Eliminating Fears: An Intervention that Permanently Eliminates the Fear of Public Speaking

To get a pdf file of the study published in the journal Clinical Psychology and Psychotherapy Click Here.

NOTE: The html version of this report found below is missing graphics and tables found in the original.


Victoria Cunningham, Ph.D.Lee Sechrest, Ph.D.Department of Psychology, University of Arizona (VC, LS);

Morty LefkoeLefkoe Institute, Fairfax, CA (ML).


This study examined the ability of a novel psychological treatment to eliminate anxiety using fear of public speaking as a test case. The treatment was designed specifically to eliminate beliefs and de-condition stimuli that are responsible for dysfunctional behaviors and emotions. A random half of the subjects received treatment immediately while wait-listed control subjects were treated three to four weeks later. After having engaged in an actual public speaking experience, subjects’ self-reported ratings showed significant reductions in fearfulness, physical sensations and cognitive difficulties often associated with speaking in public. The evidence strongly suggests that fear of public speaking was virtually eliminated and we propose that it holds promise as an intervention that might be effective in treating many other disorders.


Several surveys indicate that Americans rank speaking in public as their number one fear (Bruskin Associates, 1973; Motley, 1988; Richmond & McCroskey, 1995). This fear can be socially debilitating, and is often cited as a primary reason why someone is unable to advance in his or her career. Fear of public speaking may be related to a more general social anxiety, but it is not coterminous with it as many people appear to have quite specific fear of public speaking in the context of otherwise normal social relationships. Different therapeutic approaches have been developed to help people overcome or deal with such fears as of public speaking. One such approach is The Lefkoe Method (TLM). The Lefkoe Method, developed by the second author, aims to eliminate, quickly, long-held beliefs and “de-condition” the stimuli that produce fear and other negative emotions, e.g., the fear of speaking in public. Lefkoe has discovered that the fear of public speaking is typically caused by a) specific beliefs, such as “Mistakes and failure are bad” and “If I make a mistake, people will laugh at me” and b) conditioning, such as automatically experiencing fear whenever one is, or perceives oneself to be, in a position to be criticized or judged. Two processes in TLM, the Lefkoe Belief Process and the Lefkoe Stimulus Process, are used to address fear of public speaking.

Many, if not most, psychologists contend that long-held beliefs can be totally eliminated, if at all, only after extensive time, effort, and specific retraining. TLM challenges that assumption and contends that even beliefs formed early in childhood can be permanently eliminated in a matter of minutes. The basis for this claim is clients who state that a belief that was experienced as true is no longer experienced as true and that the behavior and emotions that result from the belief are permanently eliminated. Moreover, TLM contends that emotions that result from conditioned stimuli, for example, fear that is always experienced when one makes a mistake or is rejected, can be quickly and permanently stopped by de-conditioning the stimuli. That also can be accomplished in a matter of minutes. Those are bold claims, but they can be empirically tested. Because of the extensive clinical experience with TLM, we have elected to test it in clinical settings rather than a laboratory. This article reports on the results of application of TLM to reduction of fears associated with public speaking.

TLM, despite its fairly lengthy history, is not widely known, partly because no systematic attempt has ever been made to publicize it. TLM is, however, publicly available at no charge on a website (www.decisionmaker.com). If it can be shown to be effective in terms of scientific evidence, it has the potential to be highly useful since it is, as will be seen, a brief intervention that may be delivered even without face-to-face contact for at least some conditions.

Although the Lefkoe Belief Process (LBP) is similar in some ways to cognitive therapy approaches (CT), there are many unique aspects that distinguish it from other such approaches. First, some versions of CT attempt to change beliefs by challenging the validity of the evidence that the client uses to support them. With LBP no attempt is made to get clients to see that a current belief is wrong or not true, to see it as illogical, or to accept that it does not make sense, or to reject it as self-defeating. The LBP actually validates people for forming the belief earlier in life by assisting them to realize that most people probably would have made a similar interpretation under similar circumstances. It ensures that people realize that their belief actually is one valid meaning of their earlier circumstances.

