You are on page 1of 18




Although the effects of negative reinforcement on human behavior have been studied for a number
of years, a comprehensive body of applied research does not exist at this time. This artide describes
three aspects of negative reinforcement as it relates to applied behavior analysis: behavior acquired
or maintained through negative reinforcement, the treatment of negatively reinforced behavior, and
negative reinforcement as therapy. A consideration of research currently being done in these areas
suggests the emergence of an applied technology on negative reinforcement.
DESCRIPTORS: aversive stimulation, avoidance, escape, negative reinforcement

Research published in the Journal of Applied acteristic of applied behavior analysis (cf. Baer,
Behavior Analysis (JABA) has, for 20 years now, Wolf, & Risley, 1968), is reflected less and less in
demonstrated how knowledge about environment- the research that journals such as JABA publish.
behavior interactions, particularly those involving The general accuracy of these criticisms, as well as
response-contingent events and correlated stimuli, their basis and implications, will continue to be the
may be used for the benefit of individuals and the subject of periodic debate (Baer, 1981; Cullen,
larger society. In doing so, applied research has also 1981; Michael, 1980). Nevertheless, it is possible
made significant contributions to the general science at this point in the development of our field to
of behavior by providing extension and external identify specific and well-established areas of basic
validation of experimental findings from the basic research for which little parallel exists in the applied
research laboratory (Baer, 1978). literature, and vice versa.
Along with the development of the applied field The thesis of this article is that research on neg-
and its expansion into a number of areas in which ative reinforcement provides one of the dearest and
the outcome of an experiment often has immediate most immediately relevant examples of a case in
social implications (e.g., business and industry, de- which consideration, replication, and extension of
velopmental disabilities, education, medicine, men- basic research would benefit the applied area. Along
tal health, public affairs), there has been growing with positive reinforcement and punishment, neg-
concern of a widening gap between basic and ap- ative reinforcement has long been considered one
plied behavior analysis. Critics (Deitz, 1978; Pierce of the elementary principles of behavior. A volu-
& Epling, 1980) have indicated that the emphasis minous amount of research on negative reinforce-
of contemporary applied behavior analysis has shift- ment exists in the basic literature (see reviews by
ed away from the study of conditions that produce Bolles, 1970; Herrnstein, 1969; Hineline, 1977,
change to the production of change per se, and that 1981, 1984; Hoffman, 1966; Schoenfeld, 1969;
"relevance to basic principle," a supposed char- Sidman, 1966), and its inclusion as a distinct topic
in texts on experimental analysis (e.g., Honig &
Staddon, 1977) justifies its status as a major or-
Appreciation is expressed to Ed Malagodi, Jack Michael,
Gary Pace, Terry Page, and Terri Rodgers, whose work or ganizing principle. For example, acquisition, main-
helpful suggestions assisted in the preparation of this article. tenance, extinction, and stimulus control all have
Support was provided by Grant HD 16052 from the National been studied using negative reinforcement as the
Institute of Child Health and Human Development.
Reprints may be obtained from Brian A. Iwata, Depart- operant mechanism of interest.
ment of Psychology, University of Florida, Gainesville, Flor- Sandler and Davidson (1973) reviewed some of
ida 32611. this basic research and discussed its relevance to the

development and treatment of pathological human situations involving the use of negative reinforce-
behavior. They concluded that ". . . the escape and ment and have forced us to consider more generally
avoidance paradigms are still plagued by a number the relevance of negative reinforcement in applied
of unresolved issues .. . " (p. 254) that they hoped behavior analysis. Our experience and our exami-
would be darified by additional basic research and nation of the basic and applied research suggest
extension to the world of humans. Since that time, that the answer to each of the above questions is
a number of investigations on negative reinforce- "No." In fact, it appears that negative reinforce-
ment with humans have been conducted, yet a ment plays a central role in the development of
systematic and comprehensive body of applied re- many behaviors, appropriate as well as inappro-
search still does not exist. Consider the two most priate, and that its application in a number of
recent texts on aversive control with humans (Ax- studies has not been formally acknowledged. In
elrod & Apsche, 1983; Matson & DiLorenzo, what follows, I will describe three aspects of neg-
1984). Both provide thorough discussion of topics ative reinforcement as it relates to the applied sit-
such as positive reinforcement, extinction, time-out, uation: first, undesirable behavior acquired or main-
response cost, and contingent aversive stimulation. tained through negative reinforcement; second, the
Thus, one might expect that these texts would be treatment of negatively reinforced behavior; and
the most likely sources of information on negative third, negative reinforcement as therapy. This or-
reinforcement as well, but this is not the case. One ganization departs somewhat from that used in
text (Axelrod & Apsche, 1983) devotes less than reviews of the basic research literature and has been
a half dozen of over 300 pages to the topic of adopted here to highlight the relevance of particular
avoidance, and the discussion always is limited to issues to the applied researcher. Much of the re-
avoidance as a side effect of punishment. No men- search included here has been done with the de-
tion is made of escape or avoidance as directly velopmentally disabled population because there is
produced performances. The second (Matson & a high prevalence of significant behavioral disorders
DiLorenzo, 1984) describes the hypothetical fea- in this group and because it provides a narrow but
tures of escape and avoidance training on two pages adequate focus for discussion.
but does not cite any applied references.
The relative absence of integrated material on
negative reinforcement with humans raises several CURRENT CONCEPTUALIZATION OF
questions concerning generality and utility. Is hu- NEGATIVE REINFORCEMENT
man behavior relatively insensitive to contingencies Before proceeding, it may be helpful to clarify
of negative reinforcement? Are naturalistic human terminology and to delineate the defining features
situations typically characterized by the absence of of negative reinforcement. The purpose of this
stimuli that can function as negative reinforcers, or digression is to show that the task of determining
opportunities to escape from or avoid these stimuli? whether a given contingency is an example of neg-
What types of performances are likely to be ac- ative reinforcement may not always be a simple
quired through negative reinforcement? Finally, do one. Although there has been little confusion re-
procedures based on the application of negative garding the effect of negative reinforcement, de-
reinforcement, unlike those based on positive re- scribing its operations has posed a challenge to
inforcement and punishment, have little therapeutic many beyond the level of the beginning student.
or pragmatic value? The process of negative reinforcement typically
For the past few years my students, colleagues, involves the removal, reduction, postponement, or
and I have been conducting a series of investigations prevention of stimulation; these operations
in two areas-self-injurious behavior and pediatric strengthen the response on which they are contin-
feeding disorders. Curiously, these very different gent (Hineline, 1977). Removal and reduction of
problems have brought us into direct contact with ongoing stimulation typically produce behavior that
is called "escape," whereas postponement and pre- tify unambiguously a stimulus change as one in-
vention of stimulus presentation produce behavior volving presentation (e.g., of physical contact) or
that is called "avoidance." "Typically" is used as removal (e.g., of a token) and to determine whether
a qualifier throughout because the terms negative or not the response of interest occurs in the presence
reinforcement, escape, and avoidance are subject to or the absence of stimulation. However, because
confusion under certain conditions, as the following research outside ofthe laboratory is subject to great-
will illustrate. er variation of and less control over a multitude of
In commenting on the distinction between pos- potentially relevant stimuli, the motivational fea-
itive and negative reinforcement, Michael (1975) tures of some stimulus changes are difficult to spec-
reviewed a number of historical points related to ify. Consider, as a case in point, Osborne's (1969)
terminological usage. More important, he noted "Free-time as a reinforcer in the management of
that some stimulus changes associated with an in- classroom behavior," which examined the out-of-
crease in behavior are difficult to dassify as "pre- seat behavior of six students. During the baseline
sentation" (positive reinforcement) versus "remov- condition, the students worked for approximately
al" (negative reinforcement), and that the use of 4 hours daily without recess, and data showed that
either description may be nothing more than an the target behavior occurred frequently. During
arbitrary and incomplete abbreviation for the static treatment, students could earn 5 min of free time
"prechange" and "postchange" stimulus condi- at the end of every 1 5-min work period by re-
tions as well as for what transpires in between. For maining in their seats, and the data showed a de-
example, is a change in temperature more accurately crease in out-of-seat. It is interesting to note the
characterized as the presentation of cold (heat) or target behavior. Defined and recorded as "out-of-
the removal of heat (cold)? Problems such as this seat," free time was made available for its absence;
led Michael to suggest that "The distinction be- this type of contingency usually is described as
tween two types of reinforcement (positive vs. neg- differential reinforcement of other behavior (DRO).
