INTERPERSONAL INTERACTION AND PERSONALITY
CHANGE IN GROUP PSYCHOTHERAPY*
W. H . COONS, Ph.D. and E . P.
1
In recent years many of the traditional
assumptions about the nature of psycho­
therapy have been subjected to new and
critical scrutiny. One of the challenged as­
sumptions is concerned with insight as
a crucial condition for personality change.
For example, Collier (3), and Mendelsohn,
Yates and Peterson (14) have contended
that strong positive affect is more important
than insight. Mellinger (13) has stressed
trust. Coons (5, 4) has claimed that inter­
personal interaction in a warm accepting
environment is the dynamic of change.
Strupp (20) has noted that therapist warmth
toward patients is accompanied by more
favourable diagnostic and prognostic expect­
ations and greater communication of em­
pathy. And Snyder's (19) term 'relationship'
seems to refer to much the same phe­
nomenon.
Concurrent with the challenge to the
traditional explanations of the nature of
psychotherapy has been an increasing con­
cern about who can be therapists. Albee (1)
and Schofield (18) have stressed the impor­
tance of finding alternatives to psycho­
therapy given by professionals. Rioch and
her colleagues (16), and Carkhuff and Truax
(2) have adduced evidence that, with training,
nonprofessionals can be effective thera­
pists, while Poser (15) has made the same
claim for therapists with no training or
special knowledge.
These developments have important im­
plications for the practice of psychotherapy
and for the organization of hospital therapy
programs. If positive affect or trust or
interpersonal interaction can produce per­
sonality change, and if untrained therapists
"The data for this study were collected while both
authors were at the Ontario Hospital, Hamilton. The
research was supported by Federal Mental Health Grant
No. 605-5-292 and Ontario Mental Health Foundation
Grant No. 78.
Manuscript received April, 1969.
'Professor of Psychology, York University and Psy­
chological Consultant, Whitby Psychiatric Hospital.
2
Staff Psychologist, Whitby Psychiatric Hospital.
Canad. Psychiat. Ass. 3. Vol. 15 (1970)
347
PEACOCK, M.A.*
can be the change agents, the range of
potential therapists is greatly extended. In
view of the arguments by Albee (1), Scho­
field (18) and others about the hopelessness
of ever meeting the social need with pro­
fessional therapists, the extension of the
therapeutic role to untrained personnel has
great practical significance.
The present research is an extension of
that reported by Coons (5), who compared
the effects on mental hospital patients of
two types of group psychotherapy conduct­
ed by professional therapists. In one type
(Interaction Therapy), the technique stressed
interaction among members of the group
without reference to personal difficulties
of individual patients. In the other condition
(Insight Therapy), the technique stressed
cognitive understanding of personal difficul­
ties. It was found that patients in the Inter­
action Therapy condition showed signifi­
cantly greater improvement in Wechsler-
Bellevue and Rorschach performance than
did those patients in the Insight Therapy
condition. Patients receiving Insight Ther­
apy did not differ significantly from those
in control groups who had received no
planned therapy at all.
These results led to the general hypo­
thesis that the systematic promotion of
interpersonal interaction in small stable
groups as part of the patient's everyday ex­
perience could be a significant factor in
helping him toward improved adjustment.
Research design
The design was a simple 2 x 2 x 2 fac­
torial randomized groups design (6). Four
male and four female groups of seven pa­
tients each were assigned at random to one
of four treatment combinations, each of
which lasted for six weeks: Formal Group
Psychotherapy plus Random Ward Inter­
action, Formal Group Psychotherapy plus
Organized Ward Interaction, Random Ward
Interaction only (with no Formal Group
 348 CANADIAN P S Y C H I A T R I C ASSOCIATION J O U R N A L
Psychotherapy), and Organized Ward Inter­
action only (with no Formal Group Psycho­
therapy).
Organized Ward Interaction was defined
operationally as the interaction which oc­
curred as a function of membership in an
identifiable group to which seven patients
were assigned on a hospital ward. The group
identity was maintained for most of the
group's daily hospital activities. The mem­
bers of the group ate at the same table,
slept in the same dormitory and had their
recreational, occupational and industrial
therapy as a distinct group under the full-
time direction of a group leader who was
a ward nursing assistant of the same sex as
the members of the group. The goal of the
group activities was the development of high
group cohesiveness. A psychologist met
with the group leaders at regular weekly
meetings and was available for emergency
consultations.
Random Ward Interaction referred to the
uncontrolled socialization which is typical
of most mental hospital wards. Under this
condition each patient was free to engage
selectively in interpersonal interaction with
<anyone on the ward according to his own
inclination, or as he was required to become
involved by the ward staff. His recreational,
industrial and occupational therapy were
arranged individually without reference to
his membership in the research group.
Random Ward Interaction was a control
condition which was designed to provide a
baseline for the evaluation of the treatment
procedures.
Formal Group Psychotherapy referred to
30 hours of formal group psychotherapy
over a period of six weeks. The therapy was
designed to foster intra-group interaction in
a warm, accepting and permissive atmos­
phere. The technique was .basically non-
directive (17, 8). The same professional
psychotherapist conducted all the Formal
Group Psychotherapy groups.
To justify the retention of the control
subjects (Random Ward Interaction) in hps-;
pital. for the experimental period it, was
Vol. 15, No. 4
necessary to offer some special form of
treatment. The design of the experiment
required that this treatment be unrelated to
the interpersonal interaction variables being
studied and common to all subjects. Con­
sequently, the program for all subjects, ex­
perimental and control, included viewing a
series of six mental health films. This treat­
ment was designated Audio-Visual Therapy.
None of the subjects was receiving (or
had recently received) electroconvulsive
therapy during the experimental period.
Whether or not a subject received psycho­
tropic medication during the course of the
experiment was determined by his physician
on the basis of his ward behaviour and
without reference to his status as a subject
in the study. AH patients were free of psy­
chotropic medication for at least 48 hours
prior to the pre- or post-assessments and
throughout the assessment periods.
Criterion Measures
The Wechsler Adult Intelligence Scale
(WAIS), the Hospital Adjustment Scale
(HAS) and the Rorschach were employed
as the criterion measures. The WAIS and
the Rorschach assessments were carried out
by the same examiner. He was not informed
of which patients were in which group and
he deliberately tried to avoid conversations
which might reveal this information. During
the post-therapy assessment he occasionally
suspected which treatment condition a pa­
tient had undergone. For scoring purposes
all protocols were coded so that the exam­
iner did not know whether the protocols
came from the pre- or post-assessment.
The Wechsler Adult Intelligence Scale
(21) was used to measure pre- and post-
treatment intellectual efficiency. Improved
intellectual efficiency is assumed to indicate
improvement since it reflects a freeing of
intellectual resources with a consequent in­
crease in. the- individual's ability to cope
intellectually with problem situations. Dif­
ferent but' approximately equivalent forms
of this test were employed by Coons (5),
and the• results significantly discriminated
between his experimental groups.
 August 1970 GROUP PSYCHOTHERAPY 349

