Group Title: interaction of the A-B status of psychiatric aids
Title: The interaction of the A-B status of psychiatric aids
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00102839/00001
 Material Information
Title: The interaction of the A-B status of psychiatric aids
Physical Description: Book
Language: English
Creator: Taylor, Donald John, 1944-
Copyright Date: 1975
 Record Information
Bibliographic ID: UF00102839
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 02811379
ltuf - AAT2014

Full Text



















THE INTERACTION OF THE A-B STATUS
OF PSYCHIATRIC AIDS AND EFFECTIVE COMMUNICATION
OF EXPECTANCIES TO SCHIZOPHRENICS






By





DONALD JOHN TAYLOR


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE
DOCTOR OF PHILOSOPHY







UNIVERSITY OF FLORIDA
1975

















To CDJaneDICt, anAi to Oy Ns C. nd i n Taylor

To Janet, and to my parents John C. and Helen Taylor.
















ACy1.~v~LyDC~ ,~


My warm thanks go to Dr. Vernon Van De Riet, who

showed much patience and understanding as Chairman of

both my Master's and Doctoral committees. I am also

grateful for the help and guidance of Dr. Harry Grater,

Dr. Don Avila, Dr. Franz Epting, and Dr. Richard Swanson.

Special thanks go to Dr. Art Wells for his help in

smoothing the way for me at the Northeast Florida State

Hospital where the data for this research was gathered.

Others for whose encouragement and advice I am

grateful are Dr. Anthony Zold and Dr. Frank Rath of

William Beaumont Army Medical Center.


iii














TA ,BLE OF' COTEI OTS

Page


ACf 'OWLE DGMENTS ................................... iii

LIST OF TABLES .............. .......... ......... vi

ABSTRACT ........................................... ix

INTRODUCTION ...................................... 1

The A-B Variable and the Interaction Hypothesis.. 1

Later Clinical Outcome Studies .................. 5

Clinical Studies of A-B Therapist Behavior ...... 10

Analogue Studies of A-B Psychotherapy Behavior .. 14

Studies of A-B Therapist Attitudes .............. 17

Personality Characteristics of As and Bs ........ 21

The A-B Variable and Unintended Interpersonal
Influence .................................... 27

Extending the A-B Variable to Females ........... 30

Implications of the Interaction of A-B Status
and Interpersonal Influence with Hospitalized
Patients ..................................... 31

Purposes of the Dissertation ..................... 33

HYPOTHESES ......................................... 37

Experimenter Expectancy Hypotheses .............. 37

Voice Quality Hypotheses .......... .. ...... ..... 38

Interpersonal Impressions Hypotheses ............ 39

METHODOLOGY ..... .... ............................... 41

Subjects ........................................ 41










Page


A ids ......................................... 41
P tiJ ents . . . . . . . . . . . . . . . . . . 42

Inst ruments ... ............ ......... .. ......... 45

Procedure ....................................... 47

Statistical Treatment .............................. 51

RESULTS ..... ....................................... 52

Experimenter Expectancy Hypotheses ............. 52

Voice Quality Hypotheses ......... ............... 56

Interpersonal Impressions Hypotheses ............ 63

DISCUSSION ......................... ................ 77

Perspective ..................................... 77

General Considerations and Conclusions .......... 81

APPENDICES ..................... .. ................. 88

A. Personal Interests Questionnaire ............ 88
B. Instructions for Psychiatric Aid ............ 91
C. Instructions to be Read to Patients ......... 92
D. Rating Scale ...................... .......... 93
E. Scale Used by Aids in Rating Patients ....... 94
F. Scale Used by Patients in Rating Aids ....... 95
G. Scale Used by Judges in Rating Vocal
Qualities of Aids ........................ 96

REFERENCES ........................................ 97

BIOGRAPHICAL SKETCH ................................. 103









LI ','T OF TABLiES


Table Page


.. Coi.) : o M e A:, vs. Male Bs, and
cFe'- ,< vs. iFe'tale Es on Experil enter
LEpiect cy Scores Atta ined with High and
Lo Sf;oc. l Cpompetence Schizophrenics ........ 54

2. Comi-iari -son of A Typre Es vs. B type Es; Male
Es vs. Fcima ec Es; and IIigh vs. Low Social
Competence Schizophrenics on Experimenter
ExpecLancies Scores Attained ................ 55

3. Comparison of Male As vs. Male Bs on Rated
Voice qualities when Reading Instructions to
Low Social Competence Ss .................... 59

4. Comparison of Female As vs. Female Bs on
Rated Voice Qualities when Reading Instruc-
tions to Low Social Competence Ss ........... 60

5. Comparison of Male As vs. Male Bs on Rated
Voice Qualities when Reading Instructions
to High Social Competence Ss ................ 61

6. Comparison of Female As vs. Female Bs on
Rated Voice Qualities when Reading Instruc-
tions to High Social Competence Ss .......... 62

7. Comparison of Male As vs. Male Bs, and
Female As vs. Female Bs on Ratings Made
by Low Social Competence Schizophrenics
on the Question "Did You Like the Aid?" ... 65

8. Comparison of Male As vs. Male Bs, and
Female As vs. Female Bs on Ratings made
by Low Social Competence Schizophrenics
on the Question "Did You Feel at Ease
with the Aid?" ............................. 66

9. Comparison of Male As vs. Male Bs, and
Female As vs. Female Bs on Ratings made by
Low Social Competence Schizophrenics on the
Question "Was the Aid Friendly?" ............. 66

10. Comparison of Male As vs. Male Bs, and Female
As vs. Female Bs on Ratings made by Low
Social Competence Schizophrenics on the
Question "Was the Aid Self-confident?" ...... 67










Table


11. Comparison of Male As vs. M-le Bs, and
Femalc As vs. Female Bs on Ratings made by
Low Social Competence Schizophrenics on
the Question "Was the Aid a Strong,
Dominant Person?" ............................ 67

12. Comparison of Female As vs. Female Bs on
Ratings made by Low Social Competence
Schizophrenics on the Question "Was the
Aid a Gentle, Kind Person?" ................... 68

13. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
on Ratings made by Male and Female As on
the Question "Did You Like This Patient?" .... 68

14. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male
and Female As on the Question "Did You
Feel At Ease With This Patient?" .............. 69

15. Comparison of Low Social Competence
Schizophrenics on Ratings made by Male
and Female As on the Question "Did Your
Patient Feel at Ease?" ....................... 69

16. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male
and Female As on the Question "Did The
Patient Cooperate?" .......................... 70

17. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male
and Female As on the Question "Did the
Patient Show Interest In Rating The Photos?".. 70

18. Comparison of Male As vs. Male Bs, and
Female As vs. Female Bs on Ratings made
by High Social Competence Schizophrenics
on the Question "Did You Like the Aid?" ...... 71

19. Comparison of Male As vs. Male Bs, and
Female As vs. Female Bs on Ratings made
by High Social Competence Schizophrenics
on the Question "Did You Feel at East
with the Aid?" ................................ 71


vii


Page











Table


20. Comparison of Male As vs. Male Bs, and
Female As vs. Female Bs on Ratings made
by High Social Competence Schizophrenics
on the Question "Was the Aid Friendly?" ....... 72

21. Comparison of Male As vs. Male Bs, and
Female As vs. Female Bs on Ratings made
by High Social Competence Schizophrenics
on the Question "Was the Aid Self-confident?".. 72

22. Comparison of Male As vs. Male Bs on
Ratings Made by High Social Competence
Schizophrenics on the Question "Was
the Aid a Strong, Dominant Person?" ........... 73

23. Comparison of Female As vs. Female Bs on
Ratings made by High Social Competence
Schizophrenics on the Question "Was the
Aid a Gentle, Kind Person?" .................... 73

24. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male
and Female Bs on the Question "Did you
Like This Patient?" ............................... 74

25. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male and
Female Bs on the Question "Did You Feel
At Ease With This Patient?" .................... 74

26. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male and
Female Bs on the Question "Did The Patient
Feel At Ease?" ................................. 75

27. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male and
Female Bs on the Question "Did The Patient
Cooperate?" ................................... 75

28. Comparison of Low Social Competence
Schizophrenics vs. High Social Competence
Schizophrenics on Ratings made by Male and
Female Bs on the Question "Did The Patient
Show Interest In Rating the Photos?" .......... 76


viii


Page












Abstract of dissertation presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



THE INTERACTION OF THE A-B STATUS
OF PSYCHIATRIC AIDS AND EFFECTIVE COMMUNICATION
OF EXPECTANCIES TO SCHIZOPHRENICS



By


Donald John Taylor


August, 1975




Chairman: Dr. Vernon Van De Riet
Major Department: Psychology


In this project, A and B type psychiatric aids were

studied in interaction with high and low social competence

schizophrenics. The effect of aid type x patient type

pairings on the communication of experimenter expectancies

from the aids to the patients was studied.

Sixteen aids and 124 patients from Florida Northeast

State Hospital were used as subjects. Aids were selected

on the basis of scores on a modified version of the Kemp

30-item A-B questionnaire. All aids participating in the

study were Caucasian. Of 17 male aids completing and

returning the questionnaire, four extreme scorers were

classified as As and four as Bs. Of 36 female aids









completing and returning the questionnaire, four extreme

scorers were classified as As and four as Bs. Patients

participating in the study were all male inpatients at the

hospital diagnosed upon last admission as schizophrenic.

High and low social competence status was assigned a

patient after a review of his chart and an interview. Five

social history variables taken from the schema of Zigler

and Phillips were used in rating level of social competence.

These were age, education, occupation, employment history,

and marital history.

The basic experimental procedure consisted of each aid

administering to each of eight or less patients the Rosen-

thal picture rating task, designed to study experimenter

expectancy effects. In this task the subject (patient) is

asked to rate on a scale ranging from -10 to +10 the degree

of success or failure the persons pictured in a series of

ten neutral photographs have been experiencing. Expectancies

were established by means of the instructions read by the

investigator to the aids, and by the investigator telling

each aid immediately prior to the aid's meeting each patient

that the patient was one of two basic personality types, and

would therefore average +5 or -5 on his photo ratings.

Dyads studied were male A type aids with low and high

social competence schizophrenics; female A type aids with

high and low social competence schizophrenics; male B type

aids with high and low social competence schizophrenics;

and female B type aids with high and low social competence

schizophrenics.









After the administration of the picture rating task

each aid and patient were separately asked to rate on five-

point Likert type scales their impressions of their partners

in the picture rating task.

Later five male and three female undergraduates rated

audiotapes of the aids reading the standardized instructions

to the picture rating task on 12 voice qualities.

Dependent variables were the expectancy scores, voice

quality ratings, and interpersonal impressions ratings from

each of the eight aid type x patient type dyads. It was

predicted that As of both sexes would be more effective

than their B counterparts in unintentionally influencing

low social competence schizophrenics toward established

expectancies, and that Bs of both sexes would be more

effective in influencing high social competence schizo-

phrenics. It was further predicted that ratings of voice

qualities and interpersonal impressions ratings would vary

in predictable ways according to the aid type x patient

type pairings.

Statistical analysis demonstrated that male Bs were

more effective than male As in unintentionally influencing

high social competence schizophrenics, and that female

Bs were also more successful with this group than were

female As. Neither male nor female As, however, were more

successful than their B counterparts in influencing low

social competence schizophrenics. Analysis of the data











gave little support to the voice quality and interpersonal

impressions hypotheses.

The findings of this study do not support previous

findings in related research utilizing many of the same

procedures. Possible explanations for this fact are

presented along with implications this study has for

future research.


xii















INTRODUCTION


The A-B Variable and the Interaction Hypothesis


One of the most interesting and potentially productive

areas of research in psychotherapy has been the study of

A and B therapist "types." The A-B variable has been exten-

sively studied in actual psychotherapy and in psychotherapy

analogue situations. The results of these studies, though

often complex and sometimes seemingly paradoxical, have made

it clear that the interaction of measurable therapist charac-

teristics with patient characteristics potently affects the

success and process of psychotherapy.

Whitehorn and Betz first distinguished A and B therapists

in 1954. Working at the Phipps Clinic at Johns Hopkins Hospi-

tal, they retrospectively examined the attributes of 14 of

the 35 psychiatric residents who worked at the clinic between

1944 and 1954, and who met the criterion of having treated

at least four neurotics and four schizophrenics. The seven

therapists with the highest success rates in treating schizo-

phrenics were designated As and the seven with the lowest

rates were designated Bs. As and Bs were equally successful

in treating neurotic patients. The 100 patients seen by these

therapists were hospitalized. They were mostly of middle and











upper-class socioeconomic status. At least half were female.

Improvement data, and data indicating that A and B therapists

were characterized by different therapeutic styles, came from

ward charts, analyses of nurse's notes, conference notes, and

discharge status reports (Whitehorn and Betz, 1954).