The “evidence” that people offer for a belief usually is not the actual reason they believe it. The evidence offered usually consists of recent observations that appear to substantiate the belief. The real source of one’s beliefs, the LBP assumes, is interpretations of circumstances earlier in life. Fundamental beliefs about one’s self and life are usually formed in childhood. After a belief has been formed, however, one acts consistently with it, thereby producing “current evidence” for the already-existing belief. In other words, life becomes a self-fulfilling prophecy. Because the evidence one presents to validate one’s beliefs usually is a consequence of the belief, not its source, challenging the validity of that evidence is not the most effective way to eliminate them.

A third element that distinguishes LBP from some versions of CT has to do with getting the client to agree to act consistently with an alternative belief to test its possible validity. Because the current belief is totally eliminated by doing the LBP, one has no need to try to act differently when one goes back “into life”; one’s behavior changes naturally and effortlessly once the belief is gone.

Still another distinction between the LBP and many cognitive approaches is that the latter frequently are a tool for the client, whereas the former is a tool for the facilitator. Cognitive approaches assist clients to think more rationally in order to act more rationally in the face of strong emotions like fear, anger, depression, hostility, etc. The LBP is used by the facilitator to assist clients in eliminating the beliefs that produce such emotions. When these emotions stop after the beliefs that give rise to them are eliminated, there is no longer a need for a tool to deal with them more effectively. Finally, the Lefkoe Stimulus Process facilitates de-conditioning the stimuli for negative emotions, which has nothing to do with beliefs. In order to get rid of the fear of public speaking, one has to extinguish the conditioned stimuli that got associated with fear, such as facing criticism, feelings that one is not meeting expectations, that one is being judged, or that one is being rejected. The point of this process is to assist the person to realize that initially the current stimulus never produced the emotion. The current stimulus got conditioned to produce the negative emotion because it just happened to be associated with the real original cause in some way.

Potential effectiveness of the Lefkoe Method

The Lefkoe Method has not previously been subjected to rigorous investigation, although there is reason to believe that it might well be effective in treating a wide range of problems. In 1994 The Lefkoe Institute, in collaboration with Sechrest, conducted a study involving 16 incarcerated youths and adults at two Connecticut institutions. The study indicated fairly strongly that using TLM, specifically, the Lefkoe Belief Process, to eliminate such beliefs as “I’m bad,” “There’s something wrong with me,” “I don’t matter,” and “What makes me okay is the power that comes from a gun,” improved the self-esteem and reduced the hostility and anti-social behavior of the subjects. In part because of the small sample, the study, although reflecting statistically significant effects, was never published; the effect was actually fairly large. The study did, however, provide impetus for Lefkoe to continue use and development of his unique intervention. He and his associates have by now treated over 2,000 people with a wide range of problems, and results as he has seen them have been consistently highly favorable. He has also trained a number of other clinicians in the use of his method, and they, too, have, in aggregate, treated, successfully, a very large number of persons. The experiences of these clinicians constitute a strong basis for more systematic testing of the effectiveness of the Lefkoe approach.

An increasing number of case studies and anecdotal reports provide evidence that TLM has been effective in resolving a wide variety of serious psychological issues, including anxiety, drug and alcohol addition, ADD, bulimia, phobias, the inability to leave abusive relationships, anger, hostility, and guilt. It also is successful with everyday issues such as worrying about what people think of you, workaholism, the feeling that nothing one does is ever good enough, procrastination, and the inability to express feelings. Whether the anecdotal reports of the effectiveness of TLM with the above-mentioned psychological issues can be replicated in controlled scientific studies remains to be seen. The significant results obtained in the 1994 study, coupled with the plentiful observational evidence support the proposition that TLM might well be both efficient and effective in treating a range of at least mild to moderately severe disorders, prompted us to conduct the present study. In searching about for a test bed for TLM, we hit upon the idea of trying it out with fear of public speaking. That problem is, apparently, not uncommon, it is often at least moderately severe, and many people who experience it are highly motivated to get rid of it. Moreover, Toastmasters clubs and similar groups provide a good entry to the recruitment of persons interested in treatment.


This study is the second in what we expect to be a number of studies designed to determine the reliability of the extensive anecdotal evidence.