ative], based in turn upon the distinction between However, the instructions given to students spec-
presentation and removal simply can be dropped" ified that they were to remain in their seats, sug-
(p. 44). An additional basis for distinguishing be- gesting "in-seat" as the functional target. If so, free
tween positive and negative reinforcement was sug- time was made available for the occurrence of in-
gested first by Catania (1973) and later by Hineline seat behavior; this type of contingency is not con-
(1984), who noted that ". . . if a stimulus or sit- sidered an example of DRO. Depending on how
uation is to be reducible or removable by some one characterizes "free time," (i.e., the availability
response, that response must occur in its presence. of preferred activities vs. the termination of non-
In contrast, positively reinforced responses neces- preferred activities), we would label the contingency
sarily occur in the absence of the stimuli upon which as one involving positive or negative reinforcement
reinforcement is based" (pp. 496-497). Such a for in-seat behavior. Osborne suggested both of
distinction is not without its own problems, as can these possibilities in his discussion and perhaps this
be seen in the previous example. Is responding prior is why he did not place an adjective in front of the
to a temperature change more accurately described term "reinforcer" in the title of the article.
as responding in the presence versus the absence of As a field, we have not attended carefully to the
heat (cold)? Another problem with this distinction important distinction that Osborne drew. His study
is encountered when one considers the difference is regarded as a seminal piece of research in the
between escape (responding in the presence of stim- applied literature for expanding our notion of what
ulation), and avoidance (responding in the absence constitutes a reinforcing event and for demonstrat-
of stimulation), both of which are examples of ing very nicely the effects of group contingencies,
negative reinforcement. although the exact nature of the contingency is still
In many applied situations, it is possible to iden- unclear. A number of interesting replications and

extensions have appeared in JABA (e.g., Aaron & term "undesirable" is used here only as a means
Bostow, 1978; Baer, Rowbury, & Baer, 1973; of classifying behaviors that are considered inap-
Barrish, Saunders, & Wolf, 1969; Harris & Sher- propriate given the usual social context.
man, 1973, 1974; Long & Williams, 1973; Ma- An initial question of particular interest to those
loney & Hopkins, 1973; Medland & Stachnik, working in applied areas relates to factors that de-
1973), but none have included further discussion termine the form of the response. Acquisition of
or analysis of free time contingencies as positive negatively reinforced behavior has been a subject
versus negative reinforcement. Although such anal- of interest to basic researchers as well because it has
yses may have little or no impact on outcome (i.e., been found that some topographies are more readily
in either case, behavior will have been increased), produced than others. For example, the treadle-
our general tendency to overlook a negative rein- press and shuttle responses of pigeons are more
forcement interpretation may lead to undue em- easily controlled by negative reinforcement than is
phasis on the numerous forms that free time may the key peck, which is highly responsive to positive
take at the expense of considering important fea- reinforcement (Ferrari, Todorov, & Graeff, 1973;
tures of the environment that free time replaces. Foree & LoLordo, 1970; MacPhail, 1968; Rachlin
That is, if free time serves as negative reinforcement, & Hineline, 1967; Smith & Keller, 1970). Similar
its only essential component may be alteration or data based on the study of different species have
termination of the preceding aversive situation. provided some support for the hypothesis that neg-
In a more general sense, the complete analysis ative reinforcement involves selective control over
and specification of conditions in effect prior and preexisting "species-specific defense reactions" to
subsequent to responding was the primary basis aversive stimulation (Bolles, 1970, 1971). This
underlying Michael's (1975) suggestion to elimi- account, however, does not provide adequate ex-
nate the distinction between positive and negative planation for the wide range of human behaviors
reinforcement. It appears unlikely that the terms that apparently is succeptable to negative reinforce-
"positive" and "negative" will be deleted from our ment. A more likely explanation is that aversive
technical vocabulary in the near future; neverthe- stimulation initially produces one or more of a
less, researchers should be cognizant of the fact that variety of responses characteristic of both human
the two are potentially interchangeable and that and nonhuman subjects, including flinching, freez-
failure to consider both possibilities may have a ing, jumping, visual scanning, and related and dif-
limiting effect on experimental procedure, inter- fuse motor activity (see reviews by Davis, 1979;
pretation, and subsequent application. Hutchinson, 1977; Myer, 1971), and that the
eventual and more elaborate form of the behavior
is determined by the individual's previous history
UNDESIRABLE BEHAVIOR ACQUIRED AND and the prevailing contingency.
MAINTAINED BY NEGATIVE Thus, many of the serious behavioral disorders
REINFORCEMENT that are seen in, for example, mentally retarded
Hineline (1977) noted that a typical negative individuals may be a function of negative rein-
reinforcement paradigm includes three features: the forcement applied to a particular behavioral rep-
presence of aversive stimulation, the availability of ertoire and shaped over time. It is possible that
a response, and a suitable contingency between the certain instructional sequences (e.g., requests or even
response and the stimulation. Any behavior thus the appearance of specific training materials or the
developed or maintained, including a variety of instructor) become discriminative for aversive stim-
disruptive, destructive, aggressive, self-injurious, and ulation in the form of physical contact, which is a
otherwise problematic acts, could be considered common element in many teaching routines. At
"normal" or "adaptive" in that it is the orderly first, the stimulation and its associated cues may
outcome of specific conditioning operations. The produce behaviors similar to those noted above. If,

however, other behaviors have been successful in condition, and a third condition in which the ex-
eliminating similar types of stimulation in the past, perimenter spoke instructions to the child-this was
their eventual occurrence should not be surprising. called the "mand" condition. Higher levels of self-
Tantrums, attempts to flee, or destruction of ma- injurious behavior were associated with the mand
terials are examples of such behavior, particularly condition. Carr, Newsom, and Binkoff(1980) con-
if the individual is unskilled at more subtle or ducted a similar analysis of aggressive behavior in
socially acceptable forms of escape. Although dis- two boys, showing that aggression was more likely
ruptive, these behaviors usually are not considered to occur when demands were present than when
insurmountable barriers to instruction. A number they were absent. Finally, Iwata, Dorsey, Slifer,
of informal and formal interventions (e.g., pro- Bauman, and Richman (1982) described a general
ceeding in spite of the tantrum, "scooting" the methodology that allowed one to differentiate self-
individual's chair under a table and backing both injury associated with positive versus negative re-
against a wall, bolting the materials to the table, inforcement. In one of the conditions, self-injury
etc.) are successful in managing disruptive behavior was followed by adult attention; in another, self-
in some cases. In other cases, the interventions may injury produced brief escape from adult demands.
provide a means for shaping more serious forms of Some subjects consistently exhibited self-injury dur-
escape. The immediate result of aggression for the ing the latter condition, suggesting that their be-
individual toward whom it is directed suggests that havior was more sensitive to and maintained by
physically harmful acts could serve as very effective negative reinforcement.
escape behaviors, and their ability to terminate aver- It is important for us to identify how environ-
sive instruction is most likely a function of the ments that we create may provide negative rein-
relative size and strength of client and trainer. Fi- forcement for undesirable behaviors. When faced
nally, self-injurious behavior, if severe enough, will with situations in which our students and clients
quickly terminate any situation. are disruptive, we should immediately examine the
Data relevant to a negative reinforcement hy- antecedent as well as the consequent conditions to
pothesis for the development of behavior disorders determine if the difference between the two provides
in the mentally retarded exist in retrospective form reduction of aversive stimulation, keeping in mind
only because it would be unethical to produce that negative reinforcers may be just as idiosyncratic
pathological behavior in humans when it does not as positive ones. If we condude that our dients and
already exist. Nevertheless, support for such a hy- students exhibit bizarre and potentially dangerous
pothesis can be found in several studies. Carr and behaviors to terminate instruction, we might ques-
Durand (1985) and Weeks and Gaylord-Ross tion whether or not our well-intentioned efforts to
(1981) showed that several different topographies teach are in our dients' best interest; at the very
of inappropriate behavior occurred more frequently least, we must question one or more aspects of our
during a "difficult task" condition when compared teaching technique. Perhaps most important from
to an "easy task" condition, suggesting that the the standpoint of contingencies, our ability to iden-
former condition contained aversive properties and tify negative reinforcement as a maintenance vari-
that the resulting behavior was escape- or avoid- able for undesirable behavior may directly influence
ance-motivated. Carr, Newsom, and Binkoff(1976) treatment selection and outcome. This is particu-
examined variables that apparently exerted stim- larly true with respect to extinction and time-out.
ulus control over the self-injurious behavior of a Their use typically calls for one or more therapist
psychotic boy. In one of their experiments, they responses (e.g., turning away from the client, re-
presented the boy with three alternating situations: moving stimul from immediate access, removing
a free-play period, a condition in which the exper- the dient from the setting, etc.) that terminate the
imenter spoke descriptive sentences to the child ongoing situation. Studies conducted with non-
(e.g., "The sky is blue")-this was called the "tact" human (Appel, 1963; Azrin, 1961; Thompson,

1964) and human (Plummer, Baer, & LeBlanc, chael, 1982, for an extended discussion of this
1977; Solnick, Rincover, & Peterson, 1977) sub- topic).
jects, however, indicate that the effects of time-out A more appropriate extinction procedure would
are highly dependent on features of the "time-in" entail continued presentation of the aversive stim-
environment. Thus, although time-out might be ulus or its cue and elimination of the consequence
an effective means of extinguishing most positively that was provided formerly (i.e., avoidance or es-
reinforced behavior, it might directly strengthen cape). In this manner, the basis for responding
negatively reinforced behavior. (aversive stimulation) remains, but reinforcement
does not (Bankart & Elliott, 1974; Coulson, Coul-
son, & Gardner, 1970; Davenport, Coger, & Spec-
TREATMENT OF NEGATIVELY tor, 1970; Schiff, Smith, & Prochaska, 1972).