Until 1956, As and Bs were classified as such solely on

the criterion of their differential effectiveness with schizo-

phrenics. Then, in a study designed in part to explore the

question of whether contrasting personal qualities distinguish-

ing As from Bs could be demonstrated and characterized, Whitehorn

and Betz began to focus on the Strong Vocational Interest Blank

(SVIB) as a research tool, because it was well standardized, and

because it was not psychopathologically oriented, but focused

on human interests (Whitehorn and Betz, 1960).

The SVIB was administered to 35 therapists, some from

the original 1954 study, some from a 1956 study, and some from

unidentified sources. It was found that A therapists differed

from Bs on four scales of the SVIV. As scored high on the

Lawyer and CPA scales and low on the Printer and Mathematics-

Physical Science Teacher scales, whereas Bs showed the opposite

pattern on these four scales. Item analysis subsequently

identified 23 items on the SVIB which reliably differentiated

between therapist types (Betz, 1962; 1967).

Scores on some combination of the SVIB items or scales

then became the defining measure of A-B therapist status in












future studies of the A-B variable.

In 1962, McNair, Callahan, and Lorr found that with

nonschizophrenic male outpatients, B therapists were

significantly more successful than A therapists on several

outcome criteria. Subjects were 55 therapists from seven

Veterans Administration clinics and their 40 male outpatients.

Rating of patient change were derived from the Taylor Manifest

Anxiety Scale, Barron's Ego Strength Scale, a symptom check-

list, a self-satisfaction rating scale, therapist ratings

of severity of illness, Interview Relationship changes, and

an Interpersonal Changes and Symptom Reduction Scale.

The combination of the Whitehorn and Betz studies

(summarized in Betz, 1962; 1967) and the McNair et al. study

(1962) generated the interaction hypothesis, which has guided

subsequent research on the A-B variable. The hypothesis, as

originally formulated, is that As achieve greater success

with schizophrenic patients and Bs achieve greater success

with neurotics.

Significant differences between the Whitehorn and Betz

(1954) and the McNair et al. (1962) studies allow for other

interpretations of the results. (Chartier, 1971). The

process and outcome measures in the two studies differed,

and the possible effects of hospitalization were uncontrolled.

Also the patient samples differed in ways other than their

diagnostic categories. Most of the McNair et al. patients











were of lower or lower-middle class background, and all

were males, whereas the Whitehorn and Betz patients were

mostly middle and upper class, and at least half were

female. Furthermore, Whitehorn and Betz found no differ-

ences in the success rates of As and Bs in the treatment

of neurotics.

In spite of this questionable foundation for the

interaction hypothesis, it has generated a great deal of

research which has found the interaction of A-B status and

patient "type" to operate in a complex and potent way as

a determiner of behavior in a variety of contexts which

have implications for fundamental psychotherapeutic proces-

ses, other interpersonal processes, and patient management.

Most studies of the A-B phenomena can be classified

as one of four types: 1) clinical studies of psychothera-

pists in interaction with patients; 2) psychotherapy

analogue studies utilizing A and B quasi-therapists inter-

acting with patients, pseudopatients, or material represen-

tative of patient symptomotology, in situations assumed to

be analogous to psychotherapy; 3) studies of attitudes of

A and B therapists and quasi-therapists toward different

patient "types"; and 4) studies of personological differences

between As and Bs. Many studies are relevant to more than

one of these areas of inquiry.

The research which this dissertation attempts to replicate







S5


and extend (Trattner and Howard, 1970), opened a new line

of inquiry into the A-B phenomena: the differential ability

of A and B psychiatric aids to unintentionally influence

relatively high and low social competence schizophrenic

inpatients (Zigler and Phillips, 1960), as measured by

the Rosenthal (1966) picture sorting task, a standardized

procedure originally designed to measure the effects of

experimenter bias on experiment outcome.

The major studies of the A-B phenomena are reviewed

in the following sections.



Later Clinical Outcome Studies


In 1956 Betz and Whitehorn reported a study of the

A-B phenomenon with 109 schizophrenic patients not included

in the 1954 study. They did not report how many, if any,

therapists in this study were participants in the 1954

study. The patients were divided into a group of 45 who

received psychotherapy only, and a group of 64 who received

both insulin shock therapy and psychotherapy. For patients

receiving therapy only, As showed an 82% improvement rate

(based on a three-part rating of discharged patients by

three psychiatrists, including the patient's psychiatrist)

as compared to the Bs' 34% improvement rate. The same

differences between As and Bs in therapeutic approach

noted in the 1954 (Whitehorn and Betz) study were again











statistically significant.

With the 64 patients who received both insulin shock

therapy and psychotherapy, As and Bs showed identical 82%

improvement rates. Furthermore, patients treated by Bs

received "active personal participation," a therapeutic

style previously found to be characteristic only of As

in dealing with schizophrenics.

Betz pointed out a further important A-B distinction

in 1963, upon reexamining all of the 93 patients at the

Phipps Clinic from 1948-1954 who stayed for at least a

month and received no shock therapy. She found differen-

tial success rates of As and Bs even more striking when

schizophrenics were divided into process and nonprocess,

or reactive types. Although process patients are

considered to have poorer prognoses than nonprocess

patients (Bleuler, 1950), As had 71% improvement rates

with process schizophrenics while Bs had 18% improvement

rates with this population. With nonprocess patients, As

had a 68% improvement rate, Bs, a 44% rate.

These findings led Carson (1967) to infer that As

are superior to Bs in working with more inaccessible,

severely disturbed patients, but that this superiority

largely disappears when patients are less inaccessible and

disturbed by virtue of shock treatment or by virtue of the

reactive nature of their disorder.











Setphens and Astrup (1965) examined 4-14 year follow-

up data for 236 patients for whom data was available, who

were hospitalized at the Phipps Clinic between 1950 and

1960. Examining the effects of insulin treatment, patients'

progress vs. reactive status, and therapist A-B type, they

found almost no effect of A-B therapist classification.

Their indices of follow-up status--letters, telephone

conversations, and personal contacts with patients and

their relatives--were inprecise, however, and no mention

was made of how consistently available each of these

sources was, or how each was weighted.

Betz (1967) argued against the validity of the

Stephens and Astrup study by pointing out that the unim-

proved category of patients became too small (about 30%)

between 1955 and 1960 to be statistically comparable to

the 50% unimproved group before 1955, and that between

1955 and 1960 ataractic drugs were utilized in patient

care, making Stephens and Astrup's sample unlike the

psychotherapy only sample of the early Whitehorn and Betz

studies (Betz and Whitehorn, 1956; Whitehorn and Betz,

1954).

Chartier's (1971) reasoning that with the introduc-

tion of ataractic drugs in the period following the original

Whitehorn and Betz studies, the relevance of the A-B distinc-

tion may have been lost, is important to note in evaluating

Betz's argument. Noting that present average annual










improvement rates for hospitalized schizophrenics have

exceeded 70%, a figure comparable to the improvement rate

attained by As in the psychotherapy only condition of the

1954 Whitehorn and Betz study, Chartier speculates that

psychotropic drugs, by virtue of their symptom-reducing

effect, may operate in such a way as to render insignificant

the special qualities of a given therapist, resulting in

high improvement rates regardless of A-B status. The

finding that with patients presumably more accessible to

therapy by virtue of having received insulin shock, As and

Bs have equivalent success rates (Betz and Whitehorn, 1956),

lends support to this reasoning.

Three recent clinical studies have produced data which

does not support the interaction hypothesis. Bendar and

Mobley (1969) selected 13 As and 11 Bs from the 141 experi-

enced therapists who agreed to participate in their study,

and who were currently treating a client with a schizo-

phrenic or neurotic diagnosis. Outcome measures included

MMPI scores; therapist, patient, and psychometrician ratings

of current adjustment; Spitzer Psychiatric Status Schedule

ratings of current distress, behavioral disturbance, impulse

control, reality testing, and total adjustment; and a Q-sort.

Only the Q-sort yielded data confirming the interaction

hypothesis. A further finding was that As were more effec-

tive than Bs in relieving subjective distress, especially

with schizophrenics, and in improving general adjustment,











and that Bs were superior in facilitating impulse control,

especially with schizophrenics. This study is of question-

able value however, because the authors used a version of

the A-B scale (Campbell, Stephens, Uhlenhuth, and Johansson,

1968) which correlates only .40 with other commonly used

versions of the scale (Kemp and Stephens, 1971).

Using therapist's post-treatment evaluations and length

of hospitalization as outcome measures, Beutler, Johnson,

Neville, and Workman (1972) found no relationship of A-B

therapist status and patient schizophrenic or neurotic status

with outcome of therapy of 31 male and female patients receiv-

ing treatment at a private university affiliated psychiatric

hospital. Beutler, Johnson, Neville, Workman, and Elkins

(1973) measured outcome of therapy with 49 male and female

schizophrenic and neurotic patients at the same institution

by means or therapist improvement ratings. Again, no

significant interaction effect was found.

In contrast, two recent clinical studies have supported

the interaction hypothesis. In one, Berzins, Ross, and

Friedman (1972) measured outcome of brief therapy in a

college clinic by means of therapist evaluation of outcome,

and patient evaluation or outcome and experienced rapport

with their therapist. Results supported the interaction

hypothesis, particularly with regard to the therapists'

appraisal of their own effectiveness. B therapists also

obtained much lower improvement and rapport ratings from











schizoid than neurotic patients, whereas As received

equally high ratings with both diagnostic categories of

patients.

In another clinical study supportive of the inter-

action hypothesis, 36 enuretic children were treated by

their mothers, who served as behavior technicians follow-

ing a one-hour training program in Mowrer's conditioning

treatment for enuresis. B type mothers were significantly

more successful in treating these neurotic symptoms than

were A type mothers. (James and Foreman, 1973)

These studies taken together do not provide a sound

basis for accepting or rejecting the interaction hypothesis.

They are often contradictory. Generalization is made

difficult by lack of uniform procedures for identifying

As and Bs, lack of uniform criteria for evaluating out-

come, and large variances in therapy settings and diagnos-

tic procedures.



Clinical Studies of A-B Therapist Behavior


Several clinical studies have attempted to determine

how, if at all, A and B therapists and their patients differ

in their behavior in therapy. These studies are rather

consistent in demonstrating that As and Bs behave differ-

ently with schizophrenic and schizoid patients and neurotic

patients in psychotherapy.

Whitehorn and Betz (1954), examining case records,











found that As and Bs differed in their treatment of patients

in several respects, four of which were significantly related

to improvement. These categories were: the type of rela-

tionship formed by the therapist, the type of personal

formulation of the patient's problems, the types of strate-

gic goals selected, and the type of tactical pattern used.

As were better able to gain the confidence of their

patients, and tended to better understand the motivation

and meaning of their patient's behavior. They more often

formulated goals in terms of having the patient develop a

better understanding of his assets for constructive conflict

resolution and developing a strong, meaningful therapeutic

relationship. These were characterized as "personality-

oriented goals" in contrast to the "psychopathologically-

oriented goals" of B therapists: supervised living, symptom

decrease, socialization, and insight into pathology. As

were characteristically actively, personally involved with

their patients while Bs adopted passive, interpretive, and/or

instructional therapeutic stances. These characterizations

were derived by a retrospective examination of case records.

In 1957, Betz again stressed the importance of active

personal participation on the part of the therapists studied,

noting that of the 71 patients in both the 1954 (Whitehorn

and Betz) and 1956 (Betz and Whitehorn) studies receiving

this kind of therapeutic behavior, 89% were rated as improved.

In the condition of B therapist x insulin plus psychotherapy,










B therapists were rated as giving active personal participa-

tion in the therapy process.

McNair et al. (1962) found that As working with out-

patient neurotics were like Bs working with schizophrenics

in the Whitehorn and Betz studies in that they significantly

more often stated goals in concrete-descriptive terms, and

emphasized insight as goals. The combination of these

studies seems to indicate that As and Bs change their

therapy behavior when dealing with differing categories

of patients.

In a study by Segal (1970), therapy scripts were content

analyzed to obtain therapist attitudinal and behavioral oral

ratings. Subjects were twenty male clinical psychology

students who were seeing patients in therapy as part of

their training. Each of the therapists completed the

Symptomatic Behavior Inventory Rating Scale for each client

he was seeing in psychotherapy. This inventory was derived

from the Phillips and Rabinovitch (1958) classification

schema designed to distinguish between neurotic, or turning-

against-self (TAS) and schizoid, or avoidance-of-other (AOS)

behavior. For each therapist, two therapeutic interactions

with one turning-against-self client were recorded. Results

showed that with this neurotic group of patients, Bs made

fewer negative comments, were more facilitative and encourag-

ing of self-exploration, and placed less emphasis on having

patients respond to specific questions or ideas. As were











more directive and interpretive.

A series of studies (Scott and Kemp, 1971; Beutler

et al., 1972; Beutler et al. 1973) attempted to relate

the A-B status of therapists to differential quality of

therapist-offered conditions (Truax and Carkhuff, 1967)

manifested in therapy with schizophrenic and neurotic

patients.