Forty volunteers were recruited primarily through Toastmaster groups located near a large metropolitan Western city and were assigned randomly to either the immediate or wait-list comparison condition. Three persons dropped out of the immediate treatment group and one from the wait-list group before beginning treatment or after one session so that the final sample size was 36. To be eligible for the study, participants had to report at least a moderate fear of public speaking, defined by a self-rating of 5 or greater on a 10-point Likert scale ranging from 1 (not at all fearful) to 10 (extremely fearful). In addition, participants had to acknowledge that their fear related only to circumstances of speaking in public and not to other broader areas of their life, have access to a telephone, be willing to give informed consent to participate in the study, and be fluent in English. Women comprised 53% of the sample. The mean age was 42.6 (s.d. = 11.8). On average, participants had 9.6 years (s.d. = 8.9) of experience with speaking in public.

Experimental Design

Eligible participants were randomly assigned to TLM or a yoked wait-list control group (WL), i.e., subjects were paired at time of assignment. Figure 1 illustrates the study design and timing of questionnaire administration.

Group Timeline of ActivitiesTLM: Pretest Treatment 1st Public Speaking Posttest — — –WL: Pretest — — Posttest Treatment 1st Public Speaking Posttest

Figure 1. Study Design and Timeline of Activities.

A wait-list control design was chosen because the primary endpoint of this study was to determine the reduction in fear associated with speaking in public. This endpoint entailed waiting until after TLM subjects completed treatment and had an opportunity to speak in public. Immediately following their public speaking experience, TLM subjects completed posttest questionnaires and notified the facilitator within 1-2 days so their yoked WL subject could be instructed to complete posttest measures. WL subjects received treatment after the waiting period and completed a second set of the same posttest questionnaires immediately following their first public speaking experience. Both TLM and WL subjects attended their Toastmaster sessions during the study and both groups had the opportunity to speak there during the course of the study.


Subjects were recruited primarily through announcements made at public speaking clubs or via emails sent to club members. After the study coordinator received a signed consent form and completed baseline measures, subjects were randomly assigned to either the TLM or WL group. Subjects in the TLM group then scheduled a series of phone calls to receive as many individual treatment sessions as would be necessary to eliminate the fear. (The range was from two to five sessions, with a mean of 3.3) About half of the 12,0,500 clients who have been treated with TLM receive help over the telephone, and the reported results are always as effective as the in-person sessions (M. Lefkoe, personal communication, March 25, 2004). The facilitator for all subjects in the study was Morty Lefkoe, who has overalmost 20 years of experience in using TLM to assist clients to get rid of a wide variety of emotional and behavioral conditions, including the fear of speaking in public. The treatment consists of one-hour sessions and is delivered according to structured treatment protocols developed by Mr. Lefkoe.

Outcome Measures

Questionnaires were emailed to research subjects. All data were collected by email or fax. Because the major “problem” being reported by the subjects was the experience of anxiety, that construct was the focus of our attempts to determine the effectiveness of the treatment. Other aspects of the “problem” include uncomfortable and unpleasant physical sensation, which we also measured. We also included one measure from an established research tradition as a way of anchoring our findings to show that they are congruent with those of other investigators.

Self-rated Performance

Subjects rated their last public speaking experience with five single-item measures including how fearful, anxious, satisfied, confident, and relaxed they felt. Items were scored on a 10-point scale from ranging from1 (not at all) to 10 (extremely).

Subjective Units of Bothersome Sensations Scale

The SUBSS consists of twelve somatic and cognitive sensations commonly reported as intrusive while speaking in public. Items were rated on a 4-point scale ranging from 0 (not at all bothersome) to 3 (severely bothersome). (See Appendix). Subjects were instructed to refer to their last public speaking experience when completing the items. Ratings are summed to generate a total score with a potential range of from 0 to 36. Cronbach’s alpha for the pretest on this study sample was .67.

Confidence as a Speaker

Confidence as a speaker was measured using the Personal Report of Confidence as a Speaker (PRCS; Paul, 1966). The PRCS is a 30-item self-report measure that assesses affective and behavioral reactions to public speaking situations. The items are answered in true-false format; half are keyed “true” and half are keyed “false” to control for acquiescent responding.

Respondents were instructed to consider each item as it related to their “most recent public speaking experience.” Scores have a possible range from a low of 0 to a high of 30; the higher the score, the greater the degree of anxiety. Cronbach’s alpha was .80 for the pretest on the study sample.

