JUDGMENT OF SPRANGER PERSONALITY

TYPES FROM THE VOICE AS TRANS-

MITTED OVER A PUBLIC
ADDRESS SYSTEM

PAUL J. FAY and WARREN C. MIDDLETON

DePauw University

Tbis study represents an attempt to discover by experimental

metbods to wbat extent, if any, speakers over a public address sys-
tem can be classified by listeners into one or anotber of tbe Spranger
dominant life-value types. Spranger believes tbat one's life-pattern
is tbe key to one's personality, and tbat tbis life-pattern is revealed in
tbe value wbicb one sets on certain tbings. He stresses six funda-
mental value types in personality. Tbese types, according to
Spranger's tbeory, appear only in social forms, but, nevertheless,
tbey obey laws wbich cannot be deduced from tbe mere fact tbat
people live in common. Tbe six types are: tbeoretic, social, aestbetic,
economic, political, and religious (8).

PREVIOUS STUDIES
Comparatively few experimental studies have been made on the rela-
tionship between voice and personality. In some experiments speakers have
been presented over a radio or a public address system; in others, no
mechanical aids have been used. In the former type of study, Pear (7)
was a pioneer in demonstrating that personality qualities can be judged
from the radio voice with considerable accuracy. Assisted by the British
Broadcasting Company, he used nine speakers of different sex, age, and
interests. He secured four thousand listeners' judgments concerning the
birthplace, occupation, age, and place of residence of each of these nine
speakers. Free description was elicited from his auditors. Although his
results are difficult to quantify, he found that age was estimated fairly well.
Sex was stated with rather remarkable accuracy, and vocation was occasion-
ally judged with a surprising degree of correctness. Pear does not con-
clude definitely that there are specific types of voices. "Judging from the
readiness with which most people judge voices, one might assume that there
would be well-known classifications into types. If so, they are hard to
find" (7, 7). However, he suggests that in England "a person's voice
 JUDGMENT OF SPRANGER PERSONALITY TYPES 145
often conforms to a stereotype" (7, 28), He speaks of an "army officer's
voice" and he believes that the stage may breed, and perhaps inbreed,
stereotypes; he suggests that with the widespread interest in voices on the
stage and in the talking film, types may be created. In England, he says,
many educational institutions encourage their pupils to speak in a certain
way. Pear is of the opinion that there is no doubt that the phenomenon of
S)mesthesis plays some part in the effect that any particular voice may have
on a listener.
Herzog (6) had radio listeners in Vienna judge the height, weight,
age, sex, and occupation of each of several radio speakers. Replies were
elicited from twenty-seven hundred auditors. She found that all of these
various characteristics were judged more accurately than one might expect
from the mere operation of chance.
AUport and Cantril (1, 3) performed eight experiments on the judgment
of personality from voice. Six experiments were performed in the Harvard
Psychological Laboratory, where a complete broadcasting and receiving unit
had been installed; two were conducted over Station WEEI, Boston, The
number of student judges in the laboratory experiments ranged from
thirty-two to eighty-five; in the WEEI experiments listeners were asked
to send in their judgments. Two hundred eighty-five replies were received
in the two radio experiments (190 in one and 95 in the other). The total
