Letters to the Editor
Psychother Psychosom 2006;75:190–191
F.M. Alexander Technique in the Treatment of
Stuttering – A Randomized Single-Case
Intervention Study with Ambulatory
Dorothea Schulte, Harald Walach
School of Social Sciences and Samueli Institute European
Offi ce, University of Northampton, Northampton , UK
The Alexander technique (AT), a body-oriented method using
mental direction of awareness, is named after its founder, the Aus-
tralian actor Frederick Matthias Alexander (1869–1955), aiming at
the modifi cation of physiologically unfavorable automated habits
and postures. Habitually executed movement patterns can be un-
favorable and damaging. Alexander himself was able to cure his
own functional dysphony by employing the principles of his tech-
nique. In addition, he also successfully worked with stutterers.
Stuttering is frequently associated with high tension in the mus-
cles involved in speaking  and with neuromuscular coordination
problems  . A reduction of such an increased muscular activity
leads to reduction of the symptoms of stuttering [3, 4] . The most
conspicuous feature of stuttering is the self-conditioned secondary
symptoms, i.e. the patients’ effort of fi ghting and overcoming stut-
tering when it occurs [5, 6] . Gradually, the strain increases and the
secondary symptoms are automated into a strenuous and burden-
ing habit. AT offers a systematic way of modifying such habits as
stuttering-related increase in muscle tension and thus a possibility
of coping with fears of expectancy.
Scientifi c evidence for AT’s effi cacy is missing despite its popu-
larity  . We decided to combine an experimental evaluation tech-
nique with the highly individual approach of AT. For ethical reasons,
we decided to recruit subjects who had experienced some therapy,
but who had not been treated for 1 year prior to the study and who
still had considerable residual problems with their stuttering. The 2
subjects came from a pool of local self-help groups and therapists
who were alerted to the study and signed up for free treatment.
We used a single-case, time series design with a randomized
beginning of the intervention, as described by Edgington [8, 9] and
Wampold and Furlong  . The random component in our study
was the randomized allocation of the beginning of the intervention
within a 30-day intervention period after a 5-day baseline. Data
acquisition was by ambulatory monitoring and in situations of ev-
eryday life, using a pocket PC by Psion, series 3a (Psion™ PCL,
London, UK)  . The subjects were asked to reply to questions
concerning their experiencing of and coping with stuttering imme-
diately after episodes of speaking 3 times per day. Questions were
related to anxiety, making contact with the body, using Alexander
directives, general feeling, stuttering, quality and acceptance of
stuttering, avoidance of words or letters, speed and trying to infl u-
ence stuttering by diverse techniques including AT.
Variables were evaluated using a randomization test according
to Edgington  . Randomization tests calculate a simple test sta-
tistic and permute this calculation through all possible arrays of
data. The test statistic we used was the mean difference of a target
variable between baseline, up to the intervention point, and the
intervention period including the postobservation period. The
number of all the hypothetical differences, which are larger or equal
to the one obtained empirically, divided by the number of all po-
tential differences gives the true probability that the empirically
observed value could have occurred by chance. In addition to the
data acquired by Psion entries, audio and video recordings of con-
versations within and outside of a treatment session were done in
order to detect modifi cations in stuttering severity.
The subjects attended 30 lessons in total, which took place 2–4
times a week. Lessons were basically structured in two sections:
(1) teaching of the basic principles of AT; (2) applying AT while
The fi rst subject was a 27-year-old female student who had been
stuttering since the age of 3 with a history of multiple therapies.
Stuttering was still a residual problem which occurred in diffi cult
situations. The second subject was a 47-year-old male professional
working in an industrial enterprise. He has been stuttering since the
age of 5 and has had multiple speech therapies and psychotherapies,
including pharmacotherapy because of depression. Both subjects
described their stuttering problem as medium in sever ity.
Although the subjects did not display high stuttering rates at the
beginning of the study, a further improvement could be demon-
strated in both subjects. The female subject’s stuttering rate was
between 2.5 and 4.6% before attending AT lessons, and only 0.3
and 0.74% after about 14 lessons to the completion of this study
( fi g. 1 ). The male subject’s rates improved accordingly (from
5.24/8% to 0.64/2%).
Table 1. Results of the randomization test for the signifi cant Psion
items out of the 17
Body contact prior to speaking?1.033.59 0.02
Attempt at positively infl uencing stuttering?
Sensing of speaking movements
AT directives prior to speaking?
AT directives during speaking?
2.89 Body contact during speaking?
Felt comfortable in the situation?4.92
Successful infl uencing of stuttering?
First line = Female subject; second line = male subject. The ab-
solute magnitude of mean values for baseline (BL) and treatment
(TR) are given as well as the error probability (p). Error probabilities
are derived by dividing the number of potential permuted differ-
ences that are equal or larger than the actual one by the number of
all possible permutations (i.e. 179).