REINFORCED BEHAVIOR Techniques derived from this type of extinction
A number of procedures based on the application actually have been used for a number of years in
of extinction, differential reinforcement, and pun- the treatment of clinical phobias and provide the
ishment have been evaluated as treatments for major theoretical basis for interventions collectively
problematic behavior of unspecified origin. Their known as "implosion therapies" (Levis, 1979).
use with behavior maintained by negative rein- An example of extinction for negatively rein-
forcement will be discussed in this section, along forced behavior was reported recently by Heidorn
with an additional procedure involving stimulus and Jensen (1984). After noting that demand-
fading. related situations were associated with an increase
in their subject's self-injurious behavior, a treatment
Extinction was developed that included the following: (a) con-
Traditional time-out will not provide for the tinued presentation of demands, (b) physical guid-
extinction of behavior that has been maintained by ance to complete the requested performance con-
negative reinforcement, but other procedures might. tingent on the occurrence of self-injury, (c)
One rather obvious possibility is elimination of the termination of the session contingent on compli-
supposed aversive stimulation and its related cues, ance, and (d) gradual increase in performance cri-
which should produce a reliable decrease in escape teria across sessions. Positive reinforcement in the
or avoidance behavior (Boren & Sidman, 1957; form of praise, food, and physical contact also was
Shnidman, 1968). However, as Hineline (1977) provided, but its role as an active component of
has noted, this procedure may not be a true ex- treatment may have been minimal. A similar pro-
tinction operation. The complete removal of aver- cedure was used in one of the experiments reported
sive stimulation during extinction of negatively by Carr et al. (1980) on the treatment of aggression.
reinforced behavior can be considered analogous to Extinction consisted of belting the subject in a chair
the continuous presence of, for example, food dur- to prevent escape while a therapist wearing pro-
ing extinction of positively reinforced behavior. Both tective gear sat across a table from him. The in-
procedures amount to noncontingent reinforce- tervention differed from that used by Heidorn and
ment, which removes the basis for responding and Jensen in that no attempt was made to deliver
indirectly reduces the frequency of behavior. That instructions during extinction sessions; instead, de-
is, if food is always present during extinction of mands were introduced after aggressive behavior
food-maintained behavior, there is no basis for re- was eliminated almost completely.
sponding; a similar situation exists if shock is always As with extinction of positively reinforced be-
absent during extinction of shock-avoidance be- havior, it is possible to foresee situations in which
havior. Following these procedures, food removal extinction of negatively reinforced behavior might
or, alternatively, reappearance of the shock should not be in the immediate best interest of either the
immediately produce the target response (see Mi- client (as in the case of severe self-injury) or the
therapist (as in the case of aggression). Extinction A second approach might involve appetitive re-
procedures may be compromised further by the inforcement for a competing behavior (DRI) with
potential effects of what procedurally may resemble the escape/avoidance contingency again operative.
noncontingent aversive stimulation (see earlier dis- Ruddle, Bradshaw, Szabadi, and Foster (1982)
cussion on acquisition of avoidance responding). studied human operant performance (button press-
To the extent that these "elicited" responses occur ing) using exactly this procedure. They presented
during the extinction of negatively reinforced be- subjects with concurrent avoidance/positive rein-
havior in applied situations, attempts to increase forcement schedules, and obtained matched re-
alternative behaviors, as well as to reduce the target sponding when the schedules were equated (this
behavior, may be disrupted. Finally, research show- was made possible by using points exchangeable
ing that time-based delivery of aversive stimulation for money). Performance shifts were correlated with
can maintain (Powell & Peck, 1969) and even schedule shifts roughly in a manner predicted by
increase (Kelleher, Riddle, & Cook, 1963; Sidman, Herrnstein's (1961) matching law. Our assessment
Herrnstein, & Conrad, 1957) the rates of avoidance research on self-injury (Iwata et al., 1982) provides
behavior suggests that schedule-related variables an approximation to the Ruddle et al. methodol-
and the subject's previous history may be important ogy. During one condition, we presented to subjects
considerations in the use of extinction. a series of instructional demands. Compliance was
followed by praise and physical contact from the
Differential Reinforcement experimenter, whereas the occurrence of self-injury
Applications of reinforcement to decrease a target produced a 30-s time-out. Data gathered during
behavior (differential reinforcement of other be- that study, as well as those collected since, indicate
havior [DROI, differential reinforcement of incom- that both responses are likely to occur; in other
patible behavior [DRI], etc.) are well documented words, positive reinforcement for compliance alone
in the applied literature, although the maintaining does not suppress avoidance-motivated self-injury.
variable for the target behavior rarely is noted. The Another example of differential reinforcement was
reinforcement contingency itself typically involves reported by Kelley, Jarvie, Middlebrook, McNeer,
the use of positive reinforcement, and discussion and Drabman (1984). They provided token rein-
here will be similarly confined. Applications of neg- forcement (stars) for reductions in the pain behavior
ative reinforcement will be addressed separately. (screaming, interfering, etc.) of two children
An experiment designed to examine the sup- undergoing open burn treatment. The procedure
pressive effects of differential reinforcement on neg- was moderately effective in that reductions in pain
atively reinforced behavior may take several forms. behavior averaged less than 50%. The findings of
First, access to an appetitive reinforcer (e.g., food) Iwata et al. and Kelley et al. are consistent with
could be made contingent on the absence of the those of Ruddle et al., indicating that positively
target behavior (DRO) while the escape/avoidance reinforced behavior competes with but does not
contingency is still operative. Although this ap- suppress avoidance or escape responding that is
proach might be considered unusual, it may resem- reinforced concurrently. In contrast, Carr et al.
ble very closely situations in the natural environ- (1980) were able to obtain almost complete elim-
ment in which DRO is implemented without ination of aggression in one of their subjects by
attempting to identify the behavior-maintaining introducing positive reinforcement to an existing
contingency. To my knowledge, this study has not demand situation. They did note, however, that
been reported in the basic literature, probably due their second subject was not responsive to the pos-
to difficulties associated with equating reinforce- itive reinforcement and that a different treatment
ment. It is possible that this type of study has been (see previous discussion of extinction) was used.
reported in the applied literature but that it was A third experiment might examine reinforce-
not explicitly identified. ment, as described in either of the above examples,

combined with extinction (continued presentation physical guidance to complete a requested perfor-
of aversive stimuli and prevention of escape). The mance, contingent on the occurrence of self-injury.
Heidorn and Jensen (1984) study on self-injury, It is interesting to note that the particular form of
described previously, is an example of this ap- stimulation used as punishment may have been
proach. From an applied perspective, their proce- exactly the same aversive stimulation whose prior
dures represent optimal treatment because contin- removal served as negative reinforcement; if so, the
gencies were provided for the inappropriate as well treatment amounted to a perfect reversal of the
as the appropriate behavior. However, here the maintaining contingency. Borreson (1980) also re-
effects of reinforcement are inseparable from those ported a case study of multiple treatment for avoid-
of extinction, and a dearer interpretation would ance-motivated self-injury; however, the punishing
require comparative analysis (reinforcement plus stimulus-"forced running" up and down a stair-
extinction vs. reinforcement alone vs. extinction way-appeared to be unrelated to the prior func-
alone). tion of the behavior.
