Filetype PDF | Posted on 16 Sep 2022 | 3 years ago
Authentication
digital resources
159x Filetype PDF File size 0.88 MB Source: www.paneecioccolata.com
Journal oj Abnormal Psychology
1968, Vol. 73, No. 2, 91-99
SYSTEMATIC DESENSITIZATION AS A COUNTER-
CONDITIONING PROCESS l
GERALD C. DAVISON
State University of New York at Stony Brook
Systematic desensitization, demonstrated in both clinical and experimental
studies to reduce avoidance behavior, entails the contiguous pairing of
aversive imaginal stimuli with anxiety-competing relaxation. If, as is widely
assumed, the efficacy of the procedure derives from a genuine countercondi-
tioning process, a disruption of the pairing between graded aversive stimuli and
relaxation should render the technique ineffective in modifying avoidance be-
havior. This hypothesis was strongly confirmed: significant reduction in avoid-
ance behavior was observed only in densensitization Ss, with none occurring
either in yoked Ss for whom relaxation was paired with irrelevant stimuli or
in yoked Ss who were gradually exposed to the imaginal aversive stimuli with-
out relaxation. Other theoretical issues were raised, especially the problem of
transfer from imaginal to actual stimulus situations.
Recent years have witnessed increasing ap- and with more objective assessment of thera-
plication of the systematic desensitization peutic outcomes (e.g., Lang and Lazovik,
procedure, as developed by Wolpe (1958), to 1963; Lang, Lazovik, & Reynolds, 1965;
the modification of a wide range of neurotic Lazarus, 1961; Paul, 1966; Paul & Shannon,
disorders. In this therapeutic method the 1966). Although results from these experi-
client is deeply relaxed and then instructed ments have confirmed the effectiveness of
to imagine scenes from a hierarchy of anxiety- systematic desensitization, they do not pro-
provoking stimuli. Initially he is asked to vide direct information on the relative con-
imagine the weakest item in the list and, tributions to the observed outcomes of the
if relaxation is unimpaired, is gradually pre- different variables in the treatment procedure
sented incremental degrees of aversive stimuli (e.g., relaxation, graded exposure to aversive
until eventually he is completely desensitized stimuli, temporal contiguity of stimulus
to the most upsetting scene in the anxiety events). Moreover, the learning process gov-
hierarchy. erning the behavioral changes has not been
In numerous publications, both Wolpe (e.g., adequately elucidated. There is some sug-
19S2, 1958) and other clinical workers (e.g., gestive evidence from Lang et al. (1965) that
Geer, 1964; Lang, 1965; Lazarus, 1963; extensive contact with an E, along with re-
Lazarus & Rachman, 1957; Rachman, 1959) laxation training, does not effect behavior
have claimed a high degree of success in change. However, one can raise questions
eliminating diverse forms of anxiety dis- about the suitability of their control for re-
orders by means of this therapeutic technique. laxation, inasmuch as Ss in this condition
These clinical claims of efficacy find some began imagining snake-aversive items, but
support in recent laboratory investigations were then led away from this theme by means
conducted under more controlled conditions of subtle manipulation of content by E. It
1 This paper is based on the author's doctoral dis- is possible that this imaginal snake avoidance
sertation written at Stanford University under Albert may have counteracted the nonspecific effects
Bandura, whose invaluable advice and direction at built into the control.
every stage of the research and composition he is Wolpe's (1958) theoretical formulation of
pleased to acknowledge. For their aid and encourage- the desensitization process as "reciprocal in-
ment, sincere thanks are also rendered to Arnold hibition" is based on Hull's (1943) drive-
A. Lazarus and Gordon L. Paul. The author is reduction theory of classical conditioning, a
especially grateful to O. B. Neresen, who made fatigue theory of extinction ("conditioned in-
available both the physical facilities and human hibition"), and Sherrington's (1906) concept
resources at Foothill Junior College, Los Altos,
California. of reciprocal inhibition, whereby the evoca-
91
92 GERALD C. DAVISON
tion of one reflex suppresses the evocation of reactions to these stimuli (cf. Bandura, in
other reflexes. The conditions which Wolpe press).