Data Analysis

To assess differences for between- and within-group treatment effects, we used a series of one-way analysis of variance tests for each of the outcome measures. The primary outcome measure was the rating of fear associated with the subject’s last public speaking experience. Secondary outcome measures included ratings of how satisfied, relaxed, anxious, and confident the subject felt during his or her last public speaking experience, level of confidence as a speaker measured by the PRCS, and bothersome sensations measured by the SUBSS.

Three TLM and one WL subject terminated the study after the first session. Obviously, the number of cases was very small, but attrition from the study did not appear to be associated with any descriptive variables or pretest data. Reasons given for terminating were insufficient time and disappointment at not being given tips about managing anxiety.


Means and standard deviations for all measures at each assessment period are presented in Table 1. Because of the somewhat exploratory nature of this study, we report data separately for each measure, partly to determine the consistency of findings.

Table 1 Means and Standard Deviations for Pretest and Posttest Measures

TLM Group (n = 17) WL Group (n = 19) Pretest Posttest Pretest Posttest PosttestbMeasure M (sd) M (sd) M (sd) M (sd) M (sd)Fear 6.65 (1.32) 1.38 (0.50) 6.53 (1.71) 6.95 (1.61) 1.53 (0.51)Anxiety 7.24 (1.75) 1.69 (0.48) 6.89 (1.63) 7.32 (0.67) 1.89 (0.87)Satisfa 4.81 (2.31) 8.63 (1.09) 4.42 (1.98) 4.11 (2.23) 7.89 (2.16)Confide 4.18 (1.51) 8.81 (0.75) 4.26 (1.76) 4.37 (1.86) 8.16 (1.89)Relaxed 3.65 (1.80) 8.88 (0.81) 3.42 (1.54) 3.74 (1.63) 8.05 (2.27)PRCS 19.88 (4.59) 4.50 (3.10) 18.79 (5.09) 20.11 (4.16) 5.32 (4.78)SUBSS 17.18 (4.54) 2.65 (1.66) 16.68 (5.95) 15.16 (6.82) 2.11 (2.47)

Note: TLM = The Lefkoe Method group; WL = wait-list control group; Posttestb = WL scores after receiving treatment; PRCS = personal report of confidence as a speaker; SUBSS = subjective units of bothersome sensations scale.

The data reported in Table 1 are remarkably consistent across measures, including the PRCS, a measure well established in the literature. Particularly noteworthy is the fact that subjects in the WL group did not change at all on any measure until treatment, after which their scores were closely equivalent to those of persons in the initial treatment group. Thus, the change cannot be attributed to effects of retesting. First, there were no significant differences between groups on any measure at pretest. Second, scores in the TLM group at posttest were dramatically different from WL posttest scores. Third, after they received treatment, subjects in the WL group had scores that were not different from those at posttest for the TLM group. Figure 2 is a graphic display of “average” results for all outcome measures. The figure shows quite well the change occurring in each group when treatment takes place and the apparent magnitude of the treatment effect which seems quite large. Just to illustrate, across the 12 items in the SUBBS, at pretest subjects would have been reporting patterns of response something like:

Ratings of 1 on 6 items and ratings of 2 on six moreor Ratings of 2 on 6 items and ratings of 3 on two moreor Ratings of 2 on 8 items

After treatment, the patterns would have been more like’

Ratings of 1 on 3 itemsor Rating of 2 on one item and rating of 1 on another itemsor Rating of 3 on one item

Figure 2. Mean ratings before (pretest) and after (posttest) treatment.

Between-Group Effects at Posttest

Table 2 presents between-group effects. Results were large differences on all outcome variables when comparing TLM and WL posttest scores. By contrast, comparisons of TLM posttest scores with scores for the WL group after having received treatment (Posttestb ) were very small and associated with uniformly small and non-significant F-values.

Table 2

Between-Group Effects (F-values) for Posttest Measures

TLM vs. WLPosttest TLM vs. WLPosttestbMeasure F p F pFear 175.55 .001 0.77 .39Anxiety 786.62 .001 0.71 .40Satisfaction 54.45 .001 1.50 .23Confidence 79.89 .001 1.69 .56Relaxed 131.75 .001 1.89 .18PRCS 153.14 .001 0.34 .56SUBSS 54.23 .001 0.58 .45

Note: TLM = The Lefkoe Method group; WL = wait-list control group; Posttestb = WL scores after receiving treatment; PRCS = personal report of confidence as a speaker; SUBSS = subjective units of bothersome sensations scale.