In the Scott and Kemp study of initial interviews

conducted by 25 senior medical students with an equal

number of neurotic outpatients, there were no significant

relationships between therapist A-B status and their rated

empathy, warmth, and genuineness. As was predicted, how-

ever, B therapists elicited greater depth of self-exploration

from their patients.

In the Beutler et al. 1972 study, fourteen neurotic and

four schizophrenic patients were seen by A therapists, and

four neurotic and nine schizophrenic patients were seen by

B therapists. Pairings were rated for therapist empathy.

Results showed As with schizophrenic and Bs with neurotic

patients displayed more empathy.

In the Beutler et al. 1973 study, initial interviews

of six experienced psychiatrists with 29 neurotic and 20

schizophrenic inpatients were transcribed. Data indicated

that A therapists were more empathetic with schizophrenic

than with neurotic patients, whereas for B therapists, no

therapist-offered conditions significantly differed with

patient type.












Analoque Studies of A-1 Pvchiotherapy Behavior


Several analogue studies, rviev;ed below, have attempted

to verify the interaction hypothesis and clarify the nature

of the interaction of A and B therapists with schizoid or

schizophrenic and neurotic patients. With one exception

(Berzins, Ross, and Cohen, 1970), these studies have used

male college students as quasi-therapists.

In one of the few analogue studies not supporting, at

least in part, the interaction hypothesis, Seelig (1969) had

male undergraduates listen to tape recorded segments taken

from interviews with normals, neurotics, and schizophrenics,

and respond therapeuticallyy" at pre-selected intervals.

No therapist type main effects or therapist x patient inter-

action effects reached statistical significance. The hypo-

thesis of the study was that As with schizophrenics and Bs

with neurotics would enter into "active personal participa-

tion" (operationalized along three dimensions) more often

than in the opposite pairings.

Vaughn (1969) used graduate students in clinical

psychology who were extreme scorers on the A-B continuum as

subject-therapists. They listened to tape recorded therapy

interviews of patients classified as either TAS types or

AOS types (Phillips and Rabinovitch, 1958), and responded

as if they were interacting with the patients in therapy.

As responding to AOS patients and Bs with TAS patients, as







15


predicted by the interaction hypothesis, were warmer and

more sympathetic than when the pairings were reversed.

In another study of this type A and B volunteer under-

graduate males responded "helpfully" and "naturally" at

four points of interruption in videotaped enactments of

TAS and AOS patient prototypes. Ratings of responses

indicated that as predicted by the interaction hypothesis,

communication of respect and empathetic understanding as

well as response duration of subject-therapists was a joint

function of A-B status and patient type. (Seidman, 1971)

Berzins and Seidman (1969) had 72 male undergraduates

write self-chosen "helpful" responses to tape recorded,

enacted patient communications intended to represent

schizoid (AOS) and neurotic (TAS) communications. Subject-

therapists were divided into As, Bs, and ABs (middle-range

scorers on the A-B scale). Relative to the opposite pair-

ings, when As were paired with the schizoid and Bs with the

neurotic patient enactments, they gave longer responses,

emitted more responses of a declarative as opposed to

questioning form, and showed more positive and fewer nega-

tive Bales social-emotional reactions. ABs did not react

differentially to the two patient types.

Dublin and Berzins (1972) found A and B males, instruc-

ted to respond verbally and therapeutically to videotaped

enactments, more direct and intense in their communications

in response to schizoid and neurotic enactments, respectively.










Berzins, Seidman, and Welch (1970) had male under-

graduate As, ABs, and Bs write helpful responses to enacted

tape recorded communications purportedly of clinic patients.

Instead of AOS-TAS enactments, enactments were intended to

represent one patient who handled anger in an extrapunitive

(EXT) way and one who handled anger in an intropunitive

(INT) way. As paired with the EXT and Bs with the INT

enactments were more satisfied with their helping performances

than were Bs oppositely paired. Several other measures of

quality of response and post-experimental self-ratings of

subjective reactions to the enactments yielded only limited

support for the interaction hypothesis.

Carson, Harden, and Shows (1964) constructed a quasi-

therapeutic situation in which A and B college students

interviewed other students. Instructions to the inter-

viewees attempted to establish interviewee sets simulating

attitudes and behaviors of schizophrenics and of neurotics--

that is, distrust-hostility-expectancy of harm (DHH) and trust-

friendliness-expectancy of help (TFH), respectively. In tape

recorded sessions, the interviewers obtained as much informa-

tion as possible in 18 different content areas and these

interviews were subsequently rated regarding the adequacy of

information obtained. The expected interaction between

interviewer type and interviewee set was not significant for

adequacy of information, but the interaction was significant

when the tapes were rated in terms of the number of content











areas explored by the interviewers.

Jacob and Levine (1968) used the same procedure in a

study of college students, with one major modification.

Instructions to interviewees for the purpose of inducing

a DIIH or TFG set were tape recorded rather than given

personally, in order to control for covert effects of

experimenter bias. Results did not support the findings

of Carson et al. (1964).

In the only analogue study of this type in which real

patients were utilized, psychiatric aids conducted brief

interviews with hospitalized narcotics addicts. Prior

to being interviewed by an aid, patients were briefed by

an assistant who attempted to induce a DIIH or TFH expec-

tancy. As predicted, A type aids with DHH patients and B

type aids with TFH patients obtained better patient self-

disclosure in personal topical areas. (Berzins, Ross, and

Cohen, 1970)



Studies of A-B Therapist Attitudes


Although the clinical studies of therapy outcome have

been equivocal in their findings, the clinical therapy

process studies and analogue therapy studies outlined above

have in general been supportive of the hypothesis that As

and Bs behave differently and are differentially effective

in therapy and interview situations with neurotic and











schizophrenic or schizoid patients. Several studies have

attempted to discover attitudinal variables and subjective

reactions to patients manifested by A and B therapists

and quasi-therapists which may in part account for this

phenomenon. Another series of studies, outlined in the

next section, have had as their aim the delineation of

stable personality variables characteristic of As and Bs

that may account for their varying behavior and success

with different types of patients.

Kemp (1966) had male college students choose thera-

peutic responses at predetermined intervals in recorded

interviews of actors playing the roles of AOS and TAS

patients. As with AOS "patients" and Bs with TAS "patients"

reported more discomfort and difficulty in selecting help-

ful responses. This finding, supported by additional

research (Carson and Klein, 1965), raised the possibility

that As and Bs were most successful with patients who

produced the greatest subjective discomfort in them. The

label "paradoxical discomfort" was given to this hypothesis.

In another investigation supportive of the paradoxical

discomfort hypothesis, medical students evaluated bogus

case histories written to represent a schizoid and an intro-

punitive-neurotic patient. (Kemp and Sherman, 1965) Compatible

pairings (As with schizoids and Bs with neurotics) were charac-

terized, as measured by a 22-item questionnaire, by less interest

on the part of the medical students in treating the patient,











less confidence in outcome, feelings that the patient was

less like an ideal patient, and anticipation of more diffi-

culty in determining etiology.

Kemp and Carson (1967) replicated the procedure of the

Kemp and Sherman (1965) study, but used psychiatric residents

as Ss. They found As, in rating the schizoid patient, and

Bs, in rating the neurotic patient, expected that they

would be more uncomfortable in treatment with the patient.

The data failed to replicate the findings of Kemp and Sherman,

however, and the authors hypothesized that paradoxical dis-

comfort may be partically offset by the effects of professional

training and experience.

Other researchers have reported findings that belie

the notion of a paradoxical discomfort phenomenon. Berzins

and Seidman (1969), although using a methodology and a

subject population very similar to Kemp (1966) reported

opposite results. As and Bs reported more satisfaction with

their helpful responses when in compatible pairings--As with

AOS and Bs with TAS enactments. Berzins, Seidman, and Welch

(1970) also reported students felt more satisfaction with

their helpful responses when in compatible pairings--As with

EXT enactments, and Bs with INT enactments. In their study

of 32 practicing psychotherapists, Bendar and Mobley (1969)

found no evidence that A and B therapists as categorized by

the Campbell et al. (1968) A-B scale differentially perceive

or prefer neurotic or schizophrenic patients.








20


Anzel (1970), and Stein, Green and Stone (1972),

included patient socio-economic status as a variable in

their studies of A and B therapist's attitudes toward

patients.

Anzel varied patient characteristics presented to 48

male experienced therapists and 80 male undergraduates on

the dimensions of type of pathology (AOS-TAS), severity of

pathology (severe-mild), and social class (upper-lower).

Patient characteristics were presented by means of excerpts

of recordings from contrived therapy interviews and by

bogus written case histories of the patients. As hypothe-

sized, A therapists responded more favorably than Bs to

AOS, severely disturbed, and upper-class patients; and B

therapists responded more favorably to TAS, mildly disturbed,

and lower-class patients. Student As and Bs were not

consistently similar or dissimilar to their therapist

counterparts.

Stein, Green, and Stone (1972) had 37 psychiatric

residents rate patient vignettes on likeability, discomfort,

interest in treating, and prognosis. Patient social class

as well as diagnosis was systematically controlled. Bs

assigned significantly higher likeability ratings to

neurotics than to schizophrenic patients while As did not

differentiate between the two diagnostic types. There

were no patient type x therapist type interactions on the

other variables. Both As and Bs in general held more








21


favorable attitudes toward neurotic and middle-class

patients than they did toward schizophrenic and lower-

class patients. This finding is consonant with previous

research which has shown a therapist bias favorable to

neurotic, non-lower-class patients. (Hollingshead and

Redlich, 1958; Kaplan, Kurtz, and Clements, 1968; Kurtz

and Kurtz, 1968)



Personality Characteristics of As and Bs


The most obvious personality differences of As and Bs

are of course their differential vocational interest patterns

as shown on the SVIB. As score high on the Lawyer and CPA

scales and low on the Printer and Mathematics-Physical

Science Teacher scales, whereas Bs show the opposite pattern

on these four scales. Carson (1967) reported a factor

analysis of SVIB findings and their data on therapeutic

style and outcome, Whitehorn and Betz (1960) drew a composite

picture of the personal characteristics of A and B therapists.

They hypothesized that As are like lawyers in that they are

problem-solvers, managing to find patients' resources and

encouraging acceptable modes of behavior. They hypothesized

that As are not regulative or coercive but instead reward

free self-determination in their patients, in contrast to

Bs, who emphasize conformity, deference, and mechanical

approaches to therapy.

These hypotheses are of course speculative. Later











personality studies of As and Bs are more solidly based,

resting upon evidence gathered from the administration of

tests of cognitive style and other personality characteris-

tics to samples of As and Bs.

Pollack and Kiev (1963), using the same male staff

members of the Phipps Clinic as did Whitehorn and Betz

in their original studies, found that in spatial orienta-

tion as measured by the Witkin Rod and Frame Task, Bs

demonstrated significantly more field independence than As.

Shows and Carson (1966) replicated this finding with college

students. Their subjects did not significantly differ,

however, with regard to scores on the Witkin Embedded

Figures Test of field dependence-independence.

Pollack and Kiev (1963) felt that their finding that

A therapists were more field dependent was consistent with

the flexible, reciprocal, more personal nature of the thera-

peutic relationship characteristic of As with schizophrenics,

and that the field independence of Bs was consistent with

the view that these therapists are more precise and detached

in interpersonal situations.

Silverman (1967) reviewed the A-B literature and the

research in field dependence-independence and offered a

speculative composite description of A and B male therapists.

Briefly, As were seen as more responsive to attributes of

the perceptual field and as a consequence more capable of

relaxing their reality orientation and responding to hunches










and intuition. Silverman theorized that As are therefore

more empathetic with schizophrenics than are Bs, who share

the practical, goal-directed point of view typical of

neurotics, and are in general more empirically than intui-

tively oriented.

Portnoy and Resnick (1972) found A type male under-

graduates conditioned significantly better in a verbal

conditioning task than did Bs. They interpreted their

results as consistent with the Pollack and Kiev (1963) and

Shows and Carson (1966) findings, reasoning that subjects

in a verbal conditioning task who are more sensitized

and dependent on external cues for their behavior would

condition better than subjects who are less field dependent.

Fancher, McMillan, and Buchman (1972), studying under-

graduate males, found evidence with a test of accuracy in

person perception and a role taking test of accuracy of

empathy that As tend to be more intuitive, empathetic,

and more in touch with emotional processes than Bs, whereas

Bs tend to be more rational and intellectual in their approach

to interpersonal situations.

Berzins, Seidman, and Welch (1970) advanced the hypothe-

sis that therapist-patient complementarity on dimensions

related to modes of reacting to stress may in part account

for the differential success of As and Bs with neurotic and

schizoid patients. In doing so they hypothesized a TAS

reaction to stress as possibly characteristic of As. In











support of this idea they cite studies which have shown

that As appear to display trusting-intropunitive responses

when under stress (Sandler, 1965); display depressive

symptomatology as clinic patients, including internaliza-

tion of anger and suicidal ideation (Berzins, Friedman,

and Seidman, 1969); and respond to stimuli of the Rosenz-

weig Picture Frustration Study with significantly fewer

extrapunitive responses than ABs or Bs (Welch and Berzins,

1968).