The Hospital Adjustment Scale (12) con­ ton. Each was selected on the basis of suit­
sists of 90 statements descriptive of the ability for group psychotherapy. The criteria
behaviour of hospitalized psychiatric pa­ used were:
tients, and purports to provide an estimate
1) Ability to maintain at least passive
of hospital adjustment of adult patients of
attention;
either sex. The scale is designed to be com­
pleted by any of the ward staff who are 2) Sufficient contact with reality to be
familiar with the day-to-day behaviour of minimally disruptive to the group;
the patient. Each of the 90 statements com­ 3) Grade V I I I education or a pre-treat­
prising the scale is marked as 'True', 'Not ment I.Q. of at least 90;
True', or, in some cases, 'Does not Apply'. 4) No known organic impairment.
The inter-rater reliability is reported to be
The therapist for the group psychother­
.84. McReynolds and Ferguson (12) refer
apy selected all the patients. Psychiatric
to unpublished studies of the validity of the
nosology was ignored in the selection pro­
HAS which indicate that it is effective in
cedure but the majority (N = 38) were
differentiating between patients approaching
diagnosed as schizophrenic.
discharge from hospital and patients judged
Successive groups (four male groups and
to be extremely disturbed.
four female groups) of seven patients each
In the current study the 90-item scale were selected. After a group had been
was completed on all 5s before and after picked the treatment was chosen randomly.
the experimental period. The ward super­
The therapist had no knowledge of which
visor was made responsible for the ratings
experimental treatment the group would
which were the consensus of those ward
receive until after his selection of the mem­
staff who best knew the patient concerned.
bers of the group had been completed.
The Rorschach (9, 10) pre- and post-
Fifty-six 5s were selected,tt ranging in
treatment protocols were analyzed 'blind'
age from 15 to 51 with a mean age of 28.6
by three experienced Rorschach judges,?
years. Female 5s were slightly older than
operating independently. Each judge knew
the males (Female mean age — 31.9, S.D.
only that each pair of protocols was from
— 9.9; Male mean age — 25.9, S.D. —
the same subject (identified by a code num­
6.9). The total duration of all hospitaliza­
ber, age, sex, education, occupation and
tions ranged from one to 169 weeks, with
marital status) and that one of the protocols
a mean total hospitalization of 45.4 weeks.
was pre-treatment and the other post-treat­
Male 5s had been hospitalized for somewhat
ment. He did not know which was pre- and
longer periods of time than female 5s (Male
which was post-treatment, or the nature of
mean duration of hospitalization — 60.0,
the treatment the patient received.
S.D. — 47.2. Female mean duration of
Each judge was instructed to determine
hospitalization — 31.2, S.D. — 26.3). In
which of each pair of protocols indicated
the total sample the range of years of
the better level of adjustment, or that there
schooling was from 8 to 16 years, with a
was 'no difference' in adjustment indicated
mean of 10.2 years. There was no signifi­
by the two protocols. When the protocol
cant difference in the mean number of years
which two of the three judges rated 'better'
of schooling for males and females. Never­
was the post-treatment record, that 5 Was
theless the mean pre-treatment WAIS Full-
considered to have shown improved adjust­
Scale I.Q. of the male 5s was significantly
ment during the therapy period. Otherwise,
5 was considered to have 'not improved'. greater than that of the females (Male
mean I.Q. — 100.0, S.D. — 11.6; Female
Subjects mean I.Q. — 88.9, S.D. — 11.6, p<.005X
The subjects used in this research were
ttDuring the experiment, one subject was lost from
all patients at the Ontario Hospital, Hamil- each of the following groups: Male Formal Group
Psychotherapy plus Random Ward Interaction; Male
tProfessor F . W. Burd, Dr. H . N. Blackwell and Dr. Random Ward Interaction; Female Random Ward
K. G. Ferguson. Interaction.
 350 CANADIAN P S Y C H I A T R I C ASSOCIATION J O U R N A L V o l . 15, N o . 4

