probably less likely
of days in baseline phases in JABA ar- for such individuals than a study
ticles from 1968 through 1976 has whose objectives include ameliorating
been reported to be between 3 and 4 a presenting problem.
(Huitema, 1986), too brief to achieve More broadly, conducting research
the objectives of the steady-state strat- aimed at creating effects of obvious
egy, which is at the heart of within- practical value is likely to contact so-
subject comparisons. cietally mediated reinforcers beyond
The above assessments and concerns the research setting. Proposing to have
suffer the inevitable constraints of developed a solution for a behavioral
summarizing broad areas of research problem often leads to professionally
and may be countered with notable ex- and personally powerful reinforcers.
ceptions. There is no attempt here to Although conducting a component
provide a comprehensive review of the analysis as part of a long-term research
characteristics of the applied behavior- program may augment a procedure's
al literature, the overall quality of the eventual effectiveness, publishing a
literature as a research enterprise, or study that shows that it is possible to
the value of its service achievements. ameliorate a behavioral problem, par-
Furthermore, this is a diverse literature ticularly if an effective procedure has
38 JAMES M. JOHNSTON
not been previously published, is likely may even burden service delivery with
to be especially reinforcing. unnecessary research obligations. The
Some features of applied research result may be areas or types of applied
literature may result from convenience research that are deficient or even ab-
compared to apparent alternatives. For sent.
example, brief phases are generally Some of the missing research agen-
less troublesome than long ones. Sim- da might be addressed by expanded ba-
ilarly, it is easier to tolerate excessive sic research interests. There has been
variability in the hope that large effects increased attention in the basic re-
will make findings clear than to search search literature over the last couple of
out the sources of variability and con- decades to questions that specifically
trol them, especially when there are concern human behavior (Buskist &
barriers to modifying applied settings Miller, 1982; Hyten & Reilly, 1992;
for experimental purposes. The famil- Johnston, 1983). Some of these re-
iarity of certain practices such as inter- search themes have not been well in-
val recording may also prompt deci- tegrated with nonhuman research (Per-
sions to use them in spite of docu- one, 1985a), and some topics have not
mented weaknesses. When convenience realized the potential that might be an-
and familiarity are not subservient to ticipated. For instance, attention to so-
the effects of thorough research train- cial phenomena (e.g., Hake & Olivera,
ing in graduate school, the many 1978) has yet to blossom into an iden-
choices involved in conceptualizing tifiable and multifaceted area of basic
and conducting a study may not always research (Cherek, 1995; Schmitt, 1995;
lead to analytical clarity and strength. Sherburne & Buskist, 1995). However,
Under such conditions, decisions about some areas are showing steady growth
the research features of applied studies and integration. Experimental interest
(e.g., measurement procedures, exper- in verbal behavior has increased con-
imental design and control) may some- siderably (reflected by the evolution of
times yield to practical limitations and The Analysis of Verbal Behavior from
service-oriented emphases imposed by newsletter to journal status), and the
some applied settings. burgeoning stimulus equivalence liter-
APPLIED ature (see Sidman, 1994) seems to be
RESEARCH realizing its potential to be an integral
VERSUS SERVICE part of the basic literature while ex-
Consequences For Applied Research tending to a wide range of basic and
This applied research issues.
brief summary of the evolution What remains underdeveloped in be-
of applied behavior-analytic research havior-analytic research is what in oth-
and some of its characteristics suggests er natural sciences is categorized as ap-
that it has operated under certain con- plied science. Applied science may be
straints, whether imposed by early defined as experimental research that is
traditions or by the contingencies un- connected to basic research through its
der which applied researchers work. experimental style and a basis in fun-
Applied behavior analysis evolved as a damental principles, directly driven by
blend of applied research and service applied issues and problems, but not
delivery. This amalgam may be useful compromised by the practical limita-
in many instances, especially when the tions or the immediate service interests
objective of a project is to ameliorate of applied settings. The traditional con-
a specific behavioral problem. As a ception of applied behavioral research
way of conducting applied science, seems to be shortchanging the larger
however, this traditional model may re- research agenda of behavior analysis
sult in less than adequate experimental by failing to adequately encourage this
pursuit of applied questions in the be- type of experimentation. Although we
havior-analytic research enterprise. It can point to a large and, in some ways,
no reviews yet
Please Login to review.