The studies described here remain prototypical Punishment of negatively reinforced behavior
for the most part because very little research has presents significant complexities not found with
been reported on the use of differential (positive) positively reinforced behavior because it involves
reinforcement with escape and avoidance behavior. aversive stimulation following responses for which
On purely ethical grounds, and for the purposes of such stimulation already plays an important role as
establishing and strengthening alternative behav- an eliciting, discriminative, and motivating event
iors, the use of positive reinforcement seems criti- (for extensive reviews of this topic, see Davis, 1979;
cally important. On the other hand, its therapeutic Fowler, 1971; Hineline, 1981; and Morse & Kel-
effects as primary treatment for negatively rein- leher, 1970, 1977). The major issues are sum-
forced behavior have yet to be demonstrated. Based marized here. First, the eliciting properties of aver-
on the small amount of data available, one might sive stimulation, described previously with respect
expect that positive reinforcement is more likely to to acquisition and extinction, are relevant in the
produce beneficial results if the negatively rein- case of punishment. Although elicited behavior may
forced behavior is extinguished concurrently or if not necessarily compromise the use of punishment,
the density of positive reinforcement is noticeably it may have a deleterious effect on the overall treat-
greater than that of the negative reinforcement. ment program. Second, punishment with the same
stimulus used during escape or avoidance training
Punishment may acquire discriminative properties for respond-
Contingent aversive stimulation for negatively ing as a result of reinstating the conditions under
reinforced behavior is the functional complement which escape originated, thereby occasioning the
of DRO for positively reinforced behavior, in that very behavior being punished. For example, several
prevention of aversive stimulation (negative rein- studies have shown response maintenance and even
forcement) is contingent on the absence of respond- facilitation when shock-preventing behavior was
ing. Procedural curiosities aside, we know very little followed by the presentation of shock (e.g., Appel,
about the effects of punishment on human escape 1960; Sandier, Davidson, & Malagodi, 1966).
and avoidance, in spite of the many applied studies Third, schedule-related variables can determine
on punishment published to date. For example, the whether contingent stimulation serves as either pun-
literature on self-injury, in which most of the current ishment or reinforcement. Kelleher and Morse
applied research on punishment can be found, con- (1968) and McKearney (1972) found that re-
tains only two studies reporting the use of punish- sponding developed as avoidance behavior was sup-
ment for behavior described as avoidance motivat- pressed under dense schedules of punishment but
ed. One of the elements in the Heidorn and Jensen facilitated under thinner schedules. Finally, it has
(1984) multiple-treatment approach consisted of been noted that punishment intensity and the pres-
ence or absence of avoidance contingencies may sponding does not occur, (c) presentation of the
have an interactive effect on behavior. Sandier, Da- altered stimuli with a gradual return to their orig-
vidson, Greene, and Holzschuh (1966) imposed inal state, and (d) extinction of escape behavior.
high-, intermediate-, and low-intensity shock as Approximations to the stimulus fading approach
punishment for ongoing avoidance behavior and can be found in two studies previously discussed.
found greater response persistence under the high- Heidorn and Jensen (1984) decreased and then
intensity condition. However, when the avoidance gradually increased the frequency with which re-
contingency was later removed (i.e., responding sponse-producing stimuli (demands) were pre-
produced shock but no longer prevented it), the sented, although they did not withdraw them en-
high-intensity condition produced the most rapid tirely at the beginning of treatment. In contrast,
response suppression. Carr et al. (1976) were able to reduce self-injury
The use of punishment should always be con- by embedding demands within entertaining stories,
sidered very carefully, and even greater caution although the stories were never faded out nor were
should be taken when there is reason to believe additional demands faded in. The results of both
that the target behavior has been maintained by studies suggest that more complete evaluations of
negative reinforcement. Findings from the basic treatment based on fading are warranted. One po-
research literature suggest, although in a very ten- tential advantage of fading over extinction and pun-
tative manner, that a stimulus different from that ishment might be the complete elimination of es-
associated with prior avoidance should be used, cape behavior from the outset of treatment.
that the schedule of punishment should be a con-
tinuous one, and that "mildly aversive" stimuli
may produce greater response suppression than more NEGATIVE REINFORCEMENT AS
intense stimulation when the prevailing avoidance THERAPY
contingency remains operative. On the other hand, I have noted previously that free-time contin-
data from the Heidorn and Jensen (1984) study gencies might function as negative reinforcement,
indicated that, within the context of their multiple- although that possibility has been seldom acknowl-
treatment approach, the relationship between stim- edged. Free time also may be one of the few con-
ulation used as negative reinforcement and punish- tingencies in the applied literature that represents
ment may not be an important one. pure escape in that the stimulation (work) is rel-
atively continuous and can be reduced or terminated
Stimulus Fading but not avoided. A great majority of applications
In contrast to approaches in which a contingency make use of time- or trial-based presentation of
is directly manipulated, fading consists of altering stimuli preceded by cues, which produces avoidance
one or more features of stimuli that occasion the behavior. Examples of negative reinforcement used
target behavior. Various types of stimulus fading to strengthen desirable behavior will be discussed
have been used for many years in the treatment of in this section, grouped according to similarities in
clinical fears and phobias, dating back to the work either procedure or problem.
of Jones (1924). Contemporary formulations differ
greatly along procedural dimensions (actual vs. rep- Behavioral Engineering
resentational stimulus presentation, the presence or The earliest examples of negative reinforcement
absence of reinforcement and punishment) as well to develop or maintain appropriate behavior pub-
as on underlying theory (respondent vs. operant lished in JABA made use of apparatus-delivered
conditioning). The operant model of stimulus fad- stimulation. Azrin and his colleagues conducted
ing to reduce escape or avoidance behavior involves two such studies. The first (Azrin, Rubin, O'Brien,
(a) initial identification of response-producing stim- Ayllon, & Roll, 1968) established automated mea-
uli, (b) stimulus alteration to the point where re- surement and control over postural slouching. An

apparatus built into a shoulder harness operated s by a 95-db tone. By remaining dry, the child
closure of a circuit when slouching occurred. This could avoid the first tone; if the child urinated, he
action produced an audible click, followed 3 s later or she could escape the first tone and avoid the
by a 5 5-db tone. The contingency was an ingenious second by immediately getting out of bed and turn-
one in that it included aspects of both free-operant ing the unit off. O'Brien, Ross, and Christophersen
and discriminated avoidance plus escape. Mainte- (1986) recently used negative reinforcement to pro-
nance of correct posture (free-operant) avoided the duce elimination. As part of an overall treatment
dick, postural correction (discriminated) during the program for four encopretic children, the authors
3 s following the dick avoided the tone, and cor- wanted to establish morning control over bowel
rection during the tone provided escape. Subjects movements. To do so, parents had the children sit
consisted of 25 adults, all of whom showed re- on the toilet each morning for 5 min; failure to
ductions in slouching while the device was worn. defecate a minimum equivalent of one-fourth of a
When the contingency was reversed for two of the cup during that time was followed immediately by
subjects, both showed increases in slouching. In a insertion of a suppository. A second administration
later study, Azrin and Powell (1969) evaluated an was given if the first did not produce the desired
apparatus to increase pill taking in six subjects. The outcome.
pill dispenser produced a 50-db tone every 30 min,
an arbitrary between-pill interval. The tone could Overcorrection
be turned off by pushing a knob on the case, which Originally designed as a means for eliminating
also delivered two pills. One might expect that this accidents during toilet training (Azrin & Foxx,
arrangement would produce escape behavior ini- 1971), overcorrection consists of a group of tech-
tially, followed by free-operant avoidance, although niques whose common feature is repetitive perfor-
data to that effect were not presented. A third study mance of motor activity. Overcorrection is one of
involving apparatus-delivered negative reinforce- the most thoroughly studied and frequently used
ment was conducted by Greene and Hoats (1969). methods for reducing the frequency of a wide range
Their subject was an adult male assigned to a cor- of undesirable behaviors (see Foxx & Bechtel, 1983,
rectional unit whose task was to sort computer for a review). The procedures are considered to be
cards, for which he earned cigarettes. He also was derivatives of punishment and are applied contin-
allowed to watch TV while performing the task. gent upon the occurrence of a target behavior. At
During the treatment condition, if the subject did the same time, overcorrection can serve as negative
not complete a task cycle within a specified interval reinforcement in at least two ways. First, because
of time (avoidance), visual and auditory output the procedure calls for performance of activities that
from the TV were distorted and remained that way apparently are aversive, a therapist always is at hand
until the work cycle was completed (escape). to ensure compliance through continued instruction
and, ifnecessary, physical guidance. Thus, the client
Toilet Training and Incontinence can avoid repeated instructions and potentially in-
The presence or absence of elimination is more trusive physical contact through continued perfor-
than the occurrence or nonoccurrence of a response. mance of the required activity. Avoidance behavior
Sphincter contraction as well as relaxation is in- is also produced when overcorrection is applied
volved, and negative reinforcement has been used contingent upon the absence of a desirable response.