number of judges in the eight experiments was 587, Objective criteria were
gathered from observation and from measurement. The criterion for in-
troversion-extroversion was the Heidbreder Personal Traits Rating Scale;
for dominant values, the AUport and Vernon Study of Values; for ascen-
dance-submission, the AUport AS Reaction Study. In each of the experi-
ments the listeners were asked to check one of several alternative descriptive
words, AUport and Cantril found that no single feature of personality was
always matched correctly with voice, nor was any individual voice always
correctly judged in every respect. They also found that "the uniformity of
opinion regarding the personalities of the speakers was somewhat in excess
of the accuracy of such opinion, showing the importance of the phenomenon
of stereotyped judgment" (3, 109),
Taylor (10) used a somewhat different technique. He had each of
twenty speakers read a three-minute newspaper editorial; these readings
were recorded on a Speakophone aluminum disk. The twenty records were
then played to twenty auditors, who rated each subject whose voice they
heard. Taylor found a high degree of social agreement in judging per-
sonality traits from voice. There was a tendency for the auditors to be
most consistent in their judgments when they agreed least with the speakers'
self-ratings of their personality traits.
In some experimental studies no mechanical aids have been used. Bona-
ventura (2), for example, attempted to determine expression of the per-
sonality in the speaking voices and in the photographs of six individuals.
A large number of judges listened to the "hidden voices" and matched each
voice with a photograph of the speaker. Each auditor made extensive
notes on the rhythm, pitch, tempo, and strength of voice, and described
his procedure in making a judgment, Bonaventura chose the speakers ac-
146 CHARACTER AND PERSONALITY
cording to the Kretschmerian morphological types (pyknic, leptosome, and
athletic), and the results indicated that these types were accurately judged
in the order here named.
Stagner (9) required his subjects (twenty-five students in a social
psychology class) to check on an eleven-point graphic rating scale certain
specific vocal characteristics (voice intensity, flow of speech, poise, clear-
ness) and the traits "nervousness" and "aggressiveness." Ten students,
previously tested on the Bernreuter Personality Inventory and the Wisconsin
Scale of Personality Traits, read typewritten excerpts from a book on per-
sonality. Stagner says that his results "clearly demonstrate the use of
specific vocal cues in the process of making judgments ahout personality''
(Italics his). There was no convenient way for Stagner to determine the
reliability of the ratings.
The authors know of but one study in which an attempt was made to
determine to what extent listeners are able to judge Spranger personality
types by the use of electrical voice transmission. In half of the Allport
and Cantril experiment on dominant values, referred to previously, the results
were positive; the reliabilities of the percentages of correct judgments were
statistically significant. "In experiment I two of the speakers were high
in both aesthetic and religious interests (as measured by the Study of
Values), and were often confused with each other. The positive result of
this experiment is due therefore to the fact that these two speakers were
scarcely mistaken for the third, whose voice clearly betrayed his political
and economic interests" (I, 46).