(19S8) specified for the occurrence of recip- PROBLEM
rocal inhibition were succinctly stated in his
basic principle: In view of the fact that the behavioral out-
If a response antagonistic to anxiety can be made comes associated with systematic desensitiza-
to occur in the presence of anxiety-evoking stimuli tion are assumed to result from counter-
so that it is accompanied by a complete or partial conditioning, evidence that such a process
suppression of the anxiety responses, the bond be- does in fact occur is particularly essential
tween these stimuli and the anxiety responses will (cf. Breger & McGaugh, 1965). To the extent
be weakened [p. 71]. that desensitization involves countercondi-
This statement appears indistinguishable from tioning, the contiguous association of graded
Guthrie's (1952) view of countercondition- anxiety-provoking stimuli and incompatible
ing, according to which notion the elimination relaxation responses would constitute a neces-
of a response can be achieved by eliciting a sary condition for fear reduction. It is pos-
strong incompatible response in the presence sible, however, that the favorable outcomes
of cues that ordinarily elicit the undesirable produced by this method are primarily attrib-
behavior: "Here . . . the stimulus is pres- utable to relaxation alone, to the gradual
ent, but other responses are present shutting exposure to aversive stimuli, or to nonspecific
out the former response, and the stimulus relationship factors. The present experiment
becomes a conditioner of these and an inhibi- was therefore designed to test directly the
tor of its former response [p. 62]." Wolpe, in hypothesis that systematic desensitization in-
fact, used the terms "reciprocal inhibition" volves a genuine counterconditioning process.
and "counterconditioning" interchangeably, The 5s were individually matched in terms
but clearly indicated a preference for the of strength of their snake-avoidance behavior
former in view of his inferences about the and assigned to one of four conditions. For
neurological process accounting for the ob- one group of Ss (desensitization), a graded
served changes in behavior. However, aside series of aversive stimuli was contiguously
from the fact that he has as yet provided no paired in imagination with deep muscle relax-
independent evidence for the existence of re- ation, as in the standard clinical technique.
ciprocal inhibition at the complex behavioral The 5s in a second group participated in a
level that he is dealing with, one must be "pseudodesensitization" treatment that was
wary of basing a neurological hypothesis, identical to the first procedure except that
albeit an ingenious one, upon a behavioral the content of the imaginal stimuli paired
system which, itself, has been shown to have with relaxation was essentially neutral and
serious shortcomings (Gleitman, Nachmias, & completely irrelevant to snakes. This group
Neisser, 19 54; Kimble, 1961; Lawrence & provided a control for the effects of relation-
Festinger, 1962; Mowrer, 1960; Solomon & ship factors, expectations of beneficial out-
Brush, 1956). comes, and relaxation per se. A third group
At the present time, it appears both un- (exposure) was presented the same series
necessary and premature to "explain" be- of graded aversive items, but in the absence
havioral phenomena in terms of an underlying of deep relaxation. This condition served as
neural process whose existence is inferrable a control for the effects of mere repeated
solely from the very psychological data which exposure to the aversive stimuli. A fourth
it is invoked to explain. It appears to this group (no treatment) participated only in the
writer more fruitful to stay closer to the pre- and posttreatment assessments of snake
empirical data and to conceptualize the avoidance.
process of systematic desensitization in terms In order to ensure comparability of stimu-
of counterconditioning, according to which lus events, 5s in the pseudodesensitization
the neutralization of aversive stimuli results and exposure groups were yoked to their
from the evocation of incompatible responses matched partners in the desensitization group,
which are strong enough to supersede anxiety whose progress determined the number of
SYSTEMATIC DESENSITIZATION AND COUNTERCONDITIONING 93
treatment sessions, the duration of each ses- Lang and Lazovik's study touched and held the
sion, the number of stimulus exposures per snake before requesting an S to do so. Evidence
session, and the duration of each exposure. that avoidance behavior can be reduced through
Within 3 days following the completion of observation of modeled approach responses (Ban-
dura, Grusec, & Menlove, 1967) suggests that the be-
treatment, all Ss were tested for snake avoid- havioral changes obtained by Lang and Lazovik may
ance as well as for the amount of anxiety reflect the effects of both vicarious extinction and
accompanying each approach response. counterconditioning via systematic desensitization.