in the management of both behaviors. Hansen For example, Foxx (1977) showed that 5 min of
(1979) incorporated escape and avoidance contin- functional movement training, involving the prac-
gencies in the treatment of nocturnal enuresis in tice of varying head positions, was superior to food
two children. When a device placed in the bed and praise in developing and maintaining eye con-
detected urination, an apparatus located 4 ft away tact in three retarded children. Examination of the
produced a 70-db tone, which was followed in 7 overcorrection literature yields a number of in-
stances in which the procedure was used to increase with statements ofdisapproval (Dunlap & Johnson,
rather than decrease behaviors, including class at- 1985; Rincover & Newsom, 1985; Schreibman,
tendance (Foxx, 1976), repetitive tasks (Carey & 1975), physical guidance (Haring, 1985; Luyben,
Bucher, 1983), sharing (Barton & Osborne, 1978), Funk, Morgan, Clark, & Delulio, 1986; Sprague
speech (Matson, Esveldt-Dawson, & O'Connell, & Homer, 1984), session-lengthening procedures
1979), and spelling accuracy (Foxx & Jones, 1978; consisting of either time-out (Barrera & Sulzer-
Ollendick, Matson, Esveldt-Dawson, & Shapiro, Azaroff, 1983; O'Brien & Azrin, 1972) or remedial
1980). learning trials (Nutter & Reid, 1978; Page, Iwata,
& Neef, 1976; Richman, Reiss, Bauman, & Bailey,
Error Correction During Instruction 1984), and so on. Thus, in addition to producing
Rodgers and Iwata (1987) recently conducted a positive reinforcement in the form of experimenter
survey whose initial focus was on response prompt- praise, correct responses also may function to avoid
ing as an adjunct during behavioral acquisition. aversive social and physical stimulation and to ef-
We quickly determined that most prompting oc- fectively reduce the duration of training sessions
curs following an error and expanded the analysis (this latter point is potentially significant, for it has
to include all events that can be made contingent been shown that complex setting events or stimulus
on incorrect responses. Negative reinforcement was situations, and not just discrete stimuli, can function
not a subject of interest at the outset, but some of as negative reinforcers [Krasnegor, Brady, & Find-
the techniques that were found suggest that it plays ley, 1971], and that reduction of avoidance-session
a much more prominent role in the instructional durations can itself serve as negative reinforcement
process than is currently acknowledged. [Mellitz, Hineline, Whitehouse, & Laurence,
The most dramatic example is a study by Kirch- 1983]).
er, Pear, and Martin (1971) entitled "Shock as
punishment [emphasis added) in a picture-naming Behavioral Replacement Strategies
task with retarded children." In one experiment, Given that aversive stimuli are ubiquitous and
two children were exposed to the following two that escape is highly adaptive in their presence, it
treatments: (a) token reinforcement for correct pic- is usually the form, rather than the function, of
ture-naming responses, and (b) token reinforcement escape and avoidance behavior that presents a prob-
for correct responses plus shock for either errors or lem. This raises the possibility of eliminating in-
a response latency greater than 5 s. The shock appropriate forms of escape by negatively reinforc-
condition produced superior results. Because the ing appropriate alternatives; in essence, replacing
token reinforcement remained constant across the one behavior with another but not eliminating the
two conditions and because, regardless of how one function of the original. The concept is rather
defines the target behavior (i.e., errors vs. correct straightforward in principle and is analogous to
responses, inattention vs. attention), the desirable lever switching by nonhuman subjects following a
performance was a correct picture name, the pro- change in reinforcement schedule (e.g., De Villiers,
cedure dearly represents an avoidance contingency 1974).
in that correct responses made within 5 s of the Carr and Durand (1985) recently provided an
cue prevented the delivery of the shock. example of this strategy by teaching three children,
The Kircher et al. (1971) study represents a who had tantrums when faced with difficult tasks,
rather extreme use of negative reinforcement to how to request help from the teacher. When a child
increase desirable behavior, one that cannot be de- exhibited the appropriate response ("I don't un-
fended on ethical grounds today. However, less derstand"), brief escape was provided in the form
dramatic but analogous situations are quite com- of teacher assistance.
mon in the literature on instructional technology. Another example is drawn from our work on
It has become standard practice to follow errors feeding disorders. While attempting to increase the

oral acceptance of food in four children, Riordan, chronic and withdrawn schizophrenic male. The
Iwata, Finney, Wohl and Stanley (1984) found researchers targeted three aspects of his conversa-
that one child, who was fed through a gastrostomy tional behavior: voice loudness, duration of verbal
tube, did not respond well to positive reinforcement responding, and keeping his hands on the armrests
because her baseline rate of acceptance was virtually of his chair while speaking. A therapist approached
nonexistent. She also had resisted a number of the client, called his name, and asked him to con-
forced feeding regimens; these practices were aver- verse about one of several predetermined topics.
sive and it appeared that her success in defeating During treatment, the staff member continuously
them constituted negative reinforcement. Her treat- monitored the target behaviors; if any failed to meet
ment consisted of the following components: (a) criterion, the staff member would nag loudly one
the presentation of a redundant cue-"Take a or more of the following: "Longer!," "Louder!,"
bite"- immediately followed by (b) the presen- or "Put your hands on the armrests of the chair!,"
tation of food on a spoon. If acceptance of food and would continue to nag at 3-s intervals until
did not occur within 3 s, (c) her mouth was held the target behavior occurred. Although effective, it
open and the food was deposited. This procedure should be noted that the contingencies used by
thus resembled very dosely a discriminated avoid- Fichter et al. would not be considered typical con-
ance contingency in which one avoidance behavior sequences for social interaction; in fact, it is entirely
(active food refusal) was replaced with another possible that the appearance of the therapist would
(opening the mouth and accepting food) by allow- become discriminative for withdrawal. Data similar
ing it to prevent forced feeding. to those provided by Lovaas et al. (1965) showing
that social behavior (e.g., conversations initiated by
Other Examples the subject or his response to approach by a ther-
In addition to the Kircher et al. (1971) study apist) improved outside of treatment sessions would
on academic performance described earlier, one can have been informative with respect to this question.
find instances in which negative reinforcement has One can only assume that there was no generalized
been used-in a highly intrusive manner-to in- improvement, based on a comment made by the
crease appropriate social behaviors. Lovaas, Schaef- authors:
fer, and Simmons (1965) used escape and avoid- i. ... his [the subject's) last interaction before
ance in training two autistic children to approach ... [being discharged from the) ... unit was
adults. Prior to treatment, the children frequendy to tell one staff member how much he disliked
engaged in stereotypic behavior and showed no the unit and the staff" (Fichter et al., 1976,
social responsivity or appropriate play. Treatment pp. 384-385).
consisted of presenting the instruction, "Come here,"
followed by shock delivered through a floor grid.
The shock was terminated when the child moved SUMMARY
toward the therapist (escape). Both children quickly
learned to approach the adult in response to the In this artide I have attempted to point out a
verbal instruction (avoidance). In defense of their number of ways in which negative reinforcement
use of electric-shock avoidance, the authors pre- is relevant to behavioral development and its sub-
sented data indicating that increases in approach sequent modification in the applied situation. My
behavior were accompanied by more frequent dis- review has been a selective one in that I made no
plays of affection and decreases in stereotypy and attempt to summarize the large and varied literature
aggression. on aversion, implosion, and desensitization thera-
A less intrusive but similar procedure was used pies often used in the treatment of alcoholism,
by Fichter, Wallace, Liberman, and Davis (1976) smoking, phobic reactions, sexual disorders, and
in an attempt to improve the social skills of a related clinical problems. Still, the applied examples
represent a thorough cross section of research pub- gency can be found. In other cases, however, im-
lished in JABA over the past 20 years, and a plementation of treatment without consideration of
number of general condusions and implications can developmental factors, or treatment selection based
be drawn from the work described here. on a consideration of topography alone, may pro-
Historically, applied analyses of behavior have duce a number of unnecessary failures and subse-
failed to acknowledge escape and avoidance as po- quently may limit our ability to determine the basis
tentially common and powerful sources of rein- for differential outcome.
forcement. Evidence of this can be found in work Laboratory researchers solve problems related to
on severe behavioral disorders such as self-injury, history by controlling its course in a naive animal.