METHOD

More modern voice transmission apparatus was available wben

tbe present laboratory study was made tban wben Allport and Can-
tril conducted tbeir well-controlled experiment on tbe relationship
between voice and dominant values. Tbe apparatus used in the
Harvard Psycbological Laboratory was tbe best obtainable at tbe
time tbe experiment was performed, but public address systems bave
been markedly improved in recent years, providing tbe experimenter
witb more refined equipment. Tbe Allport and Cantril apparatus in-
cluded a micropbone, a control panel, a loud-speaker, and a system
of signal devices. Tbe loud-speaker was tbe 540 AW cone-sbaped
type (Western Electric). A two-button carbon micropbone was
used.
Witb tbe aid of modern public address equipment, bigb fidelity
voice transmission was acbieved in tbe present experiment. In fact,
tbe mecbanical transmission of tbe voice probably did not materially
reduce tbe ability of tbe listeners to judge correctly. Allport and
Cantril, in several control presentations of voice, in wbicb the speak-
ers read a passage from bebind a screen in tbe same room in wbicb
tbe judges were seated, found an average difference of approxi-
 JUDGMENT OF SPRANGER PERSONALITY TYPES 147

mately 7 per cent in favor of the normal voice. They say: "Even
though the broadcaster can be assured that most people readily
adapt to the figure-ground situation which the radio creates, our
experiments do show a slight loss in the accuracy of matching. On
the average, the natural voice is somewhat more revealing of per-
sonal qualities than is the radio voice. The loss represents perhaps
only such imperfections in transmission which mechanical improve-
ments in the radio may in time remove" (3, 120).
In the present experiment an 8-watt Knight amplifier, located
in a sound-resistant room, was used. The amplifier was well
grounded and the current source was filtered; it had an input for
microphone and phonograph and an output for dynamic speaker and
head phone. A qualified radio technician acted as monitor, operating
the control panel, adjusting voice volume, and playing phonograph
records when transcribed music was desired. A crystal microphone
(dual diaphragm) was employed. It had an output of 65 db., and
was attached to an adjustable floor stand. The microphone was
located in a sound-resistant room separate from the control room,
and was connected to the amplifier by a shielded cable and through
a shielded wall connection. The technician in the control room
signaled to the speakers in the microphone room by means of a
system of colored lights.
The audition room seated sixty people, and was favorably located
so as to make it comparatively free from distracting sounds. The
position of the dynamic speaker was determined after several acous-
tic tests had been made. The speaker was an 8-inch Jensen (2,000-
Ohm field), housed in a wall case and connected to the amplifier
in the control room by a built-in cable. An audition director was in
charge of the experiment. A signal button in the audition room
was connected with a buzzer in the control room, making it possible
for the audition director to signal the monitor regarding the volume
of sound, etc.
The announcer for the experiment prepared his script in minute
detail, using regular broadcasting continuity methods. The script
was carefully timed before the experiment began. Duplicate copies
were placed in the hands of all the speakers, the monitor, and the
audition director. The announcer spoke into one side of the micro-
phone, the individual speakers into the other side. During the ex-
periment the announcer kept close watch on the monitor's signal
light, and, in turn, signaled to the various speakers when they were
to read their materials.
148 CHARACTER AND PERSONALITY

Twelve individual speakers (eight male and four female col-

lege students) were used. Two speakers represented eacb of tbe
six Spranger value types, as measured by tbe Allport and Vernon
Study of Values. Eacb speaker scored over 40 on tbe Spranger
type tbat be, or sbe, represented, and not more tban 25 on the otber
Spranger types. Eacb speaker read over tbe public address system
a paragrapb wbicb took approximately thirty seconds to read. In
an effort to prevent boredom on tbe part of tbe listeners, four differ-
ent selections were cbosen from a current magazine article; tbe
selections may, tberefore, be assumed to be comparable, since tbey
were excerpts from tbe magazine writing of a single author. The
particular paragraph which each individual speaker read was deter-
mined in a purely random fashion. The speakers practiced reading
into the microphone before the experiment in order to decrease the
possibility of "mike fright" and to equate voices for volume. They
were not told the purpose of the experiment; artificial changes in
voice were thus avoided. During the practice periods the monitor
marked on his script the proper amplifier volume for each speaker.
During the experiment the audition director also checked on volume
of the speakers and signaled to the control room technician if the
volume was not properly equated.
The listeners were forty-five advanced students in a class in
applied psychology (twenty-one men and twenty-four women).
They were not asked to sign their names on their rating sheets.
In order to produce a desirable setting, the experiment started with
transcribed music. The listeners were asked by the announcer to
write on a sheet of paper, previously supplied by the audition director,
the names of the six Spranger types. The following brief type
descriptions, patterned after those presented by Vernon and Allport
(11), were given, each description being read twice:
1. The Theoretic
Is essentially a thinker. He generalizes, reasons, systematizes. His
reactions are essentially objective rather than subjective—cognitive rather
than affective. He obtains his thrills from solving problems, formulating
theories, and systematizing knowledge. Here we find the great teachers,
scholars, researchers.
2. The Economic
Stresses practical values, utilities, life-preserving goods and forces. He
may be a producer or a consumer, a marketer of goods or a "credit man."
He thinks in terms of agriculture, trade, industry, money. He is likely to
ask: "Of what use is such and such a fact or discovery?" He is interested
primarily in the useful and the prodvictive.
 JUDGMENT OF SPRANGER PERSONALITY TYPES 149