TABLE I

M E A N S AND STANDARD D E V I A T I O N S F O R F U L L - S C A L E I . Q.
C H A N G E S FROM P R E - TO P O S T - T R E A T M E N T ASSESSMENTS

Male Female

FGP FGP OWI RWI FGP FGP OWI RWI

plus plus (No (No plus plus (No (No
OWI RWI FGP) FGP) OWI RWI FGP) FGP)

Mean 8.7 6.9 2.7 2.3 4.9 5.4 4.4 3.6

S.D. 5.2 3.0 3.9 5.2 . 2.5 3.2 3.7 4.4

F G P — Formal Group Psychotherapy

OWI — Organized Ward Interaction
R W I — Random Ward Interaction

Hypotheses tion would more than match the greater

Basic to the research is the assumption
that interpersonal interaction is an impor­
tant component of the therapeutic process.
Since interaction is systematically fostered
by both the Formal Group Psychotherapy
and the Organized Ward Interaction treat­
ment conditions, subjects in those groups
should show greater improvement on all
three criterion measures than do the sub­
jects in the Random Ward Interaction con­
dition.
The relative effectiveness of the Formal
Group Psychotherapy and the Organized
Ward Interaction conditions hinges on the
frequency and quality of the interaction.
These two conditions differ on several
dimensions which include: a) the skill of
the staff member attending the group (pro­
fessional therapist or nurse-aide) in pro­
viding a consistently warm and accepting
social environment and in reinforcing inter­
action among the group members; b) the
potency as a reinforcer of the staff member;
and c) the time units in which the inter­
action takes place (one hour a day for the
Formal Group Psychotherapy condition
compared to eight hours a day for the
Organized Ward Interaction subjects). Be­
cause of the importance traditionally attach­
ed to the skill of the therapist in his role
as a social reinforcement machine program­
med by prior training and experience (11),
it was expected that the professional skill
of the Formal Group Psychotherapy condi­
time spent by the nonprofessional group
leaders of the Organized Ward Interaction
condition.
Thus it was hypothesized that on all three
criterion measures the amount of improve­
ment would be greatest for the combined
Formal Group Psychotherapy — Organized
Ward Interaction groups, followed by the
Formal Group Psychotherapy — Random
Ward Interaction groups, followed by the
Organized Ward Interaction groups, with
the Random Ward Interaction groups show­
ing least improvement. No differences in
improvement were expected between male
and female subjects.
Results
Wechsler Adult Intelligence Scale*
A summary of the pre- to post-treatment
changes in Full-Scale I.Q. scores is present­
ed in Tables I and I I . The effect of Formal
Group Psychotherapy was significant beyond
the 1% level (Table II). For 5s receiving
Formal Group Psychotherapy, the mean
change in Full-Scale I.Q. was 6.4 (S.D. —
4.0); for 5s not receiving Formal Group
Psychotherapy, the mean change in Full-
Scale I.Q. was 3.3 (S.D. — 4.6).
A similar three-way analysis of variance
of pre- to post-treatment changes in WAIS
Verbal I.Q. revealed no significant interac­
tions or main effects; analysis of WAIS
JBartlett's tests supported the assumption of homo­
geneity of variances of the Full-Scale, Verbal-Scale and
Performance-Scale data.
 August 1970 GROUP PSYCHOTHERAPY 351

TABLE II

SUMMARY T A B L E O F R E S U L T S O F 3-WAY A N A L Y S I S O F V A R I A N C E O F W A I S
F U L L - S C A L E I . Q. C H A N G E S F R O M P R E - TO P O S T - T R E A T M E N T ASSESSMENTS

Source df SS MS F P

Formal Group
Psychotherapy (A) 1 144.643 144.643 7.32 <.01
W a r d Interaction
Program (B) 1 5.785 5.785
Sex (C) 1 4.571 4.571
A X B 1 0.000 0.000
A X C 1 60.071 60.071 3.04 <.10(n.s.)
B X C 1 3.501 3.501
A X B X C 1 7.143 7.143
Error 45 889.143 19.759

Performance I.Q. indicated that the main
effect of Formal Group Psychotherapy was
significant at below the 5% level of con­
fidence. The mean change in Performance
I.Q. of 5s receiving Formal Group Psycho­
therapy was 8.0 (S.D. — 5.8); that of 5s
not receiving Formal Group Psychotherapy
was 3.9 (S.D. — 6.5).
Hospital Adjustment Scale
A three-way analysis of variance on the
pre-treatment HAS scores for all 5s reveal­
ed no significant effects. It was therefore
concluded that there were no initial differ­
ences in ward adjustment between male
and female groups or between patients sub­
sequently assigned to the various combina­
tions of Formal Group Psychotherapy and
Ward Interaction Programs.
The 2 x 2 x 2 analysis of variance per­
formed on the pre- to post-treatment HAS
change scores is presented in Table III. The
main effects of sex and ward program were
significant (<.05) indicating that female pa­
tients showed significantly more improve­
ment in hospital adjustment than males and
that patients receiving Organized Ward In­
teraction improved more than those in the
Random Ward Interaction condition. The
interaction between Formal Group Psycho­
therapy and Ward Interaction Program was
also significant « . 0 5 ) . Using Scheffe's mul­
tiple comparison procedure (6), it was found
that Formal Group Psychotherapy plus
Organized Ward Interaction resulted in
significantly greater improvement in hospital
adjustment than Formal Group Psychother­
apy plus Random Ward Interaction (F =

TABLE III

SUMMARY T A B L E OF 2 X 2 X 2 A N A L Y S I S O F V A R I A N C E O F P R E - TO P O S T - T R E A T M E N T CHANGE
SCORES ON T H E H O S P I T A L A D J U S T M E N T SCALE