On the assumption that the temporal con- Any S who, on the pretreatment assessments, suc-
junction of relaxation and anxiety-provoking ceeded in touching the snake barehanded was ex-
cluded from the study. Eligible 5s were matched
stimuli is essential for change, it was pre- individually on the basis of their approach behavior
dicted that only 5s in the desensitization and then assigned randomly to the different treat-
condition would display significant decrements ment conditions so as to constitute "clusters" of
in avoidance behavior, and would also be equally avoidant Ss across groups. Initially it had
been planned to include an equal number of matched
superior in this respect to Ss in the three Ss in the no treatment control group. However,
control groups. since preliminary findings, as well as data reported
by Lang and Lazovik (1963), revealed virtually no
METHOD changes in nontreated controls, it was decided to
Subjects enlarge the size of the three treatment conditions.
Therefore, eight 5s were assigned to each of the
The 5s were 28 female volunteers drawn frjm three treatment groups, while the nontreated control
introductory psychology courses at a junior college. group contained four cases. The experimental design
Students who reported themselves very much afraid is summarized in Table 1.
of nonpoisonous snakes were asked to assist in a
study investigating procedures for eliminating com- Treatment Procedures
mon fears. In order to minimize suggestive effects, The treatment sessions were conducted in a room
the project was presented as an experiment, rather other than the one in which the avoidance behavior
than as a clinical study, and no claims were made was measured. The Ss in conditions employing relax-
for the efficacy of the procedure to be employed. ation training reclined in a lounger, whereas for Ss
To reduce further the development of strong expec- in the exposure group the chair was set in an
tation of beneficial outcomes, which might in itself upright position to minimize the development of
produce some positive change, E was introduced as relaxed states.
a graduate student rather than as an experienced Relaxation paired with graded aversive stimuli
psychotherapist. To some extent, the results from (systematic desensitization). During the first session,
all the experiments cited above might have been these Ss received training in deep muscular relaxa-
confounded by these variables. tion by means of a 30-min. tape recording consist-
Pre- and Posttreatment Assessments of ing of instructions to tense and to relax alternately
Avoidance Behavior the various muscle groups of the body, interspersed
with suggestions of heaviness, calm, and relaxation.
These assessments were conducted by an E (Ei) This procedure, used earlier by the author (Davison,
who did not participate in the treatment phases of 196Sb), is based on Lazarus' (1963) accelerated
the study and had no knowledge of the conditions training in Jacobsonian relaxation and is very similar
to which Ss were assigned. The avoidance test was to the technique used by Paul (1966).
similar to that employed by Lang and Lazovik In the second session Ss ranked 26 cards each
(1963) except for several important changes that describing snake scenes in order of increasing aver-
were introduced in order to provide a more stringent siveness, for example, "Picking up and handling a
and sensitive test of the efficacy of the various toy snake," "Standing in front of the cage, looking
treatment procedures. First, whereas Lang and down at the snake through the wire cover, and
Lazovik used essentially a 3-item test, the present it is moving around a little," "Barehanded, picking
behavioral test consisted of 13 items requiring pro- the snake up, and it is moving around." The
gressively more intimate interaction with the snake desensitization procedure, modeled after Lazarus
(e.g., placing a gloved hand against the glass near (1963), Paul (1966), and Wolpe (1961), was ad-
the snake, reaching into the cage and touching the ministered in a standardized fashion, with a criterion
snake once, culminating with holding the snake bare- of IS sec. without signaling anxiety on each item.
handed for 30 sec.). Second, rather than obtaining (For specifics of the procedure, see Davison, 196Sa.)
a single overall estimate of felt anxiety following A maximum of nine sessions, each lasting about
the entire approach test, the examiner in the present 45 min., was allowed for completing the anxiety
study asked 5 to rate herself on a 10-point scale hierarchy.