Lorr and McNair (1966) suggested that since A-B scale

items correlate significantly with the masculinity-femininity

scale of the SVIB, sex-role identification might underlie

A-B differences, with As being less masculine than Bs.

This hypothesis received strong support from a study by

Dublin, Elton, and Berzins (1969).

Dublin et al. gave the Omnibus Personality Inventory

(OPI), the American College Test, and an A-B scale to

over 1200 male and almost as many female freshman college

students. Ninety-eight A and ninety-nine B males, along

with ninety-eight female As and seventy-three Bs were

selected for the study. These groups consisted of persons

scoring + one standard deviation from the mean on the A-B

scale.

For both sexes Bs and ABs (persons randomly drawn from

middle-range scorers on the A-B scale) showed progressively

higher scores on masculinity than did As. Additionally,










the OPI scale which best discriminated the female groups

was Theoretical Orientation. This scale suggests masculine

pursuits in that it is a cognitive style scale which measures

interest in science and scientific activity, and a logical,

rational, critical approach to problems. Male Bs, consis-

tent with their higher masculinity scores, also scored

higher on natural science abilities while male As outscored

Bs on verbal abilities. For both males and females, Bs

scored significantly higher on the Cognitive Complexity

scale of the OPI, which the authors describe as a scale

measuring the ability to "withstand and persevere in the

face of new, novel, uncertain, or ambiguous events." (Dublin

et al., 1969)

Berzins, Barnes, Cohen, and Ross (1971), and Berzins,

Dove, and Ross (1972), gave the Personality Research Form

(Jackson, 1967) to large samples of As and Bs. The 1971

study of 233 male undergraduates and 50 male professionals

(psychologists, psychiatrists, counselors, and social workers)

found the B pole of the A-B dimension was related to social

ascendency and openness to complex experiences, whereas the

A pole was related to caution, social ineptness, and a

restricted cognitive scope. The authors observed a strong

masculinity-femininity component in the variables differen-

tiating As and Bs, with Bs at the masculine pole.

The Berzins et al. 1972 study involved four samples

of As and Bs: 94 male professionals, 661 male undergraduates,











114 male college clinic patients, and 720 female under-

graduates. Although these groups differed at least with

regard to vocational commitment, training, age, sex, and

degree of adjustment, As and Bs, defined for each sample

as scorers in the outer quartiles of the A-B distribution

for that sample, were found to be extreme and opposite

scorers on five Personality Research Form scales: Harm-

avoidance, Dominance, Change, Sentience, and Succorance.

Briefly, A type Ss in each sample could be described as

relatively cautious, submissive, uninclined to seek

variety or sensual pleasure for its own sake, and as some-

what succorant. The B type Ss on the other hand showed a

risk-taking, dominant, variety-seeking and counterdependent

orientation.

This study is an affirmation of the hypothesis under-

lying the many analogue studies cited above in which A and

B undergraduates were assumed to be personologically similar

to A and B therapists. It is also important in combination

with the Dublin et al. (1969)research in indicating that

female As and Bs vary along the same personological dimen-

sions in the same directions as do male As and Bs, who have

been the subjects of nearly all research involving the inter-

action hypothesis.







27


The A-B Variable and Unintended Interpersonal Influence


In studies utilizing the methodology originated by

Rosenthal (1966) to examine the effects of experimenter

expectancies on the outcome of psychological research,

Trattner and Howard (1970), and Persinger, Knutson, and

Rosenthal (1968), found that A and B psychiatric aids are

differentially effective in unintentionally influencing

psychiatric patients of differing diagnostic categories.

In the prototypical Rosenthal experiment, Es present

to each of their Ss a series of 10 standardized photographs

of male and female faces. The Ss judge the degree of

success or failure the person pictured seems to be experi-

encing. Plus scores, up to +10, are assigned by the S if

he judges the person pictured to be successful. Minus

scores up to -10 are given if he judges the person pictured

to be failing. When a third party instills an expectancy

in E as to the predicted performance of S, E tends to elicit

that performanre--that is, he tends to "bias" the responses

of S to a task that statistically has been shown to normally

elicit neutral responses. The experimenter may be told,

for example, that S is of a personality type that ordinarily

averages scores of +5 on the task, or that he is of a type

that ordinarily averages -5, whereas in fact an average

score of O is the normal neutral response to the task when

experimenter bias is not instilled. (Rosenthal, 1966)











Trattner and Howard (1970) chose two As and four Bs

from 28 of the 118 male attendants at Boston State Hos-

pital who returned a Kemp 31-item A-B questionnaire. (Kemp

and Stephens, 1971) Fifty male schizophrenics were randomly

assigned to the attendants. The attendants administered

the Rosenthal picture rating task to the patients. Just

prior to the testing of each patient, attendants (Es)

were told that the patient (S) was either of a personality

type that averaged +5 or of a type that averaged -5 in

rating the success of the persons pictured. The patients

were classified as belonging to a relatively high social

competence group or a low social competence group on the

basis of the Zigler and Phillips (1960) classification

system. As predicted, A attendants biased low social

competence patients in both + and directions signifi-

cantly more than did Bs, while Bs biased high social

competence patients more than did As.

Social competence scores are derived from case

history data about the patient's age, intelligence, educa-

tion, occupation, employment history, and marital history.

(Zigler and Phillips, 1960) The concept of level of social

competence is similar to the concept of degree of severity

of psychological disturbance. TAS behavior is more typical

of relatively higher levels of social competence, and

AOS behavior is more typical of relatively lower levels

of social competence. With regard to hospitalized








29


schizophrenics, level of social competence is very similar

to the concept of premorbid adjustment. Lower levels of

social competence reflect poorer quality of premorbid

adjustment, whereas high social competence scores reflect

better premorbid adjustment and a greater likelihood that

the disorder is reactive in nature.

In a finding supportive of the Trattner and Howard

(1970) findings, Jenkins (1966), using college experi-

menters and subjects, found that B type Es, as predicted,

showed greater effects of their expectancies than did A

type Es. This is supportive of Trattner and Howard's

study in that college student Ss are closer to neurotic

or high social competence status than to schizophrenic

or low social competence status.

In the Persinger et al. research (1968), a variety of

patients served as subjects while both male and female

ward personnel served as experimenters. Contrary to prior

A-B research, patients were not selected on the basis of

severity of disturbance, but on the basis of primary cate-

gorization as relatively more hostile than anxious (paranoid

and character disorder) or relatively more anxious than

hostile (schizophrenic and neurotic). Results showed that

greater expectancy effects were exerted by A type male Es

and B type female Es with patients categorized as relatively

more anxious. With patients categorized as relatively more











hostile, B type male Es and A type female Es showed greater

unintended influence.



Extending the A-B Variable to Females


The Persinger et al. study (1968) is unusual in that

it is one of the few studies in which female As and Bs

were subjects in research on the A-B interaction phenomena.

Due to the nature of the A-B scale, which on its face

measures relative interest in mechanical, engineering, and

manual activities (Carson, 1967), nearly every one of the

A-B therapy or therapy analogue studies has used only males

as therapists or quasi-therapists. Two exceptions in addi-

tion to the Persinger et al. study are Berzins, Ross, and

Friedman (1972), in which two of the six therapists studied

in a college clinic were female; and James and Foreman (1973),

who studied the relative effectiveness of A and B mothers in

utilizing a conditioning technique to treat enuresis in their

children. Both studies supported the A-B schizophrenic-

neurotic interaction hypothesis.

Additionally, research by Berzins, Dove and Ross (1972)

and Dublin, Elton, and Berzins (1969), both extensive studies

of personality traits of As and Bs, found that A and B female

college students differ from each other in personalogical

characteristics as measured by the Personality Research

Form and the Omnibus Personality Inventory in the same ways










as do male college students, male outpatients, and male

professional psychotherapists. These findings imply that

it will be fruitful in future A-B studies to include thera-

pist or psychiatric aid sex as an independent variable.

Implications of the Interaction of A-B Status

and Interpersonal Influence with Hospitalized Patients

A strong case has been made that traditional verbal

psychotherapy is a method of persuasion--the therapist

does not use force to get his way, but uses his verbal

interpersonal skills as tools in trying to effect change

in his client's behavior. (See London, 1969, for a discus-

sion of verbal psychotherapy as a method of behavior control

characterized by subtle persuasion.) For this reason the

findings of Trattner and Howard (1970) and of Persinger et

al. (1968) are significant to the study of psychotherapy,

for they imply that As and Bs are differentially effective

in influencing differing types of patients to conform to

their expectancies. Separate studies dealing with indi-

vidual and group therapy with client populations of neurotics,

psychotics, juvenile delinquents, and the physically and

mentally disabled, have all converged in demonstrating that

therapists high in persuasive potency produce better client

personality and behavioral improvement than equally trained

therapists who are lower in persuasive potency or personal

social influence value (Truax, Fine, Moravec, and Millis,

1968; Truax, 1969).











\Apparently t-le persuasive potency of As and Bs fluctu-

ate according to the kind of patient with whom they are

interacting. Of course the Trattner and Howard and Per-

singer et al. studies dealt with psychiatric aids rather

than actual psychotherapist, but findings indicate that

personality correlates of A or B status are constant across

a wide range of age, sex, level of adjustment, and profes-

sional training. (Berzins, Dove, and Ross, 1972; Fancher,

McMillan, and Buchman, 1972; Dublin, Elton, and Berzins,

1969; Shows and Carson, 1966; Pollack and Kiev, 1963)

Perhaps more important than the implications which

these studies hold for the understanding of individual

psychotherapy, are their implications with regard to psychi-

atric hospital milieu therapy and patient management. Lack-

ing the resources for, and uncertain as to the relative

value of formal one-to-one psychotherapy, especially with

schizophrenics, most psychiatric hospitals today rely on

ataractic drugs and milieu therapy as primary modes of

treatment. Nurses, aids, and other ward personnel spend a

great deal of time with patients. Their attitudes toward

the patients, their ability to make them feel at ease,

and their ability to effectively talk to them and to thera-

peutically guide their behavior, are all important to

effective milieu therapy.

It seems practical then, if findings such as those of

Trattner and Howard (1970) can be replicated and perhaps










extended, to match wherever possible aids and other ward

personnel with patients with whom they are most effective.

Any one of the A-B scales now in use takes less than five

minutes to fill out, and is essentially of a non-threatening

nature, so there is little practical reason not to adminis-

ter it to incoming personnel. If, as in an effective

hospital it must be, the generalized expectation is that

patients will improve, or that particular programs will

prove beneficial to a patient, the persons in intimate

contact with the patient in his everyday hospital activities

should be able to effectively communicate that expectancy.



Purposes of the Dissertation


There were four main research goals for this dissertation:

1. The first purpose was to determine if the findings of

Trattner and Howard (1970) are replicable; i.e., if by means

of the Rosenthal picture rating task, A type male aids influ-

ence low social competence (S-C) male hospitalized schizo-

phrenics to a significantly greater degree than do male Bs,

and B type male aids influence high S-C male schizophrenics

to a significantly greater degree than do male As.

2. The second aim of the study was to determine how the

interaction of A and B type with high and low S-C schizo-

phrenics is effected by the sex of the aids. Since so many

psychotherapists and hospital personnel are female, and since

the Persinger et al. study (1968) indicates that A and B type










females interact with patients on the picture rating task

in different ways than do males, this is an important

area to explore.

3. The third aim of the study was to contribute to an

understanding of how vocal qualities of the aids in reading

the standardized instructions of the Rosenthal picture

rating task are related to the various pairings of patients

and aids. Trattner and Howard (1970) found that vocal

qualities of Es are effected by the pairings and that this

phenomenon seems to be related to effectiveness in biasing

patients toward the expectancies induced in the aids (Es).

Rosenthal (1966) reports that, contrary to hypotheses

that E's reinforcement of expected responses to the picture

rating task mediate the communication of expectancy effects,

data indicate that the very brief period in which the E

greets, seats, and instructs the S is when expectancy

effects are generated. Initial impressions that the E

and S form of each other, and the way in which E reads the

standardized instructions to the task, seem to be crucial.

Trattner and Howard (1970) had 13 Harvard College

males rate tapes of Es reading the standardized instruc-

tions to patient-subjects on nine qualities: discomfort,

coldness-distance, sophistication, self-confidence, dominance,

professionalness, masculinity, warmth-friendliness, and

awareness of the other. (The last variable was later dropped

from the analysis of the data.) Judges were not given











definitions of any of these vocal qualities. There were

no main effects on any of the qualities between A and B Es.