Source df SS MS F P

F o r m a l Group
Psychotherapy (A) 1 462 88 462.88 1.75 n.s.
W a r d Interaction
Program (B) 1 1370 16 1370.16 5.18 <.05
Sex (C) 1 1390 02 1390.02 5.25 <.05
A X B 1 1197 87 1197.87 4.53 <.05
A X C 1 24 44 24.44
B X C 1 355 02 355.02 1.34 n.s.
A X B X C 1 80 16 80.16
Error, 45 11904 29 264.54
 352 CANADIAN P S Y C H I A T R I C ASSOCIATION JOURNAL Vol. 15, No. 4

TABLE I V

SUMMARY OF 2 X 2 CONTINGENCY T A B L E S SHOWING T H E F R E Q U E N C Y OF " I M P R O V E D " AND " N O T

IMPROVED" RORSCHACH PROTOCOLS WITH THEIR ASSOCIATED X 2
V A L U E FOR E A C H PAIR OF T R E A T ­
MENT CONDITIONS

Treatment Condition Direction of Rorschach Protocol Change X 2

Improved Not Improved

F G P plus R W I 10 3 4.52 <.05

F G P plus O W I 4 10

F G P plus R W I 10 3 <1 n.s.

RWI 8 4

F G P plus O W I 4 10 2.39 n.s.

RWI 8 4

F G P plus O W I 4 10 0 n.s.
OWI 5 9

F G P plus R W I 10 3 4.49 <.05

OWI 5 9

RWI 8 4 2.10 n.s.

OWI 9
5

F G P — Formal Group Psychotherapy

O W I — Organized W a r d Interaction
R W I — RandomjWard Interaction

9.14; p<.05). Patients in Formal Group 4.52; p<.05) and Organized Ward Interac­
Psychotherapy plus Random Ward Inter­ tion alone (x = 4.49; p<.05).
2