3. The Aesthetic
Is essentially the "feeling" type. He lives keenly in his emotions. He
usually takes the subjective viewpoint. His values are to be found chiefly
in harmony, beauty, grace, and symmetry. His impressions are enjoyed for
their own sake. He may be a creator of art or one who merely appreciates
art. He responds to the sensuous, the rhythmic, the imaginative. His aims
are self-enjoyment, self-realization, self-fulfilment.
4. The Social
Is characterized by love—love of people. Here we may find philanthropic
love at its best. He is kind, sympathetic, and considerate for others. He is
unselfish and ready to serve those in need.
5. The Political
He has a wish for dominance, for power, and for control over others.
He is always fighting, competing, forcing. He has a very strong drive
toward self-assertion. He likes to conquer things, situations, people.
6. The Religious
Is partially exemplified in the mystic who finds God everywhere. Is
characterized by self-denial and meditation.
Following tbese brief descriptions of the various Spranger per-
sonality types, the listeners were asked to write the numbers 1 to
12 in a column on the answer sheet, leaving enough space after each
number to write the type name. The announcer always gave the
number of each voice before and after tbe reading of each selection.
Also, after each voice he paused several seconds so that listeners
could write their answers. Transcribed music was played during
this pause; there was never any silence. As stated previously, four
different paragraphs of magazine material were read, and, in order
to avoid monotony, these readings were presented in a mixed order.
The order of presentation of the six personality types was : aesthetic,
social, theoretic, political, religious, economic, religious, social, aes-
thetic, theoretic, economic, political. Precaution was taken to insure
that the speakers would be unknown to the listeners; all participants
were warned not to tell anyone that they were appearing in the ex-
periment. Although tbe authors cannot be absolutely positive tbat
eacb speaker was not known by any listener, tbere was considerable
evidence that the voices were not recognized.

RESULTS AND DISCUSSION

The percentage of listeners rating each of the twelve speakers
over the public address system in each of the six Spranger value
types is shown in Table 1. The difference between the percentages
150 CHARACTER AND PERSONALITY

^ Ih * o
NO j Q O

o
o
toIh tN
-Ih rs o
- Ih 00
1®
II- -Ih Ih ° Ih
00 o o
bT CS
o ON

o o
S
-Ih O = ||o iO

" Ih h o to NO
00
fs.
VO

o o
y
h H to
Ih to
Ih ° II -
1 bT ON

I = "1 - o
- ° 00
to
ON

Ih Ih o ON QO j 00

a.
» i II"" CS
ON o o o 2 2 o
a a II s tN
CS
1
^
ON
^
ON ON

a
o
hi"* - II"" -Ih i Ih a
Ih -I = o
Religious

CS CS

I
CO
^ il — o tN o ts 00 ON
CS
to ON
CS
Ih to 1
o tN

u o o o o
Ih
vo 00 to
S
IO ON 00 ON ON ts 00
cs 2 o o
ts tN to tN lO

o 11 ^ t o
'g -Ih -i- O

h ^Ih to
^ II - tN to
esthe tic

to
RAT

-Ih h ,h -Ih
00 00
- tN

o o o o
-Ih o
LO O 00 ON CO

?
ERCEI

E?ih ON
cs
h- CN
CS
O
h CS cs •-*•

cs
to
h c
Socia

•^ II t o 00 o
h
00 o
vo to - 12 cs
lO to

o
S !S
-Ih 00 5
1" 2 «o O o
.l.lh o

h o I cs
ON vo CO o ON
H ts tN 2 ON
Theoretic

cs

s: ON
CS
IO
cs ft
s1* 00 NO o 00 ON
CS
to oo to

s ts
to
to
ON
cs
vo o 00
2 cs
(N
CS a ON
CS
O o >o
A-V*
Score

o CS
Speaking
Order of

o CO 00 cs - ON IO NO cs

s
Sex

S
- s
Speaker

d
Mean

cd
Mean
Mean

1
Mean

6 JS c
o
1
Actual Spr anger
Type

Theoretic

Economic
Aesthetic

Religious

Political
Social
 JUDGMENT OF SPRANGER PERSONALITY TYPES 151

—16
—17

+30
—15

+46
c

+61
—IS
—11
—13

l—\6
+ 11

-IS
1 1
O\
•K o 1 tn
T
1 1 1 II o+ -t- 1 1 1 1 1 1 1 1+ + +
tn o cs
7 II _ o CS Ov