However there were significant patient x experimenter inter-

actions on all the rated voice qualities. As with low S-C

schizophrenics and Bs with high S-C schizophrenics were

rated higher on all qualities but discomfort and coldness,

on which they were rated significantly lower. Inspection

of the intercorrelations among the remaining variables

suggested that there was one major dimension being measured,

which the authors retrospectively labeled "social control."

In the dissertation research, vocal qualities of both

male and female Es reading the standardized instructions

were rated. Since it was thought probable that female Es

would communicate expectancy effects with different vocal

qualities than would male Es, the qualities of femininity,

softness, and kindness were rated along with the eight

qualities included in the Trattner and Howard experiment.

The voice quality of enthusiasm was also rated.

4. As was mentioned above, the interaction between Es and

Ss in the picture rating task is minimal: The Es greet

the patients, seat them, and read the instructions. The

patients then rate each picture, and the Es mark down their

answers. In addition, care was taken in the Trattner and

Howard study and in the present study to see that the aids

and patients had not met each other prior to the experiment.

These facts imply that the relative ability of each aid to











influence each type of patient is related to the initial

impressions each of the participants in the experimental

dyad form of their partners.

The fourth purpose of this dissertation, then, was

to assess, immediately after the picture rating task

was completed for each dyad, simple impressions that

the participants formed toward each other during the

brief (usually less than six minutes) time of their

interaction.

On five-point Likert-type rating scales, the patients

were asked to rate the degree to which they liked the

aid that gave them the "test," their feelings of com-

fortableness with the aid, the warmth-friendliness of

the aid, the dominance of the aid, the gentleness-kindness

of the aid, and the self-confidence of the aid.

Also on Likert-type rating scales, the aids were

asked to rate the degree to which they liked the patient,

their feelings of comfortableness with him, their impres-

sions of his warmth-friendliness, the degree to which the

patient was at ease in the interaction, his cooperativeness,

and his degree of involvement in the task.
















HYPOTHESES


There are three series of hypotheses. The first

series involves the interaction of aid (E) type with

patient (S) type in the degree of experimenter expectancy

effects manifested. The second series of hypotheses has

to do with the voice qualities that the four E types

manifest when reading the standardized instructions for

the picture rating task to each of the two S types.

The third series deals with impressions that the four

E types and two S types form of each other during the

administration of the picture rating task.



Experimenter Expectancy Hypotheses


Al. A type male Es will exert greater expectancy effects

with low S-C Ss than will B type male Es.

A2. A type female Es will exert greater expectancy effects

with low S-C Ss than will B type female Es.


Bl. B type male Es will exert greater expectancy effects

with high S-C Ss than will A type male Es.

B2. B type female Es will exert greater expectancy effects

with high S-C Ss than will A type female Es.












Voice (,ullity yp-oth lse


A3. Judges will rate the taped voices of both male and

female A type Es higher than their B type counterparts

in sophistication, self-confidence, profcssionalness,

warmth-friendliness, and enthusiasm, when reading instruc-

tions to low S-C Ss.

A4. Judges will rate the taped voices of both male and

female A type Es lower than their B type counterparts in

discomfort and coldness-distance when reading instructions

to low S-C Ss.

A5. Judges will rate A type male Es higher than B type male

Es in masculinity and dominance when reading instructions

to low S-C Ss.

A6. Judges will rate the taped voices of female A type

Es higher than female B type Es in femininity, kindness,

and softness when reading instructions to low S-C Ss.

A7. Judges will rate the taped voices of female A type Es

lower than female B type Es in dominance when reading

instructions to low S-C Ss.


B3. Judges will rate the taped voices of both male and

female B type Es higher than their A type counterparts in

sophistication, self-confidence, professionalness, warmth-

friendliness, and enthusiasm, when reading instructions to

high S-C Ss.

B4. Judges will rate the taped voices of both male and










female B type -1;; lowcr;: thlan their A type counterparts in

discomfort and col'n :s-diwstance when reading instruc-

tions to high S-C Ss.

B5. Judges will :rate i.] type male Es higher than A type

male Es in masculinity and dominance when reading instruc-

tions to high S-C Es.

B6. Judges will rate the taped voices of female B type

Es higher than female A type Es in femininity, kindness,

and softness when reading instructions to high S-C Ss.

B7. Judges will rate the taped voices of female B type

Es lower than female A type Es in dominance when reading

instructions to high S-C Ss.



Interrpersocnal Impressions Hypotheses


A8. Low S-C Ss will rate themselves as liking more, and

feeling more comfortable with both male and female A type

Es than with both male and female B type Es. They will also

rate the A type Es as more friendly and more self-confident.

A9. Low S-C Ss will rate male A type Es as more dominant

than B type male Es.

A10. Low S-C Ss will rate female A type Es as more gentle

and kind than female B type Ss.

All. Both male and female A type Es will rate themselves as

liking more, and beinq more comfortable with low S-C Ss.

They will also rate these patients as more at ease, coopera-

tive, and involved in the task.











B8. High S-C Ss will rate themselves as liking more and

feeling more comfortable with both male and female B type

Es than with both male and female A type Es. They will

also rate the B type Es as more friendly and self-confident.

B9. High S-C Ss will rate male B type Es as more dominant

than A type male Es.

B10. High S-C Ss will rate female type Es as more gentle

and kind than female B type Es.

Bll. Both male and female B type Es will rate themselves

as liking more, and being more comfortable with high S-C

Es. They will also rate these patients as more at ease,

cooperative, and involved in the task.















METHODOLOGY


Subjects


Aids: Thirty-item A-B questionnaires consisting of the

23-item Whitehorn and Betz scale (Whitehorn and Betz, 1960)

plus seven of the eight MIMPI items added by Kemp in his

version of the A-B scale (Kemp and Stephens, 1971) were

distributed to all available Florida Northeast State

Hospital aids working the day shift. Of 110 question-

naires given out, eighty-six were returned in usable form.

A total of fifty-three women and thirty-three men returned

the questionnaires. Questionnaires were scored for A

responses; i.e., high scores indicate a high number of A

responses. Total female scores ranged from 6 to 25, with

a median score of 14.5. Male aids' scores ranged from

7-19, with a median score of 12.

In order to control for effects of race in the aid-

patient interaction, only white aids were used for the

study. Scores for the 36 white female aids ranged from

6 to 23, with a median score of 15. Scores for the 17

white males ranged from 7 to 19, with a median score of 11.

The four highest and four lowest scorers among both

the female and male groups agreed to participate in the











study. Females dosign-ated As ha', scores of 23, 22, 22,

and 22. Female Bs had scores of 6, 8, 9, and 9. Males

designated As had scores of 13, 10, 14, and 14. Hales

designated Bs had score. of 7, 3, 8, and 11.

Female As had an average of 5 years and 10 months'

experience as aids, with a range of from 1 year to 12

years, 5 months' experience. Their ages were 42, 29, 35,

and 25. Female Bs averaged 6 years' experience, with a

range of from 4 years and five months to 9 years' experi-

ence. Their ages were 40, 35, 26, and 28.

Male As averaged 2 years and 4 months' experience,

with a range of from 8 months to 6 years and 3 months'

experience as aids. Their ages were 45, 52, 24, and 23.

Male Bs had an average of 2 years and 5 months' experi-

ence with a range of from 8 months to 4 years and 6 months'

experience. Their ages were 52, 48, 25, and 29.


Patients: One hundred-twenty-four male inpatients diagnosed

upon last admission by the staff of the Northeast Florida

State Hospital as schizophrenic were used as patient-subjects.

Subcategories of schizophrenia such as catatonic, simple, etc.,

were ignored for purposes of this study. Patients with

organic complications or severe mental deficiency were ex-

cluded. Sixty-six of the patients were white and fifty-

eight were black.

Patients were classified as of relatively higher or







43


relatively lower social competence with regard to the total

sample on the criteria established by Zigler and Phillips

(1960). Social competence is defined in terms of six

objective biographical variables. These are age, intelli-

gence, educational level, occupational level, employment

history, and marital status. The categories of each

variable and their order from low to high are presented

below.

1. Age--24 and below, 25-44, and 45 years and above.

2.* Intelligence--I.Q.s obtained on a standard intelli-

gence test of 84 or less, 85-115, and 116 and above.

3. Education--none or some grades including ungraded or

special classes or finished grade school; some high

school or finished high school; and some college or

more.

4. Occupation--Tne Dictionary of Occupational Titles

(United States Government Printing Office, 1965) was

employed to place each occupation into the categories

of unskilled and semiskilled; skilled, service, clerical,

and sales; and professional and managerial.

5. Employment history--usually unemployed; seasonal,

fluctuating, frequent shifts, part-time employment;

and regularly employed.


*This variable was not used in the classification schema
of the present study. See text for explanation.










6. Marital status--single; separated, divorced, remarried,

and widowed; and single continuous marriage (Phillips

and Zigler, 1963, p. 138).

By means of a search of the files of 13 male wards

at the hospital, 175 potential patient-subjects were

chosen. The number selected from each ward ranged from 2

to 24 patients. Data from the patient files was insufficient

in the majority of cases to establish a social competency

score for the patient, thus necessitating an interview with

individual patients to gain supplementary social history

data. In no case was intelligence included as a social

competence variable because I.Q. scores were not available.

For any of the indices, placement in the lowest cate-

gory resulted in a score of 0 for that index. Assignment

to the middle category resulted in a score of 1, and place-

ment in the highest category resulted in a score of 2. The

overall social competence score for each patient was the

mean of the scores obtained on the individual indices.

This averaging procedure was necessary because data were

not available for every patient on all six variables. Thus

the final social competence score for any individual could

range from 0 to 2.

The median social competency score for the 124 patients

actually used in the study fell between .6 and .8. For

purposes of the research a score of .6 or lower was desig-

nated as a low social competence score. This resulted in a











total of 66 high social competence patients and 58 low

social competence patients.

Scores for the total 124 patients ranged from a low

of .2 to a high of 1.8, with a mean of .723, and a standard

deviation of .12. Whites comprised 62% of the high social

competence patients and 42% of the low social competence

patients. Ages of the patients ranged from 19 to 69, with

a mean age of 42.67. The diagnosis of schizophrenic,

chronic undifferentiated type, had been given to 46.8% of

the patients. The remainder carried diagnoses of paranoid

type, simple type, childhood type, catatonic type, hebe-

phrenic type, or schizo-affective type.



Instruments


Instruments used in the experiment were a 30-item A-B

questionnaire administered to the aids; written instructions

to the aids as to how to administer the picture-rating task;

score sheets for the picture rating task; standardized

instructions read by the aids to the patients as to how to

rate the photographs; a tape recorder; a rating scale shown

by the aids to the patients as a visual aid in making judg-

ments in the picture rating task; five-point Likert-type

rating scales covering five qualities on which the aids

rated their impressions of the patients; five-point Likert-

type rating scales covering six qualities on which the











patients rated their impressions of the aids; and score

sheets used to rate voice quality.

The 30-item A-B questionnaire given the aids was the

Kemp (Kemp and Stephens, 1971) version of the A-B scale,

less one item, discarded because it was thought the content

of the item (e.g., I have had no difficulty in starting or

holding my bowel movement) might unnecessarily arouse hos-

tility or suspicion. Kemp's A-B questionnaire consists of

the 23 items of the original White horn and Betz (1960)

scale plus 8 items from the MIMPI found to correlate highly

with those items. The questionnaire is shown in Appendix A.

Within the 30 items of the questionnaire were the 23

SVIB items of the Whitehorn and Betz (1960) scale, and the

23 items of the Schiffman, Carson, and Falkenberg (Kemp

and Stephens, 1971) modification of the Kemp version, which

is more internally consistent than either of the other two

scales. Answers were scored according to these (Whitehorn

and Betz, 1960; Schiffman, Carson, and Falkenberg) versions

of the A-B scale as well as to the total 30-item scale, with

the finding that the 16 aids used in the study maintained

the same rank order regardless of which scoring system was

used.

Standardized instructions were read to each aid prior

to the aid's administering the picture rating task. These

instructions are in Appendix B.

Standardized instructions were read by the aids to the











patients in administering the picture rating task. These

instructions are in Appendix C. In addition, a rating scale

was shown to each patient by the aid in order to help the

patient understand the instructions. The rating scale is

shown in Appendix D.

Ten standardized photographs of male and female faces

used in the picture rating task were provided courtesy of

Dr. Robert Rosenthal, Harvard University.

Each aid and each patient were provided a sheet con-

taining questions, in Likert rating-scale form, on which they

were instructed to rate their impressions of their partner

in the picture rating task. These are shown in Appendices

E and F.

A sample of the score sheets used by undergraduate judges

in rating the audiotape of aids' voices when reading the pic-

ture rating task instructions to patients is shown in Appendix

G.