action showed significantly poorer hospital

adjustment than patients in the other treat­
ment conditions (F = 10.30; p<.05).
Rorschach
A 2 x 4 Chi square test for independent
samples revealed that the proportion of pa­
tients rated 'improved' and 'not improved'
varied significantly among the four treat­
ment conditions (x = 8.76; p<.05). Fur­
2
ther two-way comparisons between treat­
ment pairs were made. These data are pre­
sented in the 2 x 2 contingency tables of
Table IV; as can be seen from this Table,
Formal Group Psychotherapy plus Random
Ward Interaction resulted in a significantly,
greater proportion of improved Rorschach
protocols than Formal Group Psychother­
apy plus Organized Ward Interaction (x =
Discussion
On all three criterion measures, Formal
Group Psychotherapy emerges as a signifi­
cant agent of change. Analysis of Wechsler
Adult Intelligence Scale scores shows that
Formal Group Psychotherapy was signi­
ficantly more efficacious in producing
change than was any other treatment or
combination of treatments. These results
confirm our earlier finding (5) that Formal
Group Psychotherapy produces a significant
improvement in intellectual functioning.
The '• inference is that this type of treatment
results in a freeing of intellectual resources
so that patients are better able to cope
intellectually, at least with structured prob­
lem situations of the sort sampled by the
2
WAIS. ,
 August 1970 GROUP P S Y C H O T H E R A P Y
The Rorschach results also indicate that
Formal Group Psychotherapy produces
greater improvement than Organized Ward
Interaction, either alone or in combination
with Formal Group Psychotherapy. Inspec­
tion of the Rorschach data presented in
Table I V further supports the finding that
Organized Ward Interaction has a detrimen­
tal effect on the impact of the Formal Group
Psychotherapy. The reason for this is not
entirely clear but it seems plausible to
assume that the professional group therapist
and the ward-staff group leaders work to­
ward conflicting goals. The ward staff tend
to view 'good patients' as those who con­
form to the rules and regulations of the
ward; professional therapists tend to define
improvement in terms of greater independ­
ence of action which may run counter to
some of the restrictive demands of the ward
regime. The Hospital Adjustment Scale
data appear to reflect this conflict. The main
effects of both sex and ward program were
significant as was the interaction between
Formal Group Psychotherapy and Organ­
ized Ward Interaction. The criterion mea­
sure here was completed by the ward staff
who were involved regularly with the pa­
tients in the study and knew the nature
of the treatment condition (unlike the
Rorschach judges and the WAIS examiner).
The Hospital Adjustment Scale evaluations
can be expected to reflect any bias in favour
of conformity to rules imposed by the ward
staff and may well reflect a bias in favour
of those groups participating in programs
in which the ward staff were heavily in­
volved, that is those which included Organ­
ized Ward Interaction in their treatment
condition. Thus, two sources of judge-bias
(positive evaluation of conformity behaviour
and halo effect) may account for the diver­
gent Hospital Adjustment Scale results.
Nevertheless, whatever the validity of this
speculation, it is clear that Formal Group
Psychotherapy is the only treatment condi­
tion which had a significant impact, alone
or in combination with other treatment, on
all three criterion measures.
How are we to rationalize the finding
that our professional therapist, who met his
353
groups for only an hour per day, five days
a week, produced greater improvement than
did our group leaders who were involved
for eight hours per day, five days a week?
After all, these group leaders were en­
thusiastic and well-motivated persons who
performed their functions with unusual
diligence.
It is conceivable that the most obvious
explanation holds: that the professional
therapist's skill was the crucial factor in
bringing about the improvement. There is,
however, another possibility which is re­
ceiving increasing attention among psycho­
therapists, that of the expectancies patients
bring to the therapeutic situation. The
group leaders were lower-echelon hospital
staff who were stereotyped by the patients
as having low status and minor professional
competence. By contrast the professional
therapist was viewed as having high status
and substantial professional competence. It
seems likely that the subjects in the Formal
Group Psychotherapy treatment condition
entered therapy with a higher expectation
of favourable outcome than did those as­
signed to the Organized Ward Interaction
groups. A consequence of this patient atti­
tude might well have been that the profes­
sional therapist possessed greater reinforcing
value than did the group leaders, and thus
was able to shape more efficiently the be­
