Psychology Learning Resources
Joseph Wolpe and Systematic Desensitization

Systematic Desensitization: Concept Definition

Systematic desensitization is a therapeutic procedure designed to help clients overcome anxiety that is disproportionate to the actual situation. The procedure contains four core elements:

  1. Subjective Units of Discomfort (SUDs): The client learns how to report on a subjective scale how he or she feels when confronted with certain situations.
  2. Competing Response: The client is taught a response to counteract anxiety [usually relaxation].
  3. Hierarchy Construction: Feared situations are ordered with respect to the amount of anxiety they elicit.
  4. Desensitization Proper: The client repeatedly visualizes the feared situations in order of the hierarchy, while he or she performs the competing response.

(cf., Paul, 1969a; Spiegler & Guevremont, 1998; Wolpe, 1982; Wolpe, 1990)

1. Subjective Units of Discomfort (SUDs)

Systematic desensitization requires that the client make known to the therapist his or her discomfort level to various situations related to the problem. One possibility is for the client to make statements such as "a lot" or "very little." Wolpe (1990) questions the value of such verbal descriptions; instead, he advocates the use of a subjective anxiety scale. The scale ranges from zero to 100. The client is told to assign the number zero to an imagined sense of absolute calm and the number 100 to his or her worst anxiety. With these two reference points, the client rates his or her current state. Over time clients become increasingly confident in their subjective assessments. The unit is the SUD (subjective unit of discomfort). In all the tables thus far presented in this tutorial, the numbers listed under the "Anxiety" column are SUD values.

SUD values are not necessarily consistent between clients: a value of 50 for one client may be indicative of more discomfort than it is for another client. Thus, it is not possible to compare comfort levels between clients based on their SUD reports (Spiegler & Guevremont, 1998). What is important is that there is consistency within the reports of a single client: a situation assigned a value of 50 today should produce the same amount of discomfort than a situation assigned a value of 50 yesterday. If so, then SUD values can be more confidently used to assess the efficacy of treatment over time for that client.

Wolpe (1990) presents evidence that scores on this subjective scale correlate well with psychophysiological parameters of anxiety, such as peripheral vasoconstriction and heart rate.

2. Competing Response

Deep muscle relaxation is typically taught to clients in order to compete with negative emotional reactions engendered by the feared situation. Wolpe's (1990) relaxation procedure, which he adapted in abbreviated form from Jacobson (1938), consumes about a third of a session, takes about six lessons to complete, and requires that clients practice at home twice a day for 10 to 15 minutes. During therapy sessions the client is taught to alternately tense and relax different sets of muscles (arms; face; jaw; eyes; neck and shoulders; back, abdomen and thorax; and legs), and in the process he or she learns to differentiate between feelings of tension and relaxation. Wolpe (1990) advises that while there is no sacred sequence, the one adopted should remain orderly. By the end of training, the client can be brought to a state of deep relaxation using only a brief set of instructions.

Other ways of teaching relaxation include: autogenic training (repeated suggestions of heaviness and warmth), transcendental meditation, yoga, and electromyographic biofeedback (Wolpe, 1990).

The competing response need not be relaxation. Alternative competing responses have included sexual arousal, assertive behaviors, eating, pleasant thoughts, humor, and physical activity (Spiegler & Guevremont, 1998; Wolpe, 1990). Thus, systematic desensitization is still possible for those persons who, despite their and their therapist's best efforts, cannot achieve deep muscle relaxation. Suppose this were true for Nancy. If so, her therapist may choose to try humor as the competing response. He might present scenes to her with details that elicit laughter. For example, while Nancy is picturing herself giving a presentation, her therapist might describe how a particularly serious student in her class is making a funny face at the teacher (cf., Ventis, 1973). Note that Nancy would not have to learn to laugh, which is an advantage of using humor rather than relaxation as the competing response.

While some authors now question the role of relaxation as a critical component in the desensitization process, Wolpe (1990) points out that it was never his view that this particular competing response was indispensable to the success of the procedure.

3. Hierarchy Construction

Around the same time that relaxation training begins, the therapist and client engage in discourse in order to identify situations that make the client feel anxious. Questionnaires may also be employed. For example, Joseph Wolpe uses a Fear Survey Schedule: objects and events that typically elicit unpleasant feelings are listed and the client is asked to rate his or her discomfort level to each one (Wolpe, 1990, Appendix C, Fear Survey Schedule).