following the successful performance of each task. Relaxation paired with snake-irrelevant stimuli
Third, the examiner stood at all times not closer (pseudodesensitization). The Ss assigned to this
than 2 ft. from the cage, whereas the tester in group received the same type and amount of
94 GERALD C. DAVISON
TABLE 1
SUMMARY OF EXPERIMENTAL DESIGN
Group Pretreatment assessment Treatment procedure Posttreatment assessment
(£2) (Bi)
Desensitization" Avoidance test with Relaxation paired with graded Avoidance test with
anxiety self-reports aversive stimuli anxiety self-reports
Pseudodesensitization" Same Relaxation paired with Same
snake-irrelevant stimuli
Exposure" Same Exposure to graded aversive Same
11 stimuli without relaxation
No treatment Same No treatment Same
Jtf = 8.
relaxation training as 5s in the above-mentioned imaginal exposure. Thus, Ss undergoing pseudo-
group. Similarly, in the second session they also desensitization received the same number and dura-
ranked 26 stimulus items, except that the depicted tion of pairings during each session as their
scenes were entirely unrelated to snakes. Because desensitization mates, with the important exception
of the widespread belief that exploration of child- that snake-irrelevant stimuli were contiguously
hood experiences may be important in alleviating associated with relaxation.
objectively unrealistic fears, it was decided to em- Exposure to graded aversive stimuli without
ploy descriptions of common childhood events, which relaxation (exposure). The 5s in this group were
Ss were asked to rank chronologically. Some of the administered the same series of snake-aversive stim-
items were essentially neutral in content ("You are uli in the same order and for the same durations as
about age six, and your family is discussing where determined by their respective partners in the
to go for a ride on Sunday afternoon, at the dinner desensitization group to whom they were yoked.
table.")) while the others had mild affective proper- However, exposure 5s received no relaxation train-
ties ("You are about five years old, and you are ing (hence, had one session less with E), nor did
sitting on the floor looking sadly at a toy that you they engage in anxiety-competing relaxation while
have just broken."). The use of generic content thus visualizing the aversive situations. Because of the
made it possible to use snake-irrelevant stimuli yoking requirements, on those occasions when 5s
without reducing the credibility of the treatment signaled anxiety, they were instructed to maintain
procedure. the images until E asked them to discontinue. Co-
As in the desensitization condition, 5s were deeply operation in this obviously unpleasant task was
relaxed and asked to imagine vividly each scene obtained through friendly but cogent reminders that
presented by the E until told to discontinue the such visualization was important for the experimental
visualization. Each 5 in this condition, it will be design.
recalled, was yoked to her matched partner in the No treatment group. The Ss assigned to this
desensitization group, whose progress defined the group merely participated in the assessments of
number of treatment sessions, the length of each avoidance behavior at the same time as their
session, as well as the number and duration of each matched partners in the desensitization condition.
TABLE 2 RESULTS
CHANGES IN SNAKE-APPROACH BEHAVIOR DISPLAYED Table 2 presents the change scores in ap-
BY SUBJECTS IN EACH OP THE TREATMENT proach behavior for each S in each of the
CONDITIONS
eight matched clusters.
Condition
Matched Between-Group Differences
cluster Desensi- Pseudo- No treat-
tization desensi- Exposure ment Because of the unequal number of Ss in the
tization no treatment group, these data were not in-
1 3 2 2 0 cluded in the overall statistical analysis.
2 3 -1 0 . — . Two-way analysis of variance of the change
3 6 0 -1 -1
4 5 1 -5 0 scores obtained by the three matched treat-
S 0 1 2 — ment groups yielded a highly significant
6 6 8 1 0 treatment effect (F = 6.84; p < .01).
7 12 0 0 —
8 7 1 1 — Further, one-tailed comparisons of pairs of
M 5.25 1.50 0.0 -0.25 treatment conditions by t tests for correlated
means revealed that Ss who had undergone
no reviews yet
Please Login to review.