Procedure


The heart of the experimental procedure was the Rosenthal

(1966) picture rating task, a task originally designed to

measure the effect of covert experimenter bias on experiment

outcome. The task in the present study involved each aid (E)

presenting to each of eight patients* seen consecutively a


*One female A and one male A saw only seven patients each.
One male B saw only six patients.











standardized series of ten photographs of male and female

faces. (Rosenthal, 1966) Each E read a standardized set

of instructions to each S, asking him to rate each photo-

graph on the apparent success or failure that the person

pictured had been experiencing. Each S was asked to say

where on a scale running from +10 (extreme success) to

-10 (extreme failure) each person pictured would fall.

The aid scored the patient's responses.

Expectancies were established by means of the instruc-

tions read to the Es, and by the author telling the E

immediately prior to the E's meeting each patient that

the patient was one of two basic personality types, and

would therefore average +5 or -5 on his photo ratings.

It was made clear to the Es that the Ss had never taken

the picture rating task before, but that previously

administered personality tests indicated that they would

probably score in the predicted direction. Approximately

half of the eight Ss seen by each E were randomly assigned

to the +5 expectancy group and half to the -5 expectancy

group. Without experimenter expectancy being established,

Ss should in fact produce a mean score of zero in rating

the pictures. (Rosenthal, 1966)

Each aid was brought into an interview room and was

given the instructions for the picture rating task (see

Appendix B) and familiarized with the procedure which would

follow and the various instruments needed to carry out the











experiment. The aid was told that "we" were in the process

of developing a test of how well a person can judge the

success or failure of another person, and advised that the

author would tell him what scores to expect from each patient

just prior to his meeting the patient.

Patients in groups of from eight to twelve members,

from a ward other than the one on which the aid worked, were

brought to a waiting room and told that they would meet an

aid who was going to show them some pictures and have them

rate them. They were assured that the task was not diffi-

cult and would take very little time to complete.

After introducing each patient to the aid in the

interview room, the author left the room and waited outside.

The aid turned on a tape recorder and read the instructions

to the patient. After completing the picture rating task,

each patient was escorted by the author to a third room

where he was asked to respond to the questionnaire concern-

ing his impressions of the aid who had shown him the photo-

graphs. Then the patient was asked for any relevant social

history information not included in his file. Meanwhile

the aid filled out a questionnaire concerning his impressions

of the patient. This procedure was followed until the aid

saw eight patients who were willing to cooperate and able

to follow the relevant instructions.

The author did not know which patients would be classi-

fied as high social competence or low social competence











patients prior to introducing themi to the aid. Post-

experimental analysis showed that female As saw 15 low

social competence and 16 high social competence patients.

Female Bs saw 16 high and 14 low social competence

patients. Male As saw 16 high competence and 15 low

competence patients, and male Bs saw 18 high competence

and 13 low competence patients.

Later undergraduate psychology students were

recruited to rate voice qualities of the aids. The

five male and three female undergraduates who served

as judges of voice quality heard audiotapes of each aid

reading the instructions once under each of four condi-

tions: reading to a low social competence patient with

an established expectancy of +5; reading to a low social

competence patient with a -5 expectancy; reading to a

high social competence patient with a +5 expectancy; and

reading to a high competence patient with a -5 expectancy.

These four reading were randomly sampled from each E's

eight Ss.

Judges were given a booklet of rating sheets on

which to rate each voice heard. The twelve qualities

were varied in order of presentation for each voice heard

in order to minimize rating bias. Judges were not given

any definitions of the twelve vocal qualities they were

required to rate.











Statistical Treatment


The results of each hypothesis were tested for signifi-

cance by one tailed t-tests for independent samples. The

significance level for these t-tests was set at .05.

Hypotheses Al, A2, Bl, and B2 predict differential

experimenter expectancy effects produced under four separate

experimental conditions. In order to simplify computation

of the picture rating task data upon which these hypotheses

rest, a score of 10 was assigned to a patient who averaged 0

on the task. A score in the direction of induced expectancy

was added to 10, and a score in the direction opposite of

induced expectancy was subtracted from 10. For example,

if a patient for whom a +5 expectancy was given scored a

mean of +2.7 on the ten pictures, his score was converted

for computational purposes to 12.7. If a patient scored

a mean of -2.7 on the ten pictures, his score was converted

to 7.3. Thus there were no negative scores with which to

work.

The data related to the series of hypotheses A3-A7 and

B3-B7, dealing with effects of E and S pairings on the voice

quality of Es, was generated by judges listening to and

rating audiotapes on a five point Likert-type scale. Analysis

of inter-rater reliability was accomplished by Ebel's formula

for interclass correlation, which yields essentially an

average intercorrelation of ratings from n raters.















RESULTS


Experimenter Expectancy Hypotheses


Hypothesis Al predicted that A type male Es would

exert greater expectancy effects with low S-C Ss than

would B type male Es. Analysis shows there is no statis-

tically significant difference between the mean scores

obtained from low S-C Ss by A and B male aids. Hypothesis

A2 predicted a greater expectancy effect would be exerted

on low S-C Ss by A type female Es than by B type female

Es. This hypothesis was not supported by the data.

Although As more effectively biased low S-C Ss than did

Bs, this relationship was not statistically significant.

Results are summarized in Table 1.

In a comparison of means not required by the hypo-

theses, it was found that female As exerted greater

expectancy effects with low S-C Ss than did male Bs. Mean

expectancy scores for low S-C Ss were 10.98 with female

As serving as Es, and 8.77 with male Bs, with a standard

deviation of 3.23, and a t-score of 1.80(p< .05).

Hypotheses Bl and B2 predicted that male B type Es

and female B type Es, respectively, would exert greater

expectancy effects on high S-C Ss than would their A type











counterparts. Both hypotheses were supported by the data.

Results are summarized in Table 1.

Additional analysis shows that there were no main

effects of social competence level, experimenter type,

or experimenter sex on expectancy scores (Table 2).

Thus the interaction effect with regard to experi-

menter expectancy did not take place fully as predicted,

in that As were not found to be significantly more effec-

tive than their B counterparts in influencing the ratings

of high S-C Ss. The interaction effect was partially

confirmed however, in that B aids of both sexes were

more effective than their A counterparts in influencing

the ratings of high S-C Ss.













TABLE 1

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON EXPERIMENTER EXPECTANCY SCORES ATTAINED WITH HIGH AND
LOW SOCIAL COMPETENCE SCHIZOPHRENICS



Low S-C Ss High S-C Ss


Mean S.D. t Mean S.D. t

Male A
type Es 9.46 9.09
2.72 .68 N.S. 2.84 1.86*
Male B
type Es 8.77 10.92



Female A
type Es 10.98 8.69
3.24 .15 N.S. 3.44 1.75*
Female B
type Es 10.80 10.82


*p< .05








'55


TABLE 2


COMPARISON OF A TYPE Es vs. B TYPE Es; MALE Es vs. FEMALE
Es; AND HIGH vs. LOW SOCIAL COMPETENCE SCHIZOPHRENICS
ON EXPERIMENTER EXPECTANCY SCORES ATTAINED



Mean S.D. t



A type Es 9.53
3.17 1.52 N.S.

B type Es 10.41


Male Es 9.64
3.14 1.15 N.S.


Female Es 10.30



Low S-C Ss 10.03
3.15 .14 N.S.

High S-C Ss 9.91











Voice Quality Hypotheses


Ebel's formula for inter-rater reliability (Guilford,

1954) yielded an average r of .31 with a range of from .20

to .36 for each voice quality x aid type x patient type

condition. This low inter-rater reliability renders any

findings regarding the various voice quality hypotheses of

doubtful validity. The following results, which are

summarized in Tables 3, 4, 5, and 6, are reported with

this caveat in mind.

Hypothesis A3 predicted that judges would rate the

taped voices of both male and female A type Es higher than

their B type counterparts in sophistication, self-confidence,

professionalness, warmth-friendliness, and enthusiasm, when

reading instructions to low S-C Ss. None of the components

of this hypothesis was confirmed with one exception: voices

of male type As were judged to have significantly more

enthusiasm than those of male type Bs in reading the instruc-

tions to the picture rating task to low S-C Ss. There were

no significant differences found between the voices of

female As and Bs when paired with low S-C Ss. (Tables 3 and

4)

Hypothesis A4 predicted that judges would rate both

male and female A type Es lower than their B type counter-

parts in the voice qualities of discomfort and coldness-

distance when reading to low S-C Ss. Only one of the four








S57


components of this hypothesis was confirmed: Judges rated

female Bs significantly higher in discomfort than female

As when paired with low S-C Ss. (Table 4)

The predictions of hypothesis A5 were not upheld by

the data. Judges did not rate A type male Es higher than

B type male Es in masculinity and dominance when reading

to low S-C Ss. (Table 3)

Hypothesis A6 was concerned with female As and Bs.

The predictions were that taped voices of As would be rated

higher than those of Bs in the qualities of femininity,

kindness, and softness when reading instructions to low

S-C Ss. This hypothesis was not confirmed by the data.

(Table 4)

The last voice quality hypothesis dealing with

instructions read to low S-C Ss predicted that with these

Ss the voices of female As would be rated lower in dominance

than those of female Bs. The data did not support this

hypothesis. (Table 4)

Hypotheses B3-B7 were concerned with the effect on

judge's rating of the pairings of As and Bs with high S-C Ss.

Hypothesis B3 predicted that judges would rate the

taped voices of both male and female B type Es higher than

their A type counterparts in sophistication, self-confidence,

professionalness, warmth-friendliness, and enthusiasm, when

reading instructions to high S-C Ss. The data did not support

this hypothesis. In the only statistically significant











finding, voices of male As were rated higher in profes-

sionalness than male Bs when reading to high S-C Ss.

This results is opposite to that predicted. (Tables 5

and 6)

Hypotheses B4-B7 all failed to be confirmed by the

data. Hypothesis B4 predicted that voices of both male

and female Bs would be rated lower than those of As in

discomfort and coldness-distance when reading instructions

to high S-C Ss. Hypothesis B5 predicted that voices of

male Bs would be rated higher than those of male As in

masculinity and dominance when reading to high S-C Ss.

Hypothesis B6 predicted that in this condition voices of

female Bs would be rated higher than those of female As

in femininity, kindness, and softness. The final hypo-

thesis of this series predicted that judges would rate

the taped voices of female Bs lower than female As in

dominance when reading to high S-C ss. (Tables 5 and 6)

In summary, there was little confirmation of the voice

quality hypothesis. Of forty comparisons of means, only

three t-tests indicated significant results. One of these

results was in the direction opposite that predicted by

the hypothesis. Low inter-rater reliability places the

value of these results in question.











TABLE 3


COMPARISON OF MALE As vs. MALE Bs ON RATED VOICE
WHEN READING INSTRUCTIONS TO LOW S-C Ss


QUALITIES


Mean S.D. t


Sophistication


Self-Confidence


Professionalness


Warmth-Friendliness


Enthusiasm


Discomfort




Coldness-Distance


Masculinity




Dominance


Male As

Male Bs


Male As

Male Bs


Male As

Male Bs


Male As

Male Bs


Male As

Male Bs


Male As

Male Bs


Male As

Male Bs


Male As

Male Bs


Male As

Male Bs


1.10


1.13


1.21 N.S.


.76


1.65

1.48


1.39


1.14


1.28


.61 N.S.


1.27 N.S.


.97


1.52

1.29


1.47


1.19


1.0


.72 N.S.


2.08*


.97


1.67

1.89


2.16

2.45


2.39

1.93


2.31

2.02


1.18


1.21


1.45


1.52


.79 N.S.


.91 N.S.


1.17 N.S.


.76 N.S.


*p< .025








60

TABLE 4

COMPARISON OF FEMALE As vs. FEMALE Bs ON RATED VOICE QUALITIES
WHEN READING INSTRUCTIONS TO LOW S-C Ss

Mean S.D. t


Sophistication



Self-Confidence



Professionalness



Warmth-Friendliness



Enthusiasm



Discomfort



Coldness-Distance



Dominance



Femininity



Kindness


Softness


Female
Female

Female
Female

Female
Female

Female
Female

Female
Female

Female
Female

Female
Female

Female
Female

Female
Female

Female
Female


As
Bs

As
Bs

As
Bs

As
Bs

As
Bs

As
Bs

As
Bs

As
Bs

As
Bs

As
Bs


Female As
Female Bs


1.53
1.41

1.90
1.76

1.80
1.65

2.11
2.31

1.50
1.62

1.23
1.62

1.80
1.47

1.49
1.44

2.39
2.41

2.23
2.15

2.07
2.09


1.36


1.13



1.30



.94



1.15



.93



1.13



1.08



1.37



.95


1.13


.22 N.S.



.50 N.S.



.45 N.S.



.80 N.S.



.46 N.S.



1.77*



1.17 N.S.



.19 N.S.



.06 N.S.



.33 N.S.


.07 N.S.