Notable fear-producing situations are then classified according to themes and assigned SUD values. Within each theme, the situations are ranked ordered according to these values on what is called an anxiety hierarchy. You have already seen examples of several anxiety hierarchies including that in the case of Nancy. The theme of her hierarchy could be called "fear of speaking in class." Deeper probing into Nancy's problem by her therapist may reveal that not only does she have a "fear of speaking in class," but she also has "fear of being devalued by others." Thus, Nancy and her therapist would likely also formulate another hierarchy based on this second theme. Items might include: "Being awarded a C+ for what she feels is an excellent essay" and "Making a relevant point that is ignored during a discussion within her study group."

Here are some key points to keep in mind about anxiety hierarchies:

  • The client need not have experienced each situation included in the hierarchy. For example, although Nancy had never given a full presentation in front of a class before, this does not exclude the situation from the hierarchy. The question posed to the client concerns whether he or she would expect to feel anxious if confronted by that situation (Wolpe, 1990).
  • The hierarchy is not carved in stone, and may be changed at any point in the procedure. Suppose the lowest anxiety-arousing situation in Nancy's hierarchy was assigned a SUD value of 30. Repeated presentations of this first scene may not diminish Nancy's anxiety to it. A solution would be to search for items dealing with the same theme that produce less anxiety and add them to the hierarchy. For example, a less fear-provoking situation than "Ask professor questions about presentation in his office" could be "Write down questions about presentation to be asked to professor in his office."
  • The hierarchy should contain evenly spaced items according to their SUD values, their distance being no more than 5 to 10 SUDs (Martin & Pear, 1998; Wolpe, 1990). If the distance is greater than this, the inclusion of intervening items should be considered. Wolpe (1990) explains the rationale: "If items are too similar, time will be wasted; if adjacent items differ too widely in anxiety-evoking potential, progress will be halted upon moving from the lesser to the greater" (p. 164). For example, suppose Item #5 on a hierarchy differs from Item #4 by only 2 SUD units. In this case, desensitization of Item #5 may have no effect on the desensitization of Item #6; in other words, Item #5 could be bypassed altogether with adversely affecting the outcome. On the other hand, suppose Item #4 differs from Item #6 by 30 SUD units; here, bypassing Item #5 may prove detrimental in that desensitization to Item #6 would have occurred much faster had Item #5 been desensitized first.
  • The hierarchy is based on relative SUD value assigned to each situation by the client and not on common sense. For example, you may have noticed from Nancy's "fear of public speaking" hierarchy that she reports greater anxiety thinking about giving a presentation to five of her friends in the classroom than she does in the presence of 10 of her friends. This may not make sense "logically" (most of us might believe the larger the group the greater the stress); nevertheless, Nancy's feelings are the determining factor and the "10 friends" situation is ranked below the "5 friends" situation on the hierarchy.
  • Miltenberger (1997) recommends between 10 to 20 situations per hierarchy, while Martin & Pear (1998) recommend between 10 to 30.

4. Desensitization Proper

After the client has learned deep muscle relaxation (or another competing responses) and has constructed a hierarchy of fears centered on a particular theme, the desensitization procedure begins. The session begins by inducing a deep state of relaxation; then, the therapist asks the client to imagine certain scenes from the hierarchy while maintaining this relaxed state.

Wolpe (1990, pp. 171-172) provides a transcript that illustrates the first session for Mrs. C, a 24-year old art student with examination anxiety.

THERAPIST: I am now going to ask you to imagine a number of scenes. You will imagine them clearly and they will interfere little, if at all, with your state of relaxation. If, however, at any time you feel disturbed or worried and want to draw my attention, you can tell me so. As soon as a scene is clear in your mind, indicate it by raising your little finger about one inch. First I want you to imagine that you are standing at a familiar street corner on a pleasant morning watching the traffic go by. You see cars, motorcycles, trucks, bicycles, people, and traffic lights; and you hear the sounds associated with all these things.

(After a few seconds the patient raises her left finger. The therapist pauses for five seconds).

THERAPIST: Stop imagining that scene. By how much did it raise your anxiety level while you imagined it?

Mrs. C: Not at all.

THERAPIST: Now give your attention once again to relaxing.