1 1 + + 1 1 1 1 1 II 1 1 1 + + +

T
Ov [ O o 2 II t VO .o
^ 1
+31 +29 +30

u 1 1
u o
+ 11 1 1tn1 1 11 + 1 -I-l 1
ss AND CHAN

zlL ;IT 1 T

—17
+12 +21
ON

—17 —17
—10
tn

—13

iz-i-
+22

R
[1. *
T 7 r rs cs
s1 cs 1

+21

z -1-
7
oi-l-
—12

<z
-\\ 1 1
tn tn tn 1 u^ m lo 1 00 CS
^ 1 CTv vO ^.
1 1 •j' 1 1 l' 1 1 + + + 1 + 1+ + + 1 1 1 1
.9 cs <s 1
tn wi ON Ov
_j_
"^
elif

1 1 'l 1 1 1 1 + + 1+ 1 + 1 + + 1 1 1
H a 1 tn
s 7 1 1 1 1 1 + +J + + + + + 1 1
M
r.. If - c^ 1
^, u
H ,, 1 7 1 1 +1 + + 1 + 1 1 1 7 7
cq
1 rn 7
tn
- 1 tn o VO 1 a a Ov o CO
S: Ov
7 7
tn tn
-FEREtICE

1 1 1 1 1 I -1- + + + + 11 1 1 1 1 7
r

—14
1+29

—12
zi-l-
+40
+26
+10 —17
—17

tn -j-
1 1 1 1 + 1
Q m

—17
—10

00
—17

NJ

? 1;
+27

+52

OO VO
H
1 1 1 7 1 1 7
S
ial

VO tn \ o 00 f. c cs tn 00 CS o
PERCE

=8 + 1 + + 1+ 1 1 1 1 + 1 1 1 1
tn r>l
s I ^7
—17
—12

s
+64
+40

+19

H
1 1 1 + 1 1 1 1
1° 00 VO VO ... f . 00 <M 00

1 ++ + 1 1 1 + 1 + 1 -1- 1 1 1

-I?
o CTv
1
+12

00
+11

—13

s 1 1 1 1 1 + 1
H VO (S Ov CS cs o o 1^ (S

1 ++ + • 1 + + + 1 1+ + 1 1 1 1
rated
ntage
ces of

r„
s g. So VO VO VO VO So VO VO VO NO VO NO VO »o VD NO VO
VO VO VO VO VO NO VO VO VO VO VO NO NO VO \o VO
• '