*p< .05


--










TABLE 5


COMPARISON OF MALE As vs. MALE Bs ON
WHEN READING INSTRUCTIONS TO


RATED VOICE QUALITIES
HIGH S-C Ss


Mean S.D. t


Sophistication



Self-Confidence


Professionalness


Warmth-Friendliness



Enthusiasm


Discomfort


Coldness-Distance



Masculinity



Dominance


Male As
Male Bs

Male As
Male Bs

Male As
Male Bs

Male As
Male Bs

Male As
Male Bs

Male As
Male Bs

Male As
Male Bs

Male As
Male Bs

Male As
Male Bs


1.20
1.01

1.96
1.84

1.43
1.90

1.70
1.51

1.74
1.76

1.68
1.51

1.91
2.19

2.20
2.25

1.87
2.05


1.05



1.24


1.10


1.23



.95


.90


1.32



1.30



1.22


.73 N.S.



.46 N.S.


2.70*


.73 N.S.



.10 N.S.


.85 N.S.


.64 N.S.



.18 N.S.



.75 N.S.


*p< .005


--








62


COMPARISON OF FEMALE
WHEN READING


TABLE 6
As vs. FEMALE Bs ON RATED VOICE
INSTRUCTIONS TO HIGH S-C Ss


QUALITIES


Mean S.D. t


Sophisticated



Self-Confidence


Professionalness


Warmth-Friendliness


Enthusiasm



Discomfort



Coldness-Distance



Dominance



Femininity



Kindness


Softness


Female As
Female Bs

Female As
Female Bs


Female As
Female As


Female
Female


Female As
Female Bs

Female As
Female Bs

Female As
Female Bs

Female As
Female Bs

Female As
Female Bs

Female As
Female Bs


Female As
Female Bs


1.24
1.38


1.38
1.44


1.64
1.64


2.19
2.16

1.48
1.48

1.76
1.62

2.04
1.94

1.44
1.51

2.38
2.28

1.99
1.80


1.97
2.05


1.30



1.30


.47 N.S.



.20 N.S.


1.44


1.30


.88



1.22



1.28



1.30



2.07



1.43



1.23


.10 N.S.


.05 N.S.



.15 N.S.



.36 N.S.



.23 N.S.



.21 N.S.



.58 N.S.



.31 N.S.











Interpersonal Impressions Hypotheses


This series of hypotheses dealt with the impressions

Es and Ss formed of each other during their brief inter-

actions. It was hypothesized that different pairings would

produce different ratings as predicted by the interaction

hypothesis.

Hypotheses A8-A10 dealt with the ratings made by low S-C

Ss of their own impressions of the A and B type Es who adminis-

tered the picture rating task. Results are summarized in

Tables 7-12.

The prediction made by hypothesis A8 was that low S-C

Ss would rate themselves as liking more, and feeling more

comfortable with both male and female As, and that they would

rate both male and female As more friendly and self-confident

than their B counterparts. Statistical analysis confirms one

component of this hypothesis: Low S-C Ss did rate themselves

as liking female As more than female Bs. (Table 7) An

additional and unexpected finding was that both male and

female Bs were rated as more self-confident than As. (Table

10) This finding was opposite to that predicted. Hypotheses

A9 and A10 were not confirmed. They predicted that low S-C

Ss would rate male As as more dominant than male Bs, and

female As as more gentle and kind than female Bs.

Hopethesis All predicted that both male and female As

would rate themselves as liking more and being more comfortable










with low S-C Ss, and that they would rate these Ss as more

at ease, cooperative, and involved in the picture rating

task. This hypothesis was not supported by the data.

Results are summarized in Tables 13-17.

Hypotheses B8-B10 were concerned with the ratings

made by high S-C Ss of their impressions of A and B type

Es. Results are summarized in Tables 18-23.

According to hypothesis B8, high S-C Ss would rate

themselves as liking more and feeling more comfortable with

both male and female Bs than with their A counterparts, and

they would also rate male and female Bs as more friendly and

self-cofident. This hypothesis was not confirmed by the

data. Contrary to expectation, high S-C Ss rated them-

selves as liking A type female more than B type females.

(Table 18)

Hypotheses B9 and B10 predicted that high S-C Ss would

rate male Bs as more dominant than male As, and female Bs

more gentle and kind than female As. Neither hypothesis

was confirmed by the data.

Hypothesis Bll predicted that both male and female Bs

would rate themselves as liking more and being more comfortable

with high S-C Ss, and that they would also rate these Ss as

more at ease, cooperative, and involved in the picture rating

task. Results of statistical analysis indicate that male Bs

did rate themselves as liking high S-C Ss more than low S-C

Ss, and that they saw high S-C Ss as more cooperative in the











experimental session (Tables 24 and 27). Both results

are as predicted. The other predictions of Hypothesis Bl1

were not confirmed by the data (See Tables 24-28).

In summary, of forty comparisons of means undertaken

to test the series of hypotheses concerning interpersonal

impressions formed by Es and Ss, four were significant in

the predicted direction, and three were statistically

significant in the direction opposite that predicted. The

findings were essentially threefold: 1) Female As were

liked better by both high and low S-C Ss than were female

Bs. 2) Low S-C Ss rated As of both sexes less self-

confident than their B counterparts. 3) Male Bs liked

high S-C Ss more than they liked low S-C Ss, and saw high

S-C Ss as more cooperative during the picture rating task

administration.


TABLE 7

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY LOW SOCIAL COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "DID YOU LIKE THE AID?"1

Mean S.D. t

Male A type Es 3.2
1.18 .44 N.S.
Male B type Es 3.0

Female A type Es 3.53
.93 1.91*
Female B type Es 2.86


Ratings were made on a five-point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).
*p< .05












TABLE 8

COMPARISON OF MALE Ss vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY LOW SOCIAL COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "DID YOU FEEL AT EASE WITH THE AID?"'



Mean S.D. t


Male A type Es 2.73
1.42 .68 N.S.
Male B type Es 2.54


Female A type Es 3.2
1.16 .33 N.S.
Female B type Es 3.14



TABLE 9

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY LOW SOCIAL COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "WAS THE AID FRIENDLY?"1



Mean S.D. t


Male A type Es 3.06
1.11 1.0 N.S.
Male B type Es 3.15


Female A type Es 3.47
.64 .42 N.S.
Female B type Es 3.57



1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).












TABLE 10

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY LOW SOCIAL COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "WAS THE AID SELF-CONFIDENT?"1


Mean S.D. t


Male A type Es 2.67
1.10 1.69*
Male B type Es 3.38


Female A type Es 2.6
.96 2.5**
Female B type Es 3.57


*p< .05
**p< .005



TABLE 11
COMPARISON OF MALE As vs. MALE Bs ON RATINGS MADE BY LOW
SOCIAL COMPETENCE SCHIZOPHRENICS ON THE QUESTION "WAS THE
AID A STRONG, DOMINANT PERSON?"



Mean S.D. t


Male A type Es 2.93

1.08 1.10 N.S.

Male B type Es 3.38



1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).











TABLE 12

COMPARISON OF FEMALE As vs. FEMALE Bs ON RATINGS MADE BY LOW
SOCIAL COMPETENCE SCHIZOPHRENICS ON THE QUESTION "WAS THE
AID A GENTLE, KIND PERSON?"1


Mean S.D. t


Female A type Es 3.07
1.09
Female B type Es 3.07




TABLE 13

COMPARISON OF LOW SOCIAL COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE
AND FEMALE As ON THE QUESTION "DID YOU LIKE THIS PATIENT?"1

Male As
Mean S.D. t


Low S-C Ss 3.33
.64 .42 N.S.
High S-C Ss 3.43

Female As
Mean S.D. t


Low S-C Ss 2.93
.83 1.07 N.S.
High S-C Ss 3.25


scale ranging


1Ratings were made on a five point Likert-type
from 0 (Not at all) to 4 (A lot).










TABLE 14

COMPARISON OF LOW SOCIAL COMPETENCE SCHIZOPHRENICS vs.HIGH
SOCIAL COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE
AND FEMALE As ON THE QUESTION "DID YOU FEEL AT EASE WITH
THIS PATIENT?"


Male As

Mean S.D. t


Low S-C Ss 3.67
.64 .42 N.S.
High S-C Ss 3.5



Female As

Mean S.D. t


Low S-C Ss 3.33
.68 1.0 N.S.
High S-C Ss 3.57


TABLE 15
COMPARISON OF LOW SOCIAL COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE AND
FEMALE As ON THE QUESTION "DID YOUR PATIENT FEEL AT EASE?-"1

Male As
Mean S.D. t


Low S-C Ss 2.13 .62 .77 N.S.
High S-C Ss 3.00

Female As
Mean S.D. +
Low S-C Ss 3.47.68 1.0 N.S.
High S-C Ss 3.25


1Ratings were made on a five point Likert-type scale ranging
from (Not at all) to 4 (A lot).












TABLE 16

COMPARISON OF LOW SOCIAL-COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL-COMPETEL;CL SCHIZOPHRENICS ON RATINGS MADE BY MALE AND
FEMALE As ON THE QUESTION "DID THE PATIENT COOPERATE?"1



Male As
Mean S.D. t

Low S-C Ss 2.92
1.24 .16 N.S.

High S-C Ss 3.0


Female As
Mean S.D. t


Low S-C Ss 3.4
.66 1.41 N.S.
High S-C Ss 3.06



TABLE 17
COMPARISON OF LOW SOCIAL-COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL-COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE AND
FEMALE As ON THE QUESTION "DID THE PATIENT SHOW INTEREST IN
RATING THE PHOTOS?"1

Male As
Mean S.D. t

Low S-C Ss 2.6
1.33 .21 N.S.
High S-C Ss 2.5

Female As
Mean S.D.

Low S-C Ss 3.2
1.37 1.31 N.S.
High S-C Ss 2.56

IRatings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).











TABLE 18

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY HIGH SOCIAL COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "DID YOU LIKE THE AID?"1


Mean S.D. t


Male A type Es

Male B type Es


Female A type Es

Female B type Es

*p<.05


3.38

2.78


3.0

2.44


1.31



.86


1.3 N.S.



1.86*


TABLE 19

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY HIGH SOCIAL COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "DID YOU FEEL AT EASE WITH THE AID?"1


Mean S.D. t

Male A type Es 3.13
1.52 1.08 N.S.
Male B type Es 2.56


Female A type Es 2.75
1.24 1.63 N.S.
Female B type Es 3.06


1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).








72


TABLE 20

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY HIGH SOCIAL-COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "WAS THE AID FRIENDLY?"l


Mean S.D. t


Male A type Es 3.31
1.31 .43 N.S.

Male B type Es 3.11


Female A type Es 3.44
.73 .23 N.S.
Female B type Es 3.38



TABLE 21

COMPARISON OF MALE As vs. MALE Bs, AND FEMALE As vs. FEMALE
Bs ON RATINGS MADE BY HIGH SOCIAL-COMPETENCE SCHIZOPHRENICS
ON THE QUESTION "WAS THE AID SELF-CONFIDENT?"1


Mean S.D. t


Male A type Es

Male B type Es


Female A type Es

Female B type Es


3.31

3.11


3.13

3.25.


1.31


.43 N.S.


.38 N.S.


1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).












TABLE 22


COMPARISON OF MALE As vs. MALE Bs ON RATINGS BY HIGH SOCIAL
COMPETENCE SCHIZOPHRENICS ON THE QUESTION "WAS THE AID A
STRONG, DOMINANT PERSON?"1



Mean S.D. t


Male A type Es 3.0
1.57 .71 N.S.
Male B type Es 2.78



TABLE 23


COMPARISON OF FEMALE As vs. FEMALE Bs ON RATINGS MADE BY
HIGH SOCIAL COMPETENCE SCHIZOPHRENICS ON THE QUESTION "WAS
THE AID A GENTLE, KIND PERSON?"1



Mean S.D. t


Female A type Es 3.31
.87
Female B type Es 3.31



1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).











TABLE 24


COMPARISON OF LOW SOCIAL-COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL-COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE
AND FEMALE Bs ON THE QUESTION "DID YOU LIKE THIS PATIENT?"1

Male Bs

Mean S.D. t

Low S-C Ss 2.38
.96 2.08*
High S-C Ss 3.11

*p<.05

Female Bs


Low S-C Ss


High S-C Ss


2.36


2.44


.83


.27 N.S.


TABLE 25

COMPARISON OF LOW SOCIAL-COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL-COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE AND
FEMALE Bs ON THE QUESTION "DID YOU FEEL AT EASE WITH THIS
PATIENT?"1
Male Bs
Mean S.D. t

Low S-C Ss 2.77
.88 1.93 N.S.
High S-C Ss 3.39


Female Bs

Low S-C Ss 2.64
1.03 .44 N.S.
High S-C Ss 2.81

1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).
*P<.05











TABLE 26

COMPARISON OF LOW SOCIAL-COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL-COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE
AND FEMALE Bs ON THE QUESTION "DID THE PATIENT FEEL AT EASE?"1

Male Bs
Mean S.D. t


Low S-C Ss 2.08
1.17 1.0 N.S.
High S-C Ss 2.50


Female Bs
Mean S.D. t


Low S-C Ss 2.29
.85 1.25 N.S.
High S-C Ss 2.69



TABLE 27

COMPARISON OF LOW SOCIAL-COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL-COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE AND
FEMALE Bs ON THE QUESTION "DID THE PATIENT COOPERATE?"1

Male Bs

Mean S.D. t

Low S-C Ss 2.77
1.03 2.14*

High S-C Ss 3.56


1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).
*p< .05








76


TABLE 28

COMPARISON OF LOW SOCIAL-COMPETENCE SCHIZOPHRENICS vs. HIGH
SOCIAL-COMPETENCE SCHIZOPHRENICS ON RATINGS MADE BY MALE
AND FEMALE Bs ON THE QUESTION "DID THE PATIENT SHOW INTEREST
IN RATING THE PHOTOS?"'