Letters to the Editor
Statistical evaluation of the Psion entries demonstrates a sig-
nifi cant effect (p ! 0.05) of AT on the successful infl uence on stut-
tering. Eight variables showed signifi cant results in the randomiza-
tion test with a signifi cance level of 5% ( table 1 ).
Our study, which to our knowledge is the fi rst intensive single-
case study of AT using ambulatory monitoring, has shown that in
2 cases of moderate residual stuttering 18–30 lessons of AT have
led to a further improvement both in symptoms and in self-per-
ceived coping and self-effi cacy. The qualitative analysis showed
that the subjects felt more able to control their stuttering and more
By way of self-criticism one has to note the following limita-
tions. Since this was the fi rst study of its kind, we used a rather
broad band of outcome variables. This leads to a potential problem
with multiple testing. In both cases, 5 out of 17 variables show sig-
nifi cant effects below the conventional limit (from p = 0.02 to p =
0.04). With 20 variables, only 1 would be expected to be signifi cant
AT is a complex intervention, and an argument can be made
that there is no single component that produces the effect but the
interaction of all elements  . Thus, our study does not say any-
thing about any particular component, which could only be studied
by large dismantling studies. Meanwhile, AT as a whole seems to
have stood one empirical test of therapeutic effectiveness in re-
sidual problems of stuttering.
The authors want to thank the subjects of this study for the
strains they put on themselves by complying with the request of
entering their Psion data. This study would never have been pos-
sible without their efforts and their open-mindedness and honesty.
D.S. wants to thank Elisa Ruschmann for her supervision of her
work with the subjects, as well as speech therapist Karl Schneider
for his supervision regarding stuttering therapy. We are grateful to
Paul Hüttner who helped with programming the Psions. H.W. is
supported by the Samueli Institute.
FI 1FI 2FI 3 ATL
Stuttering frequency (%)
Stuttering frequency during free speech
Stuttering frequency during phone calls at home
Fig. 1. Stuttering frequency in the female subject dur-
ing free speech with the investigator at the beginning of
AT lessons (ATL) and during phone calls during natural
situations at home. First interview (FI) values are mean
values from 1- to 2-min talking sequences from the be-
ginning, the middle of and the end of conversations,
respectively. The fi rst talking minute from the begin-
ning of AT lessons and from phone conversations was
1 Fox PT, Ingham RJ, Ingham JC, Hirsch TB, Downs JH, Martin C, Jera-
bek P, Glass T, Lancaster JL: A PET study of the neural systems for stut-
tering. Nature 1996; 382: 158–162.
2 Peters HFM, Hulstijn W, van Lishout P, Johannsen HS, Springer L: Tim-
ing and coordination of speech motor processes in stutterers; in Johannsen
HS, Springer L (eds): Stottern: Münster 19.–22.5.1993. Ulm, Verlag Pho-
niatrische Ambulanz der Universität Ulm, 1993, pp 22–42.
3 Bloodstein O: A Handbook on Stuttering, ed 5. San Diego, Singular Pub-
lishing Group, 1995.
4 Van Riper C: Die Behandlung des Stotterns. Köln, Demosthenes-Verlag
der Bundesvereinigung Stotterer-Selbsthilfe eV, 1999.
5 Starkweather CW, Givens-Ackerman J: Stuttering. Austin, Pro-Ed,
6 Bloodstein O, Eisenson J: Stuttering as an anticipatory struggle reaction;
in Eisenson J (ed): Stuttering: A Symposium. New York, Harper & Broth-
ers, 1958, pp 1–69.
7 Ernst E, Canter PH: The Alexander technique: a systematic review of con-
trolled trials. Forsch Komplementärmed Klass Naturheilkd 2003; 10: 325–
8 Edgington ES: Randomized single-subject experiments and statistical
tests. J Counsel Psychol 1987; 34: 437–442.
9 Edgington ES: Randomization Tests, ed 3. New York, Dekker, 1995.
10 Wampold BE, Furlong MJ: Randomization tests in single-subject designs:
illustrative examples. J Behav Assess 1981; 3: 329–341.
11 Fahrenberg J, Myrtek M (eds): Ambulatory Assessment: Computer-As-
sisted Psychological and Psychophysiological Methods in Monitoring and
Field Studies. Göttingen, Hogrefe & Huber, 1996.
12 Verhoef MJ, Lewith G, Ritenbaugh C, Thomas K, Boon H, Fonnebo V:
Whole systems research: moving forward. Focus Alternative Complement
Ther 2004; 9: 87–90.
Harald Walach, PhD, University of Northampton
School of Social Sciences and Samueli Institute European Offi ce
Boughton Green Road, Northampton NN2 7AL (UK)
Tel. +44 1604 892952, Fax +44 1604 722067