in which discussions of etiology have focused pri- Although applied researchers rarely can exercise this
marily on attentional factors rather than on those option, they can make a unique contribution by
related to escape; in research on contingencies such developing methods for "unravelling" behavioral
as free time, in which free time as negative rein- history, to the extent that it is possible. Our real-
forcement has not been a subject of analysis; in ization that behavioral development through neg-
studies in which avoidance contingencies have been ative reinforcement can produce the same topog-
inaccurately described as punishment; and in re- raphy as that resulting from positive reinforcement,
search on instructional processes, in which a variety time and time again, may provide the impetus for
of avoidance contingencies have been used but not continued refinement in the analysis of behavioral
evaluated or even described as such. function. For example, a number of researchers
A second and more optimistic condusion sup- have conduded that unitary accounts of severe be-
portable by work described here is that an applied havioral disorders in the developmentally disabled
technology of negative reinforcement is emerging. are unsatisfactory and have begun to establish
The work is somewhat scattered at present and methodologies for identifying the functional prop-
little is known in some areas. Nevertheless, under erties (one being negative reinforcement) of disor-
each of the topics induded in the present discus- ders such as pica (Mace & Knight, 1986), self-
sion-behavioral development, treatment of neg- injury (Carr & Durand, 1985; Iwata et al., 1982),
atively reinforced behavior, and therapeutic uses of and stereotypy (Durand & Carr, 1987). Most re-
negative reinforcement-research activity has in- cently, Bailey (1987) has proposed the term "be-
creased in recent years, and we are beginning to see havioral diagnostics" to describe a general strategy
investigations of common procedures to the point for isolating the bases of problematic behavior.
where categorization is both possible and useful. Continued work in this area and extension of the
This evidence ofgrowing interest suggests that neg- relevant methodologies to other human problems
ative reinforcement may be one of the most sig- are essential if we are to develop a mature tech-
nificant areas of applied research during the coming nology of behavior. In the meantime, researchers
years. Having made that prediction, what remains should be encouraged (perhaps by editors) to seek
is to offer some prompts to help ensure its accuracy. out and indude more detail on subjects' behavioral
The area of behavioral development is particu- histories. In addition to the usual demographics
larly problematic because applied researchers often offered (e.g., age, sex, grade or functioning level,
are faced with situations in which the behavior of etc.), which provide little information relevant to
interest has a long, complex, and unknown history. a behavioral analysis, it would be helpful to provide
In fact, the most important difference between lab- some account of factors related to behavioral de-
oratory- and field-based research, at least from a velopment and maintenance. As evidence support-
behavioral standpoint, is the lack of control over ing a negative reinforcement interpretation accu-
history that is characteristic of applied research. In mulates, we will be increasingly compelled to
some cases (perhaps even most), behavioral history formalize our anecdotal observations and to confirm
may be irrelevant if a sufficiently powerful contin- these observations through manipulation.

Because applied research often is concerned with of the effectiveness of commonly used positive rein-
a problem as it actually exists, the treatment of forcers (the results of this estimate may indicate
negatively reinforced behavior will provide perhaps that more potent reinforcers are needed), (b) we
the greatest opportunity for creative work. For ex- can determine whether procedures such as remedial
ample, research on the extinction of positively rein- trials, physical assistance, and so on, serve any usefil
forced behavior has induded variation and exten- function and if that function is one of negative
sion (e.g., time-out, exclusion, sedusion, contingent reinforcement, and (c) we can base future training
observation, time-out ribbon, movement suppres- successes on the planned rather than the accidental
sion), parametric analysis (e.g., duration, delay, use of negative reinforcement. A third promising
schedule, changeover requirement), and compari- application involves further elaboration of behav-
son (e.g., with differential reinforcement, response ioral replacement strategies. If we are willing to
cost, and punishment). None of these questions entertain the assumption that it is impossible to
have been addressed adequately in applied research eliminate all sources of aversive stimulation, the use
on the extinction of negatively reinforced behavior, ofsuch stimulation to alter the topography ofescape
and a similar situation exists with respect to dif- and avoidance behavior, from an undesirable one
ferential reinforcement and punishment. to a tolerable one, makes eminent sense from a
Research on the treatment of negatively rein- clinical standpoint.
forced behavior will require consideration of issues A final cautionary note. Some of the applied
that are different than those relevant to the treat- research induded in this review was selected spe-
ment of positively reinforced behavior. These issues cifically to show that negative reinforcement can
have been noted previously, and some have been form the basis of highly intrusive intervention. In
the focus of laboratory research for several years. at least one sense, negative reinforcement might be
Stimulus selection, schedules, and intensity, for ex- considered more intrusive than punishment be-
ample, may differentially affect the outcome of con- cause, with negative reinforcement, presentation of
tingent aversive stimulation for ongoing avoidance the aversive stimulus is contingent on the absence,
behavior. Therefore, it will be important for applied rather than the occurrence, of behavior. Therefore,
researchers to become acquainted with basic find- as with punishment, we should conduct research
ings on negative reinforcement. As a result, we may on negative reinforcement with great care and under
find that methodologies and procedures developed the appropriate conditions to determine how it might
in the laboratory can be extended to the applied be used effectively and humanely, its limitations,
situation so as to facilitate analysis and treatment. and its proper role within the larger realm of cur-
Research on the use of negative reinforcement rently available treatment.
may take several interesting directions. First, neg-
ative reinforcement may provide an alternative
means for establishing behavior when attempts to REFERENCES
use positive reinforcement fail (e.g., as in the case Aaron, B. A., & Bostow, D. E. (1978). Indirect facilitation
of Riordan et al., 1984, in which a child's operant of on-task behavior produced by contingent free-time for
level of eating was nonexistent). If so, we will want academic productivity. Journal of Applied Behavior
Analysis, 11, 197.
to know the behaviors for which specific contin- Appel, J. B. (1960). Some schedules involving aversive
gencies are useful and the conditions under which control. Journal of the Experimental Analysis of Be-
they should be applied. Second, it appears that the havior, 3, 349-359.
Appel, J. B. (1963). Aversive aspects of a schedule of
acquisition of adaptive behavior in our training positive reinforcement. Journal of the Experimental
programs is at least partially a function of negative Analysis of Behavior, 6, 423-430.
reinforcement. Future research must evaluate the Axeirod, S., & Apsche,J. (1983). The efects ofpunishment
on human behavior. New York: Academic Press.
roles of escape and avoidance within the training Azrin, N. H. (1961). Time-out from positive reinforce-
context so that (a) we will have a proper estimate ment. Science, 133, 382-383.

Azrin, N. H., & Foxx, R. M. (1971). A rapid method of Carr, E. G., & Durand, M. V. (1985). Reducing behavior
toilet training the institutionalized retarded. Journal of problems through functional communication training.
Applied Behavior Analysis, 4, 89-99. Journal of Applied Behavior Analysis, 18, 111-126.
Azrin, N. H., & Powell,J. (1969). Behavioral engineering: Carr, E. G., Newsom, C. D., & Binkoff, J. A. (1976).
The use of response priming to improve prescribed self- Stimulus control of self-destructive behavior in a psy-
medication. Journal of Applied Behavior Analysis, 1, chotic child. Journal of Abnormal Child Psychology, 4,
99-108. 139-153.
Azrin, N., Rubin, H., O'Brien, F., Ayllon, T., & Roll, D. Carr, E. G., Newsom, C. D., & Binkoff, J. A. (1980).
(1968). Behavioral engineering: Postural control by a Escape as a factor in the aggression of two retarded
portable operant apparatus.Journal of Applied Behavior children. Journal of Applied Behavior Analysis, 13,
Analysis, 2, 39-42. 101-117.
Baer, A. M., Rowbury, T., & Baer, D. M. (1973). The Catania, A. C. (1973). The nature of learning. In J. A.
development of instructional control over classroom ac- Nevin & G. S. Reynolds (Eds.), The study of behavior:
tivities of deviant preschool children. Journal of Applied Learning, motivation, emotion, and instinct (pp. 31-
Behavior Analysis, 6, 289-298. 68). Glenview, IL: Scott, Foresman.
Baer, D. M. (1978). On the relation between basic and Coulson, G., Coulson, V., & Gardner, L. (1970). The effect
applied research. In A. C. Catania & T. A. Brigham of two extinction procedures after acquisition on a Sid-
(Eds.), Handbook of applied behavior analysis: Social man avoidance contingency. Psychonomic Science, 18,
and instructional processes (pp. 11-17). New York: 309-310.
Irvington. Cullen, C. (1981). The flight to the laboratory. The Be-
Baer, D. M. (1981). A flight of behavior analysis. The havior Analyst, 4, 81-83.
Behavior Analyst, 4, 85-9 1. Davenport, D. G., Coger, R. W., & Spector, 0. J. (1970).
Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some The redefinition of extinction applied to Sidman free-
current dimensions of applied behavior analysis. Journal operant avoidance responding. Psychonomic Science, 19,
of Applied Behavior Analysis, 1, 91-97. 181-182.
Bailey, J. S. (1987, May). Behavioral diagnostics: New Davis, H. (1979). Behavioral anomolies in aversive situ-
tools for applied behavior analysis. Presented as an in- ations. In J. D. Keehn (Ed.), Psychopathology in ani-
vited address at the Association of Behavior Analysis mals: Research and clinical implications (pp. 197-
Convention, Nashville, TN. 222). New York: Academic Press.