u a
ra
1
ean

ean

0
ean
ean

c
ean

.id u
K 6
2 2
anger

ao.
u
-iH •rj D
£
the

5no
ctu

'So
IS
Po

<:
Re

< %
152 CHARACTER AND PERSONALITY

of listeners rating each of the speakers in each of the Spranger

types and chance percentage (16.67) is shown in Table 2. It will
be noted from Table 1 that the mean percentages of male listeners
are superior to those of female listeners in correctly identifying the
theoretic, social, economic, and political value types, although the
sex differences are, in most cases, slight. The sexes are tied in accu-
racy of judging the aesthetic and religious types.
Tables 1 and 2 show that the two representatives of the theoretic
type are identified 11 per cent better than chance by the total number
of listeners. Speaker "Fr" (male, theoretic) is judged to be the
social type by 44 per cent of the listeners, while Speaker "Bl" (male,
theoretic) is judged to be the economic type by 47 per cent. The
mean percentage of total listeners' ratings for the two theoretic
type voices is lowest for the aesthetic, religious, and political types;
it is highest for the economic, theoretic, and social types.
The two representatives of the social type are identified 23 per
cent better than chance by the total number of listeners. Speaker
"Co" (male, social) is judged to be the political type by 60 per
cent of the listeners. Speaker "Sh" (female, social) is judged to
be the social type by 69 per cent, and not a single listener judges
her to be the economic or political type. The mean percentage of
total listeners' ratings for the two social type voices is lowest for
the religious, economic, aesthetic, and theoretic types; it is highest
for the social and political types.
The two representatives of the aesthetic type are identified 29
per cent better than chance by the total number of listeners. Speaker
"Fl" (female, aesthetic) is judged to be the aesthetic type by 44
per cent of the total number of listeners, while Speaker "Ru" (male,
aesthetic) is judged to be the aesthetic type by 47 per cent. The
mean percentage of total listeners' ratings for the two aesthetic type
voices is lowest for the political, economic, social, and theoretic
type; it is highest for the aesthetic and religious types.
The two representatives of the religious type are identified 12
per cent better than chance by the total number of listeners. Speaker
"Ho" (female, religious) is judged to be the aesthetic type by 47
per cent of the total number of listeners, while not a single listener
judges her to be the economic or political type. Speaker "Ma"
(male, religious) is judged to be the economic type by 33 per cent
of the total number of listeners. The mean percentage of total
listeners' ratings for the two religious type voices is lowest for the
 JUDGMENT OF SPRANGER PERSONALITY TYPES 153
political, theoretic, and social types; it is highest for the aesthetic,
religious, and economic types.
The two representatives of the economic type are identified only
3 per cent better than chance by the total number of listeners.
Speaker "Sn" (male, economic) is judged to be the economic type
by 38 per cent of the total number of listeners, while Speaker "Ra"
(female, economic) is judged to be the religious type by 49 per cent.
The mean percentage of total listeners' ratings for the two economic
type voices is lowest for the political, aesthetic, and social types;
it is highest for the religious, economic, and theoretic types.
The two representatives of the political type are identified 46
per cent better than chance by the total number of listeners. Speaker
"Ha" (male, political) is judged to be the political type by 47 per
cent of the total number of listeners, while Speaker "Os" (male,
political) is judged to be the political type by 78 per cent. The mean
percentage of total listeners' ratings for the two political type voices
is lowest for the aesthetic, social, and religious types; it is highest
for the political, economic, and theoretic types.
Table 2 definitely indicates that certain voices are stereotypes;
they impress listeners as being the voices of persons who could be
classified in a particular Spranger value type. Speaker "Sh" ap-
pears to be a social stereotype; Speaker "BI," economic; Speakers
"Co," "Ha," and "Os," political; Speakers "Ru" and "Ho," aesthe-
tic; Speaker "Ra," religious.
In order to discover the relationship between the listener's rat-
ings of the Spranger personality types of the speakers and the
actual personality types, as determined by the Allport and Vernon
Study of Values, the Contingency Method, developed by Pearson,
was used (5, 198). Table 3 shows the coefficients of contingency
TABLE 3
COEFFICIENT OF CONTINGENCY BETWEEN LISTENERS' RATINGS OF SPRANGER
PERSONALITY TYPES AND ALLPORT-VERNON TEST SCORES OF TWELVE
SPEAKERS OVER PUBLIC ADDRESS SYSTEM

Listeners N C P.E.*

Males 21 .65 .019

Females. 24 .71 .017

Total .... 45 .52 .016

*P. E. is computed from the formula:

PE _ 0.6745
4 , 102)
154 CHARACTER AND PERSONALITY

between the listeners' (male, female, and total) ratings of Spranger

types and the Allport and Vernon test scores. Inasmuch as the
maximum C from a 6 x 6-fold table is, according to Yule (12), .913,
and the C for the total listeners is .52 ± .016, one may conclude that
there is a medium positive correlation between the listeners' ratings
of Spranger personality types and the actual types themselves.
The results of this experiment are significant for radio broad-
casting, inasmuch as the fidelity of voice transmission over the
public address system, previously described, is equivalent to that
achieved over a radio. The listeners in this study, although all
were college students, were not selected. In radio experiments on
the relationship between voice and personality traits, only a small
percentage of the listeners answer. In several such experiments the
authors found that it was very difficult to get radio listeners to
respond, although there was evidence of widespread listener interest.
In the present study, however, judgments were made by all auditors.
From a practical standpoint, the results of the experiment indi-
cate that certain voices seem to portray a particular personality type
to a rather large percentage of listeners. Practical use might be
made of stereotyped voices in the broadcasting of dramatic materials.
The casting of characters, frequently done ineffectively, could thus
be greatly improved. For example, if, in broadcasting a dramatic
sketch, it is important that a certain speaker impress his audience
as being a religious person, then such a speaker should be chosen
because of his stereotyped "religious voice"; actually, of course, he,
or she, may be the economic or the political type rather than the
religious. The actual type, however, is not important, since in the
field of radio broadcasting, as in the field of dramatics, the chief
objective is to create a certain type and to produce a certain effect.

SUMMARY

1. In judging Spranger value types from the voices of speakers,

selected on the basis of their scores on the Allport and Vernon
Study of Values, and heard over a public address system, male
listeners are slightly superior to female listeners, but the sex differ-
ences are negligible.
2. The mean percentages of total listeners' ratings for two
representatives of each of six Spranger types (theoretic, social,
aesthetic, religious, economic, political) are superior to chance.
3. The Spranger types which are judged most accurately (in
terms of mean percentages superior to chance) are: political, 46 per
 JUDGMENT OF SPRANGER PERSONALITY TYPES 155

cent; aesthetic, 29 per cent; social, 23 per cent. The types which
are judged least accurately are: economic, 3 per cent; theoretic, 11
per cent; religious, 12 per cent.
4. There is a rather significant medium positive correlation be-
tween the listeners' ratings of Spranger value types and the actual
types themselves.
5. Certain voices are stereotypes; they definitely impress lis-
teners as being the voices of persons who might be classified in a
particular Spranger value type. Other voices are not stereotypes;
there is little consistency in judging them.
6. Practical use should be made of stereotyped voices in the
broadcasting of dramatic materials over the radio. The casting
of characters could thus be improved, and a more desirable effect
could be produced on the audience.
REFERENCES
1. ALLPORT, G. W . . & CANTRIL, H . Judging personality from voice. /. Soc.
Psychol., 1934, S, 37-55.
2. BONAVENTURA, M . Ausdruck der Personlichkeit in der Sprechstimme und
im Photogramm. Arch. ges. Psychol., 1935, 94, 501-570.
3. GANTRIL, H., & ALLPORT, G. W . The psychology of radio. New York:
Harper, 1935. Pp. 276.
4. DAVENPORT, G. B.. & ERAS, M . P. Statistical methods in biology, medicine
and psychology. New York: John Wiley & Sons, 1936. Pp. 216.
5. GARSETT, IH. E . Statistics in psychology and education. New York: Long-
mans, Green & Go., 1926. Pp. 317.
6. HERZOG, H . Stimme und Personlichkeit. Zsch. f. Psychol., 1933, 130, 300-379.
7. PEAR, T . H . Voice and personality. New York: John Wiley & Sons, 1931.
Pp. 247.
8. SPRANGER, E . Types of men. (Trans, from the 5th German ed. by P. J.
W. Pigas, 1928.) Halle (Saale) : Niemeyer. Pp. 402.
9. STAGNER, R. Judgments of voice and personality. / . Educ. Psychol., 1936,
27, 272-277.
10. TAYLOR, H . G. Social agreement on personality traits as judged from speech.
/. Soc. Psychol, 1934, S, 244-248.
11. VERNON, P. E., & ALLPORT, G. W . A test for personal values. / . Abnorm.
(Soc.) Psychol., 1931, 26, 231-248.
12. YULE, G. U . An introduction to the theory of statistics. London: Griffin
& Go., 1919. Pp. 434.