haviour of his subjects. (See Goldstein [7],
for an explication of the role of patient
expectancies in psychotherapy.)
Whatever the ultimate explanation, the
results of the present study require that we
reserve judgement about Poser's (15) con­
tention that "traditional training in the
mental health professions may be neither
optimal nor even necessary for the promo­
tion of therapeutic behaviour change in
mental hospital patients (p. 289)." Before
we can conclude that professional training
is irrelevant it is incumbent upon us to
clarify the complex interrelations of factors
influencing the therapeutic process, factors
such as the interaction of age, sex and status
of patient and therapist and the expectancies
which both therapist and patient bring to
the therapeutic transaction.
 354 CANADIAN P S Y C H I A T R I C ASSOCIATION J O U R N A L
Summary
The study was concerned with comparing
the efficacy of three forms of social inter­
action in modifying the behaviour of mental
hospital patients. Eight groups of seven
patients each were assigned, in a simple
randomized design, to combinations of
Formal Group Psychotherapy (conducted
by a professionally trained therapist), Or­
ganized Ward Interaction (conducted by
untrained group leaders), or Random Ward
Interaction (no planned therapy). Equal
numbers of male and female groups were
involved.
The dependent variables were: 1) The
Wechsler Adult Intelligence Scale; 2) The
Hospital Adjustment Scale; 3) The Ror­
schach Technique of Personality Diagnosis.
Formal Group Psychotherapy emerged as
the only treatment condition which had a
significant impact, alone or in combination
with other treatment, on all three criterion
measures. It was concluded that judgement
must be reserved on the contention of sev­
eral authors that traditional professional
training in psychotherapy is unnecessary
for the promotion of therapeutic change.
References
1. Albee, G. W.: Manpower needs for men­
tal health and the role of psychology.
Canadian Psychologist, 1965, 6a: 82-92.
2. Carkhuff, R. R. and Truax, C. B.: Lay
mental health counselling: the effects of
lay group counselling. Journal of Consul­
ting Psychology, 1965, 29: 426-431.
3. Collier, R. M.: Consciousness as a regu­
latory field: a theory of psychotherapy.
Journal of Abnormal and Social Psycho­
logy, 1957, 55: 275-282.
4. Coons, W. H.: The dynamics of change
in psychotherapy. Canadian Psychiatric
Association Journal, 1967, 12: 239-245.
5. Coons, W. H.: Interaction and insight in
group psychotherapy. Canadian Journal of
Psychology, 1957, 11: 1-8.
6. Edwards, A. L . : Experimental Design in
Psychological Research, p. 175, p. 154,
Toronto: Holt, Rinehart & Winston, 1964.
7. Goldstein, A. P.: Therapist-Patient Ex­
pectancies in Psychotherapy. New York:
Macmillan Co., 1962.
Vol. 15, No. 4
8. Hobbs, N.: Group-centered psychother­
apy. In Rogers, C. R., Client-Centered
Therapy. Boston: Houghton-Mifflin, 1951.
9. Klopfer, B. and Kelley, D.: The Rorschach
Technique. New York: World Book Co.,
1942.
10. Klopfer, B., Ainsworth, Mary D., Klopfer,
W. I . , and Holt, R. R.: Developments in
the Rorschach Technique. New York:
World Book Co., 1954.
11. Krasner, L.: The therapist as a social re­
inforcement machine. In H. H. Strupp and
L. Luborsky (Eds.), Research in Psycho­
therapy, Vol. I I . Washington, D.C.:
American Psychological Association,
1962: 61-94.
12. McReynolds, P., and Ferguson, J. T.:
Clinical Manual for the Hospital Adjust­
ment Scale. Palo Alto, California: Leland
Stanford University Press, 1953.
13. Mellinger, G. D.: Interpersonal trust as a
factor in communication. Journal of Ab­
normal and Social Psychology, 1956, 52:
304-309.
14. Mendelsohn, R. M., Yates, J. W., and
Peterson, L . : The experiential aspect of
psychotherapy. American Journal of Psy­
chotherapy, 1957, 11: 254-261.
15. Poser, E . G.: The effect of therapists'
training on group therapeutic outcome.
Journal of Consulting Psychology, 1966,
30: 283-289.
16. Rioch, M. J., Elkes, C., Flint, A. A.,
Usdansky, B. S., Newman, R. G., and
Silber, E . : National Institute of Mental
Health pilot study in training mental coun­
selors. American Journal of Orthopsychi­
atry, 1963, 33: 678-679.
17. Rogers, C. R.: Client-Centered Therapy.
Boston: Houghton-Mifflin, 1951.
18. Schofield, W.: Psychotherapy: the Pur­
chase of Friendship. Englewood Cliffs,
N.J.: Prentice-Hall, 1964.
19. Snyder, W. U.: The Psychotherapy Rela­
tionship. New York: The Macmillan Co.,
1961.
20. Strupp, H. H.: The outcome problem in
psychotherapy revisited. Psychotherapy:
Research and Practice, 1963, 1: 1-13.
21. Wechsler, D.: The Measurement and Ap­
praisal of Adult Intelligence. Baltimore:
Williams and Wilkins, 1958.
Resume
Dans cette etude, on compare l'effica-
cite de trois formes d'interaction sociale qui
viennent modifier le comportement des
malades dans les hopitaux psychiatriques.
Huit groupes de sept malades chacun ont
August 1970 GROUP PSYCHOTHERAPY 355