Bankart, B., & Elliott, R. (1974). Extinction of avoidance Deitz, S. M. (1978). Current status of applied behavior
in rats: Response availability and stimulus presentation anlaysis: Science versus technology. American Psychol-
effects. Behaviour Research and Therapy, 12, 53-56. ogist, 33, 805-814.
Barrera, R. D., & Sulzer-Azaroff, B. (1983). An alternating De Villiers, P. A. (1974). The law of effect and avoidance:
treatment comparison of oral and total communication A quantitative relationship between response rate and
training programs with echolalic autistic children. Jour- shock-frequency reduction. Journal of the Experimental
nal of Applied Behavior Analysis, 16, 379-394. Analysis of Behavior, 21, 223-235.
Barrish, H. H., Saunders, M., & Wolf, M. M. (1969). Dunlap, G., & Johnson, J. (1985). Increasing the inde-
Good behavior game: Effects of individual contingencies pendent responding of autistic children with unpredict-
for group consequences on disruptive behavior in a class- able supervision.Journal of Applied Behavior Analysis,
room. Journal of Applied Behavior Analysis, 2, 119- 18, 227-236.
124. Durand, V. M., & Carr, E. G. (1987). Social influences
Barton, E. S., & Osbome,J. G. (1978). The development on "self-stimulatory" behavior. Journal of Applied Be-
of classroom sharing by a teacher using positive practice. havior Analysis, 20, 119-132.
Behavior Modification, 2, 231-250. Ferrari, E. A., Todorov, J. C., & Graeff, F. G. (1973).
Bolles, R. G. (1970). Species-specific defense reactions and Nondiscriminated avoidance ofshock by pigeons pecking
avoidance learning. Psychological Review, 77, 32-48. a key. Journal of the Experimental Analysis of Behav-
Bolles, R. G. (1971). Species-specific defense reactions. In ior, 19, 211-218.
F. R. Brush (Ed.), Aversive conditioning and learning Fichter, M. M., Wallace, C. J., Liberman, R. P., & Davis,
(pp. 183-233). New York: Academic Press. J. R. (1976). Improving social interaction in the chron-
Boren,J. J., & Sidman, M. (1957). A discrimination based ic psychotic using discriminated avoidance ("nagging"):
on repeated conditioning and extinction of avoidance Experimental analysis and generalization.Journal of Ap-
behavior. Journal of Comparative and Physiological plied Behavior Analysis, 9, 367-386.
Psychology, 50, 18-22. Foree, D., & LoLordo, V. (1970). Signalled and unsignalled
Borreson, P. M. (1980). The elimination of a self-injurious free-operant avoidance in the pigeon. Journal of the
avoidance response through a forced running conse- Experimental Analysis of Behavior, 13, 283-290.
quence. Mental Retardation, 18, 73-77. Fowler, H. (1971). Suppression and facilitation by response
Carey, R. G., & Bucher, B. (1983). Positive practice over- contingent shock. In R. F. Brush (Ed.), Aversive con-
correction: The effects of duration of positive practice on ditioning and learning (pp. 537-604). New York: Ac-
acquisition and response reduction. Journal of Applied ademic Press.
Behavior Analysis, 16, 101-109. Foxx, R. M. (1976). Increasing a mildly retarded woman's
attendance at self-help dasses by overcorrection and in- Hutchinson, R. R. (1977). By-products of aversive control.
struction. Behavior Therapy, 6, 390-396. In W. K. Honig & J. E. R. Staddon (Eds.), Handbook
Foxx, R. M. (1977). Attention training: The use of over- of operant behavior (pp. 415-431). Englewood Cliffs,
correction avoidance to increase the eye contact of autistic NJ: Prentice-Hall.
and retarded children.Journal of Applied Behavior An- Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E.,
laysis, 10, 488-499. & Richman, G. S. (1982). Toward a functional analysis
Foxx, R. M., & Bechtel, D. R. (1983). Overcorrection: A of self-injury. Analysis and Intervention in Develop-
review and analysis. In S. Axelrod & J. Apsche (Eds.), mental Disabilities, 3, 1-20.
The efects ofpunishment on human behavior (pp. 133- Jones, M. C. (1924). Elimination of children's fears. Jour-
220). New York: Academic Press. nal of Experimental Psychology, 7, 382-390.
Foxx, R. M., &Jones,J. R. (1978). A remediation program Kelleher, R. T., Riddle, W. C., & Cook, L. (1963). Per-
for increasing spelling achievement of elementary and sistent behavior maintained by unavoidable shocks.Jour-
junior high school students. Behavior Modification, 2, nal of the Experimental Analysis of Behavior, 6, 507-
211-230. 517.
Greene, R. R., & Hoats, D. L. (1969). Reinforcing ca- Kelleher, R. T., & Morse, W. H. (1968). Schedules using
pabilities of television distortion. Journal of Applied noxious stimuli, III: Responding maintained with re-
Behavior Analysis, 2, 139-141. sponse-produced electric shocks. Journal of the Exper-
Hansen, G. D. (1979). Enuresis control through fading, imental Analysis of Behavior, 11, 819-838.
escape, and avoidance training. Journal of Applied Be- Kelley, M. L., Jarvie, G. J., Middlebrook, J. L., McNeer,
havior Analysis, 12, 303-307. M. F., & Drabman, R. S. (1984). Decreasing burned
Haring, T. G. (1985). Teaching between-dass generaliza- children's pain behavior: Impacting the trauma of hy-
tion of toy play behavior to handicapped children. Jour- drotherapy. Journal of Applied Behavior Analysis, 17,
nal of Applied Behavior Analysis, 18, 127-139. 147-158.
Harris, V. W., & Sherman,J. A. (1973). Use and analysis Kircher, A. S., Pear,J.J., & Martin, G. L. (1971). Shock
of the "good behavior game" to reduce disruptive dass- as punishment in a picture-naming task with retarded
room behavior. Journal of Applied Behavior Analysis, children.Journal of Applied Behavior Analysis, 4, 227-
6, 405-417. 233.
Harris, V. W., & Sherman, J. A. (1974). Homework Krasnegor, N. A., Brady, J. V., & Findley, J. D. (1971).
assignments, consequences, and classroom performance Second-order optional avoidance as a function of fixed-
in social studies and mathematics. Journal of Applied ratio requirements. Journal of the Experimental Anal-
Behavior Analysis, 7, 505-519. ysis of Behavior, 15, 181-187.
Heidorn, S. D., & Jensen, C. C. (1984). Generalization Levis, D. J. (1979). The infrahuman avoidance model of
and maintenance of the reduction of self-injurious be- symptom maintenance and implosive therapy. In J. D.
havior maintained by two types of reinforcement. Be- Keehn (Ed.), Psychopathology in animals: Research and
havior Research and Therapy, 22, 581-586. clinical implications (pp. 257-277). New York: Aca-
Herrnstein, R. J. (1961). Relative and absolute strength demic Press.
of response as a function of frequency of reinforcement. Long, J. D., & Williams, R. W. (1973). The comparative
Journal of the Experimental Analysis of Behavior, 4, effectiveness of group and individual contingent free time
267-272. with inner-city junior high school students. Journal of
Herrnstein, R. J. (1969). Method and theory in the study Applied Behavior Analysis, 6, 465-474.
of avoidance. Psychological Review, 76, 49-69. Lovaas, 0. I., Schaeffer, B., & Simmons, J. Q. (1965).
Hineline, P. N. (1977). Negative reinforcement and avoid- Building social behavior in autistic children by use of
ance. In W. K. Honig & J. E. R. Staddon (Eds.), Hand- electric shock. Journal of Experimental Research in Per-
book of operant behavior (pp. 364-414). Englewood sonality, 1, 99-109.
Cliffs, NJ: Prentice-Hall. Luyben, P. D., Funk, D. M., Morgan, J. K., Clark, K. A.,
Hineline, P. N. (1981). Several roles of stimuli in negative & Delulio, D. W. (1986). Team sports for the retarded:
reinforcement. In P. Harzem & M. D. Zeiler (Eds.), Training a side-of-the-foot soccer pass using a maximum-
Advances in analysis of behavior: Vol. 2. Predict- to-minimum prompt reduction strategy. Journal of Ap-
ability, correlation, and contiguity (pp. 203-246). plied Behavior Analysis, 19, 431-436.
Chichester, England: Wiley. Mace, F. C., & Knight, D. (1986). Functional analysis and
Hineline, P. N. (1984). Aversive control: A separate do- treatment of severe pica. Journal of Applied Behavior
main? Journal of the Experimental Analysis of Behav- Analysis, 19, 411-416.
ior, 42, 495-509. MacPhail, E. M. (1968). Avoidance responding in pigeons.