On average, subjects in both groups received 3 sessions

[F(1,34) = .25, p > .62]. Only three subjects in the WL group required a 5th session.

Within-Group Effects from Pretest to Posttest

Analysis of within-group changes from pretest to the first posttest showed significant treatment effects for all outcome measures in the TLM group and no such changes in the WL group (see Table 3). Across all outcome measures, after WL subjects received treatment, their scores were negligibly different from those in the TLM group.

Table 3

Within-group Effects (F-values) for Pretest to Posttreatment in TLM and WL Groups

TLMPretest vs. Posttest WLPretest vs. Posttest WLPretest vs. PosttestbMeasure F* F F*Fear 224.51 0.61 34.19Anxiety 149.81 1.08 138.85Satisfaction 35.53 0.21 26.72Confidence 122.29 0.03 43.16Relaxed 113.31 0.38 54.11PRCS 125.49 0.76 70.61SUBSS 153.38 0.54 97.19

Note: * p < .001, otherwise, F-values are non-significant.

To assess the magnitude of the treatment effect, we calculated Cohen’s d (Mpost – Mpre / SDpre) (1988). Because the pretest and post-treatment values in the TLM and WL groups were not significantly different from one another, we pooled their data. All effect sizes estimates were substantial (fear d = 3.36, PRCS d = 2.97, SUBSS d = 2.76).

Subjects were asked to rate how helpful the sessions were for them in reducing or eliminating their fear of speaking in public on a 10-point on which1 was ‘not at all helpful’ and 10 was ‘extremely helpful’. Ninety-four percent of the sample rated the treatment as 7 or higher.


The large, positive changes on all outcome measures subsequent to treatment give strong support to the claim of efficacy of the TLM for reducing fear associated with speaking in public. The fact that change was of the same magnitude even for the wait-listed subjects adds to the robustness of the evidence of TLM’s efficacy. The TLM resulted in substantial decreases or complete eliminations of fear, accompanied by positive changes in confidence and reduced negative sensations felt during speaking in public in the experimental group. Overall, the TLM appears to have potential as an effective, quick, and convenient procedure to eliminate the fear of speaking in public.

It is true that the measures we used all involve self-report, but, as noted earlier, the complaint with which people began was self-report. Moreover, we do not think that measures such as observations would necessarily be informative, and they would have been difficult to arrange in a way that would produce reliable findings. Some people with high levels of anxiety are able to cover it up very well, and other people with no anxiety at all can appear flustered if they have not prepared well. Thus, for a study that is attempting to determine if an intervention is effective in eliminating a subject’s experience of anxiety, asking the subject to rate his level of fear (and his related physical symptoms) both before and after the intervention is the best option to reliably determine whether or not the intervention is successful.Because the treatment group was tested afrterafter giving a speech and the control group was not, it might be argued that the active treatment ingredient is the exposure to public speaking, rather than TLM, given the substantial evidence of the effectiveness of exposure methods for social anxiety. There are two answers to this argument: First, subjects from both groups had spoken many times prior to the study without any significant reduction in the fear. In fact, Toastmaster membership requires regular speaking. Second, to the extent desensitization works, it requires repeated exposure, not one. It would make little sense to claim that subjects who spoke regularly at Toastmaster meetings who reported a fear level with a mean of “7” reduced the fear to a mean of “1.5” merely” merely by giving one additional talk.

We want to argue at this point that TLM is a profoundly psychological method, a direct application of psychological constructs and principles to the effecting of behavioral change (Sechrest & Smith, 1994). The method is centered on the concept of belief and a paraphrasing of the idea of isomorphism of experience and action (Campbell, 1963). That is, in general, and, we think, to a very great extent, people will act on the basis of what they believe to be true as a result of their prior experiences and mental processing of them. If people believe that another person is liable to harm them, they will want to stay away from that person. If people regard some situation as fearsome and they believe themselves to be incapable of mastering their fear, they will avoid that situation. TLM is a way of helping people recognize and eliminate beliefs that, however warranted they may once have been, are no longer relevant to the problems that face them.