Male Bs

Mean S.D. t

Low S-C Ss 2.38

1.48 .83 N.S.

High S-C Ss 2.83




Female Bs

Mean S.D. t


Low S-C Ss 2.21
1.20 .95 N.S.
High S-C Ss 2.63



1Ratings were made on a five point Likert-type scale ranging
from 0 (Not at all) to 4 (A lot).













DISCUSSION


Perspective


This study is one of many which deal with A-B x patient

type interaction phenomena. Although there have been over

eighty published studies examining various aspects of this

phenomena, including studies of psychotherapy process and

outcome; psychotherapy analogue studies; studies of atti-

tudes of As and Bs toward differing types of patients; and

studies of personality factors characterizing As and Bs,

the data that has been garnered can only be characterized

as inconclusive and sometimes contradictory.

This dissertation is a replication and extension of

a study by Trattner and Howard (1970). That study was

seen as a potentially important addition to the A-B litera-

ture primarily because the focus of study was the communica-

tion of expectancies, which is seen as a potent factor in

successful psychotherapy and case management, as well as

a phenomenon readily measurable by a well worked-out

standardized procedure. The heart of the experimental

procedure, the administering of the Rosenthal picture

rating task by an aid to a patient, effectively narrows the

range of behaviors under study so that a few variables can

be isolated as dependent on the aid x patient pairings.


77











This is because the procedure allows only a minimal and

substantially standardized interaction between the aids

and patients. Variation in content of the verbalization

directed to the patient by the aid is held to an absolute

minimum. This allows voice quality to be studied with

verbal content controlled. Essentially the aid greets

the patient, asks him to have a seat, and then reads him

a set of standardized instructions. If the expectancy

is to be communicated, it must be within this circumscribed

framework.

The present study used a larger number of patients

and aids as experimental subjects than did the Trattner

and Howard (1970) study, in an effort to increase the

generalizability of the findings. Female aids were added

in order to determine if the phenomenon being studied was

applicable to the many females working as psychiatric

aids and in various other capacities within the mental

health field. Four additional vocal qualities, along

with the vocal qualities examined by the Trattner and

Howard (1970) research, were studied. The impressions

each member of the aid x patient pairings formed of each

other were also examined in a further effort to understand

the effect of A-B x patient type pairings on the dyadic

interaction involved in the Rosenthal picture rating task.

The results of the present study are in substantial







79


disagreement with those of the Trattner and Howard (1970)

study. Although Bs of both sexes were significantly

more successful in influencing high social competence

schizophrenics than were As, As were not significantly

more successful in influencing low S-C Ss. The voice

quality findings of Trattner and Howard were not supported.

There were no meaningful patterns indicating As and Bs

vary their voices in ways measurable by this study's

methodology when speaking to different categories of

patients.

The final series of hypotheses of this study did

not relate to the question of the replicability of the

Trattner and Howard research. These hypotheses involved

the impressions, or "attitudes" the aids and patients

interacting in the picture sorting task formed of each

other during their brief encounter. It was felt that if

A and B aids were differentially effective in uninten-

tionally influencing high and low social competence

patients, that concommitantly, the members of the dyads

formed by the various aid x patient pairing would have

predictably varying impressions of their partners.

Two of the three statistically significant relation-

ships found with regard to this series of hypotheses were

of an interactional nature: Low S-C Ss rated As of both

sexes less self-confident than their B counterparts; and

male Bs liked high S-C Ss more than low S-C Ss, and saw











them as more cooperative. Only one pole of the interaction

is supported in these findings. In other words, although

male Bs liked high S-C Ss more and saw them as more coopera-

tive, the opposite pole of the interaction did not hold

true: Male As did not like low S-C Ss more than high S-C

Ss, nor did they see them as more cooperative.

Conclusions can be drawn from these results, however.

It appears that for male Bs, a patient's likeability is

associated with his perceived cooperativeness, and patients

seen as likeable and cooperative are more successfully

influenced in the Rosenthal picture rating task. If one

accepts the concept that bigh social competence schizo-

phrenics are more like neurotics than are low social

competence schizophrenics, this finding is consonant with

that of Stein, Green, and Stone (1972) where B type male

psychiatric residents assigned significantly higher like-

ability ratings to neurotic vignettes than to schizophrenic

vignettes while As did not differentiate between the two

diagnostic types.

The finding that Bs are seen as more self-confident

than As (by low social competence patients) is consistent

with the characterization of Bs as being, compared to As,

more masculine (Lorr and McNair, 1966; Dublin, 1969;

Berzins et al., 1971), risk-taking, dominant, variety-

seeking, and counterdependent. (Berzins et al., 1971).

There must of course be some reservations in giving











weight to the ratings of aids mad?, by schizophrenic

inpatients, who are very disturbed in the realm of inter-

personal perception and judgment.

Additionally, the psychiatric aids' ratings in this

study are also somewhat suspect in light of the observa-

tion (discussed more fully below) that the aids seemed to

perceive all the patients in the hospital as very much

the same; that is, as persons to be manipulated in order

to maintain efficiency and order in the hospital routine.

This resulted in a sort of I-it relationship, which being

antithetical to successful psychotherapy, severely limits

the generalizability of the aids' ratings to the psycho-

therapeutic situation.


General Considerations and Conclusions


The first series of hypotheses, predicting differen-

tial expectancy effects from the various aid x patient

pairings, were the heart of this study. The second and

third series of hypotheses were subsidiary to the first,

as they were directed at clarifying why the interaction

effect predicted by the first series of hypotheses did in

fact take place. Below is a discussion of possible reasons

for the failure of the interaction effect found by Trattner

and Howard (1970) to take place in this study.

The greatest difference between the Trattner and

Howard (1970)experiment and the present one may very well











have been the nature of the aids used as Es. In both

studies the aids were selected by the sole criteria of

their completing and returning an A-B questionnaire,

and their having scores on the extreme ends of the

distribution of questionnaire scores. It is likely

that the aids participating in the two studies, and the

hospital milieu in the two studies, were unalike in

several significant ways, and that these differences

may have affected the outcomes of the two experiments.

The investigator worked out of the office of a

clinical psychologist at the Florida Northeast State

Hospital for approximately one month while working on

various aspects of this study. During this time the

investigator was able to talk to several aids and staff

members of the hospital. These conversations, along with

personal observations of the hospital's day to day func-

tioning are the base upon which much of the following

thinking rests.

The hospital is one of the major sources of employ-

ment in the area. The aids working there were mostly

residents of Macclenny and nearby areas. Their motiva-

tion for working at the hospital, in many if not most

instances, had little to do with any interest in working

with "mentally ill" persons. Additionally, they were

not consulted by the professional staff to help in forming

treatment plans, nor were they in general treated as











significant, contributing members of the mental health

team. Perhaps for these reasons, it appeared that very

often patients were perceived by aids as nuisances, to

be dealt with in a custodial manner geared to minimize

disruption of the routine of the hospital.

This set of circumstances may have affected the

study's outcome in a fundamental way. If A and B aids

are to interact differently with different kinds of

patients, as they did in the Trattner and Howard (1970)

research, they must be able to distinguish, at some level

of awareness, between low and high social competence

patients. Trattner and Howard theorized that patients'

level of social competence was communicated to the aids

by the patients' non-verbal communications: how they

looked, moved, and talked. The aids in the present

study were unlikely to attend to a patient's unique

non-verbal communications, and were therefore unlikely

to differentiate between high and low social competence

patients.

A related matter is the impact of racial prejudice

on the attitudes of the aids toward patients and on the

aid x patient dyadic interactions. Such a problem may

have been a confounding variable in this study to a greater

degree than in the Trattner and Howard (1970) research,

which took place in Boston. The aids in the present

study were all white, and all resided in rural Florida.











Blacks comprised 380 of the high S--C patients, and 58%

of the low S-C patients. No data on the racial make-up

of the participants in the Trattner and Howard study

was published.

Another of the results of the aids' attitudes

toward patients and their work may have been a lack of

interest in carrying out the procedures required of them

in the experiment. Experimenter expectancy effect is

not likely to be manifested at all if the E in the

experiment is not interested in its outcome. (Rosen-

thal, 1966) Only one of the 16 aids participating

later asked if the patients had rated the pictures as

predicted by the investigator in giving the instruc-

tions establishing experimenter expectancy.

Some of the aids expressed defensiveness about

having to fill out the A-B questionnaire. Many asked

if they were in some way being evaluated, a fear again

expressed by several of the 16 aids serving as Es at

the time the picture rating task was explained to them.

This defensiveness and concern about evaluation of their

performance may have affected the outcome of the study.

An unexpected complication that very probably affec-

ted the outcome of the experimenter expectancy hypotheses,

as well as that of the voice quality hypotheses and the

inter-judge reliability in rating vocal qualities, was

that several of the aids had difficulty in effectively







S85


reading the instructions to the patients. Some aids of

both sexes and of both A and B types mispronounced words,

and/or misread words.

The aids' attitudes toward the patients was perhaps

reflected in the apparent lack of feeling and inflection

in their voices as they read the instructions. This

problem became apparent when the audiotapes were heard.

An important factor contributing to low inter-judge

reliability in rating the tapes was simply boredom

brought on by the frequent poor quality of the aids'

renderings of the instructions.

In conclusion, this study does not support the

findings of the Trattner and Howard study (1970), which

had opened a promising new line of inquiry in this area

of research. The general dearth of theory in A-B research,

along with the general inconsistency of experimental

findings, make it difficult to pinpoint the reasons for

the contradictory findings of this study and that of

Trattner and Howard.

One portion of the study that explored an area not

touched upon in the Trattner and Howard (1970) research

did uncover some particularly useful findings. With

regard to the question as to whether females can be

meaningfully classified as As or Bs and whether their

interactions with patients are related to their A-B

status, this study's answer is clearly in the affirmative.







86


B status was significantly related to the effectiveness

of female aids in influencing the picture rating res-

ponses of high S-C Ss. Performance differences among

A and B females in interactions with patients have not

previously been demonstrated in clinical or analogue

research with the exceptions of a study of psychotherapy

outcome in which two of six therapists involved were

female (Berzins, Ross, and Friedman, 1972), a study

in which B type mothers were more successful than A

type mothers in modifying the enuretic behavior of their

children (James and Foreman, 1973), and the unpublished

Persinger et al. (1968) research in which both male and

female ward personnel served as experimenters giving the

Rosenthal picture rating task to inpatients.

Major recent studies (Berzins, Dove, and Ross, 1972;

Berzins, Barnes, Cohen and Ross, 1971; Dublin, Elton, and

Berzins, 1969) have been successful in finding personality

correlates of A-B status that are invariate regardless of

the sex of the As and Bs studied. Future research in the

A-B area should be directed toward gathering more empirical

evidence that treatment-relevant performance differences

are not limited to male As and Bs.






































APPENDICES













APPENDIX A


PERSONAL INTERESTS QUESTIONNAIRE


Instructions: Please fill out the following questionnaire
as truthfully as possible. This is not a test, so there are
no "right" or "wrong" answers. Along with the answers given
by all of the other attendants at the hospital, your answers
will be pooled as part of some research that is being done
during the next few weeks by a psychologist from the Univer-
sity of Florida. In order to insure confidentiality of your
answers, no individual's questionnaire will be seen by any-
one on the staff of the hospital. Please place your completed
questionnaires face down in the box located in the nurse's
station marked Personal Interests Questionnaires.


Name


(please print)


Sex M F (Circle one)

Ward

Days on duty: M T W Th Sat Sun (Circle the days you are on duty)

Hours of duty shift


For the following items, please indicate the degree of
interest you would have in each of the activities, school sub-
jects, or occupations listed, by placing a circle around the
appropriate answer. Work rapidly.


1. Drilling in a company

2. Marine engineer

3. Mechanical engineer

4. Photo engraver

5. Specialty salesman

6. Toolmaker

7. Making a radio set

8. Building contractor


Like Indifferent

Like Indifferent

Like Indifferent

Like Indifferent

Like Indifferent

Like Indifferent

Like Indifferent

Like Indifferent


Dislike

Dislike

Dislike

Dislike

Dislike

Dislike

Dislike

Dislike




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.9 - mvs