Hoffman, H. S. (1966). The analysis of discriminated Journal of the Experimental Analysis of Behavior, 11,
avoidance. In W. K. Honig (Ed.), Operant behavior: 629-632.
Areas of research and application (pp. 499-530). New Maloney, K. B., & Hopkins, B. L. (1973). The modifi-
York: Appleton. cation of sentence structure and its relationship to sub-
Honig, W. K., & Staddon,J. E. R. (Eds.). (1977). Hand- jective judgments of creativity in writing. Journal of
book of operant behavior. Englewood Cliffs, NJ: Prentice- Applied Behavior Analysis, 6, 425-433.
Hall. Matson, J. L., & DiLorenzo, T. M. (1984). Punishment

and its alternatives: A new perspective for behavior Plummer, S., Baer, D. M., & LeBlanc,J. M. (1977). Func-
modification. New York: Springer. tional considerations in the use of timeout and an effective
Matson,J. L., Esveldt-Dawson, K., & O'Connell, D. (1979). alternative. Journal of Applied Behavior Analysis, 10,
Overcorrection, modeling, and reinforcement procedures 689-705.
for reinstating speech in a mute boy. Child Behavior Powell, R. W., & Peck, S. (1969). Persistent shock-elicited
Therapy, 1, 363-371. responding engendered by a negative reinforcement pro-
McKearney, J. W. (1972). Maintenance and suppression cedure. Journal of the Experimental Analysis of Be-
of responding under schedules of electric shock presen- havior, 12, 1049-1062.
tation. Journal of the Experimental Analysis of Be- Rachlin, H., & Hineline, P. N. (1967). Training and
havior, 17, 425-432. maintenance of keypecking in the pigeon using negative
Medland, M. B., & Stachnik, T. J. (1973). Good behavior reinforcement. Science, 157, 954-955.
game: A replication and systematic analysis. Journal of Richman, G. S., Reiss, M. L., Bauman, K. E., & Bailey, J.
Applied Behavior Analysis, 6, 45-51. S. (1984). Teaching menstrual care to mentally re-
Mellitz, M., Hineline, P. N., Whitehouse, W. G., & Lau- tarded women: Acquisition, generalization, and main-
rence, M. T. (1983). Duration-reduction of avoidance tenance. Journal of Applied Behavior Analysis, 17,
sessions as negative reinforcement. Journal of the Ex- 441-451.
perimental Analysis of Behavior, 40, 57-67. Rincover, A., & Newsom, C. D. (1985). The relative
Michael, J. (1975). Positive and negative reinforcement: motivational properties of sensory and edible reinforcers
A distinction that is no longer necessary; Or a better way in teaching autistic children. Journal of Applied Be-
to talk about bad things. Behaviorism, 3, 33-44. havior Analysis, 18, 237-248.
Michael, J. (1980). Flight from behavior analysis. The Riordan, M. M., Iwata, B. A., Finney, J. W., Wohl, M. K.,
Behavior Analyst, 3, 1-24. & Stanley, A. E. (1984). Behavioral assessment and
Michael, J. (1982). Distinguishing between discriminative treatment of chronic food refusal in handicapped chil-
and motivational functions of stimuli. Journal of the dren. Journal of Applied Behavior Analysis, 17, 327-
Experimental Analysis of Behavior, 37, 149-155. 341.
Morse, W. H., & Kelleher, R. T. (1970). Schedules as Rodgers, T. A., & Iwata, B. A. (1987, September). Analysis
fundamental determinants of behavior. In W. N. Schoen- of error correction procedures during behavioral acqui-
feld (Ed.), The theory of reinforcement schedules (pp. sition. In J. S. Bailey (Chair), Training research in
139-185). New York: Appleton. mental retardation. Symposium presented at the Florida
Morse, W. H., & Kelleher, R. T. (1977). Determinants Association for Behavior Analysis Convention, Sarasota.
of reinforcement and punishment. In W. K. Honig & Ruddle, H. V., Bradshaw, C. M., Szabadi, E., & Foster, T.
J. E. R. Staddon (Eds.), Handbook of operant behavior M. (1982). Performance of humans in concurrent
(pp. 174-200). Englewood Cliffs, NJ: Prentice-Hall. avoidance/positive-reinforcement schedules. Journal of
MyerJ. S. (1971). Someeffectsofnoncontingentaversive the Experimental Analysis of Behavior, 38, 51-61.
stimulation. In R. F. Brush (Ed.), Aversive conditioning Sandier, J., & Davidson, R. S. (1973). Psychopathology:
and learning (pp. 469-536). New York: Academic Press. Learning theory, research and applications. New York:
Nutter, D., & Reid, D. H. (1978). Teaching retarded Harper & Row.
women a dothing selection skill using community norms. Sandier, J., Davidson, R. S., Greene, W. E., & Holzschuh,
Journal of Applied Behavior Analysis, 11, 475-487. R. D. (1966). Effects of punishment intensity on in-
O'Brien, F., & Azrin, N. H. (1972). Developing proper strumental avoidance behavior. Journal of Comparative
mealtime behaviors of the institutionalized retarded. and Physiological Psychology, 61, 212-216.
Journal of Applied Behavior Analysis, 5, 389-399. Sandier, J., Davidson, R. S., & Malagodi, E. F. (1966).
O'Brien, S., Ross, L. V., & Christophersen, E. R. (1986). Durable maintenance of behavior during concurrent
Primary encopresis: Evaluation and treatment. Journal avoidance and punished-extinction conditions. Psycho-
of Applied Behavior Analysis, 19, 137-145. nomic Science, 6, 105-106.
Ollendick, T. H., Matson, J. L., Esveldt-Dawson, K., & Schiff, R., Smith, N., & Prochaska, J. (1972). Extinction
Shapiro, E. S. (1980). Increasing spelling achievement: of avoidance in rats as a function of duration and number
An analysis of treatment procedures utilizing an alter- of blocked trials. Journal of Comparative and Physio-
nating treatments design. Journal of Applied Behavior logical Psychology, 81, 356-359.
Analysis, 13, 645-654. Schoenfeld, W. N. (1969). "Avoidance" in behavioral
Osborne, J. G. (1969). Free-time as a reinforcer in the theory. Journal of the Experimental Analysis of Be-
management of classroom behavior. Journal of Applied havior, 12, 669-674.
Behavior Analysis, 2, 113-118. Schreibman, L. (1975). Effects of within-stimulus and ex-
Page, T. J., Iwata, B. A., & Neef, N. A. (1976). Teaching tra-stimulus prompting on discrimination learning in au-
pedestrian skills to retarded persons: Generalization from tistic children. Journal of Applied Behavior Analysis,
the dassroom to the natural environment. Journal of 8, 91-112.
Applied Behavior Analysis, 9, 433-444. Shnidman, S. R. (1968). Extinction of Sidman avoidance
Pierce, W. D., & Epling, W. F. (1980). What happened behavior. Journal of the Experimental Analysis of Be-
to analysis in applied behavior analysis? The Behavior havior, 11, 153-156.
Analyst, 3, 1-10. Sidman, M. (1966). Avoidance behavior. In W. K. Honig
(Ed.), Operant behavior: Areas of research and appli- ing on generalized vending machine use by moderately
cation (pp. 448-498). New York: Appleton. and severely handicapped students. Journal of Applied
Sidman, M., Herrnstein, R. J., & Conrad, D. G. (1957). Behavior Analysis, 17, 273-278.
Maintenance ofavoidance behavior by unavoidable shocks. Thompson, D. M. (1964). Escape from SD associated with
Journal of Comparative and Physiological Psychology, fixed-ratio reinforcement. Journal of the Experimental
50, 553-557. Analysis of Behavior, 7, 1-8.
Smith, R., & Keller, F. (1970). Free-operant avoidance in Weeks, M., & Gaylord-Ross, R. (1981). Task difficulty
the pigeon using a treadle response. Journal of the Ex- and aberrant behavior in severely handicapped students.
perimental Analysis of Behavior, 13, 211-214. Journal of Applied Behavior Analysis, 14, 449-463.
Solnick, J. V., Rincover, A., & Peterson, C. R. (1977).
Some determinants of the reinforcing and punishing Received August 25, 1987
properties of timeout. Journal of Applied Behavior Initial editorial decision August 30, 1987
Analysis, 10, 410-424. Revision received September 14, 1987
Sprague, J. R., & Horner, R. H. (1984). The effects of Final acceptance September 15, 1987
single instance, multiple instance, and general case train- Action Editor, Jon S. Bailey