Note that this scene had nothing to do with taking examinations. In fact, it was not even part of Mrs. C's anxiety hierarchy. Why was it presented? Wolpe (1990) calls this a "control" scene. It is purposefully neutral and functions to gauge the client's abilities to visualize and relinquish control.

After a pausing 20-30 seconds with renewed relaxation instructions, the scene lowest on the anxiety hierarchy is presented.

THERAPIST: Now imagine that you are home studying in the evening. It is the 20th May, exactly a month before your examination.

(After about 15 seconds Mrs. C. raises her finger. Again she is left with the scene for 5 seconds.)

THERAPIST: Stop that scene. By how much did it raise your anxiety?

Mrs. C: About 15 units.

THERAPIST: Now imagine that same scene again a month before your examination.

The therapist presents the same scene a few more times until Mrs. C's reported anxiety level is reduced to zero. Then, the scene one higher on the anxiety hierarchy is presented in the same manner, until Mrs. C's reported anxiety level to it is also zero. In this way, the therapist proceeds up the hierarchy, eventually to the most fear-inducing scene. In other words, the therapist presents a new scene only after the client's self-report to the current scene indicates no discomfort (i.e., zero SUD units). If at any point during the visualization of a scene the client becomes extremely anxious, he or she is ordered to stop visualizing and to focus on relaxing.

Therapy sessions should end on a positive note, either with complete desensitization for a particular scene or with very good progress on it (Sarafino, 1996). Each new session starts with the last scene from the previous session for which anxiety was not reduced to zero; if anxiety to the last scene was reduced to zero, then the therapist presents the scene next highest on the hierarchy (Wolpe, 1990) (Others have recommended that each new session begin with the last scene fully desensitized in the previous session: Martin & Pear, 1998; Sarafino, 1996). Throughout the process, a record is kept of all of the client's responses, which helps the therapist monitor the client's progress and ultimately provides an objective measure of the outcome

What is the end result? Martin & Pear (1998) note:

"When the client finishes the last scene is the hierarchy, he/she can generally encounter the actual feared objects without undue distress. No doubt the positive reinforcement that the client receives [perhaps from significant others] then helps maintain continued interactions with the stimuli that previously elicited intense, debilitating fear." (p. 332)

Here are some other general guidelines and quantitative considerations concerning the desensitization procedure.

  • For clients for whom multiple hierarchies were constructed, the therapist may work with up to four of them in any given session. (Wolpe, 1990).
  • Usually, presenting a scene three or four times will suffice to bring the client's anxiety level to zero, but sometimes 10 or more presentations may be required (Wolpe, 1990). When repeated presentations of a scene do not produce the desired effect, remedial measures are taken, which might include: shortening how long the client visualizes that scene, imagining the scene one lower on the hierarchy for a longer than usual time (e.g., 20 sec), and constructing and working through new items on the hierarchy with SUD values between the scene one lower on the hierarchy and the problematic scene (Martin & Pear, 1998).
  • The total number of scenes presented within a session is limited by the scheduled duration of the session as well as the client's endurance. In beginning sessions there may be time for only eight to ten presentations, but later on - with "preliminaries" taking less and less time - 30 to 50 presentations are possible (Wolpe, 1990).
  • A desensitization session typically lasts about 15 to 30 minutes (Wolpe, 1990).
  • The client is usually asked to imagine a scene for about five to seven seconds (Wolpe, 1990).
  • The interval between scenes is usually between 10 and 30 seconds (Wolpe, 1990).
  • The number of sessions to complete the process varies widely depending upon the specifics of the problem. Success has been reported in a few as one session (e.g., Ventis, 1973); or, it can take as long as 100 sessions (Wolpe, 1990).
  • The spacing of sessions does not appear to affect the outcome and very little change appears to occur between sessions (Wolpe, 1990). Sessions may be scheduled several times a day or be spaced apart by several weeks. (However, at least two authors suggest that desensitization sessions occur no fewer than once per week: Martin & Pear, 1998; Sarafino, 1996).
  • "It is important to assess the client's fear in the actual fear-producing situation, so as to be sure that the results of systematic desensitization have generalized successfully. If the results of systematic desensitization do not fully generalize, in vivo desensitization [to be discussed] may be used in addition to systematic desensitization to enhance its effectiveness and ensure generalization" (Miltenberger, 1997, p. 497).