ete assignes, selon un modele simple choisi a l'hopital; 3) la Technique Rorschach de

au hasard, a des combinaisons de psycho­
therapie de groupe reconnue (dirigee par
un therapeute ayant recu une formation
professionnelle), d'interaction organisee
dans les salles (dirigee par des moniteurs
n'ayant pas recu de formation) ou d'inter­
action au hasard dans les salles (sans thera-
pie organisee). Les groupes d'hommes et de
femmes y etaient en nombre egal.
Voici quelles etaient les variables de-
pendantes: 1) l'Echelle Wechsler d'intelli-
gence des adultes; 2) l'Echelle d'adaptation
diagnostic de la personnalite.
L a psychotherapie de groupe reconnue
s'est revelee la seule condition de traitement
qui ait eu un impact significatif, seule ou
combinee a un autre traitement, parmi les
trois mesures de criteres. On en a conclu
qu'il faut accepter avec des reserves l'af-
firmation de plusieurs auteurs qui pre-
tendent que la formation professionnelle
conventionnelle en psychotherapie n'est pas
necessaire pour provoquer une modification
therapeutique.

The meeting of two personalities is like the

contact of two chemical substances; if there
is any reaction, both are transformed.
Modern Man in Search of a Soul

Carl Gustav Jung

1875- 1961