TLM has a good bit in common with CBT, RET, and other generally cognitive methods of therapeutic intervention, but it is not simply a reformulation of any of them. In its emphasis on eliminating beliefs, rather than learning to cope with them, it is distinctive. It is also distinctive in its claim that the problems resulting from those beliefs can be eliminated entirely, not just reduced by some degree. The appropriate outcome test for TLM is a category change (from having a problem to not having any problem) rather than a reduction in the mean value of the problem. That is a bold claim, and it remains to be seen whether it can be upheld for a wide range of problems. The present results are certainly suggestive of the possibility that TLM might be able to do just what it says.

Public speaking anxiety was chosen for the sake of convenience in completing an initial test of TLM, not out of any particular interest in providing a treatment for public speaking problems, per se. We believe that the results of this study should be interpreted as demonstrating that TLM may be a useful intervention for dealing with mental and behavioral disorders that are to some extent debilitating. Our plans for the immediate future are to develop a strategy by means of which it will be possible to recruit a number of therapists trained in TLM who will agree to participate in randomized trials to assess the usefulness of TLM in treating a range of problems common in clinical settings.

Information about TLM is readily available on Lefkoe’s website (www.decisionmaker.com). The aim is to make TLM generally available to clinicians as an alternative tool for discharging their professional responsibilities.

We do need, obviously, to determine how long the effects of TLM are sustained. Six-month follow-up questionnaires available currently for 23 of the 37 subjects indicate that the TLM approach has a long-sustained effect for our primary variable of interest, the experience of fear while speaking in public. Ratings based on 23 returned questionnaires range from 1 to 4 with an average of 1.9 (s.d. = 1.0), values that are not different from those obtained at posttest. Informal conversations with a random sample of seven subjects two years following the study indicated that the fear was “still gone.”The impressions of clinicians who have used TLM are that the effects are quite durable. The effects are, for one thing, likely to be self-reinforcing. Although TLM does not make direct use of exposure in the reduction of anxiety, in fact, in vivo exposure is meant to occur in every case. In the present study of public speaking anxiety, for example, all subjects actually made a public speech of some sort after treatment, but before completing the posttest measures. That experience, if successful—and all instances seem to have been successful—should constitute a strong exposure intervention.

One additional point that is worth attention is that the intervention reported on here was conducted entirely by telephone; the facilitator, Lefkoe, never saw any of the participants. Although participants were all residents of west coast communities, mostly in the Bay area, that was solely because they were recruited from public speaking clubs identified by Lefkoe. In principle, the intervention could have been delivered anywhere in the English-speaking world. It is also important that the intervention was actually quite brief. These characteristics of TLM, for the kinds of problems exemplified by fear of public speaking, indicate that the intervention should be highly cost-effective and that it could be made widely available.

A potential limitation of this study was the dependency on self-report data. However, we find the consistent response patterns of subjects in both groups to be compelling enough to rule out demand characteristics often associated with self-report data. At the very least, we think these results provide a strong basis for recommending further rigorous testing of TLM.


Bruskin Associates (1973). What are Americans afraid of? The Bruskin Report 53, 27.

Campbell, D.T. (1963). Social attitudes and other acquired behavioral dispositions. In S. Koch (Ed.), Psychology: Vol. 6, A Study of a Science (pp.94-172). New York: McGraw- Hill.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Lefkoe, M. (1997). Re-create Your Life: Transforming yourself and your world with the Decision Maker® Process. Kansas City, KS: Andrews and McMeel. Lefkoe, M. (2001). Behavior Change Doesn’t Have to be Difficult , California Psychologist, December, 2001

Lefkoe, M. (2001). Everyone knows that you can’t eliminate fundamental beliefs quickly and permanently. Are you sure? California Therapist, May, 2001

Motley, M.T. (1988, January). Taking the terror out of talk: thinking in terms of communication rather than performance helps us calm our biggest fear. Psychology Today, 22(1), 46-49.

Paul, G.L. (1966). Insight vs. desensitization in psychotherapy. Stanford, CA: Stanford University Press.

Richmond, V. P., & McCroskey, J. C. (1995). Communication: Apprehension, avoidance, and effectiveness (4th ed.). Scottsdale, AZ: Gorsuch Scarisbrick.

Sechrest, L., & Smith, B.H. (1994). Psychotherapy is the practice of psychology. Journal of Psychotherapy Integration, 4, 1-30.