-'_:
CALIFORNIA STATE UNIVERSITY, NORTHRIDGE
EFFECTS OF RACE, SOCIO-ECONOMIC STATUS, AND
'' TYPE OF CRIME ON ATTRIBUTIONS OF GUILT
A thesis submitted in partial satisfaction
of the requirements for the degree
of Master of Arts in
Psychology
by
Preston A. Cox
.-/
January, 1980
The Thesis of Preston A. Cox is approved:
Dr.
Dr. Michele A. Wittig ~
Dr. Paul Skolnick
California State University, Northridge
ACKNOWLEDGMENTS
I would like to thank the members of my committee, Dr.
Paul Skolnick, Dr. Michele A. Wittig, and Dr. Jerry I. Shaw.
I would also like to thank Dr. Sarnual Pinneau for his
assistance with the choice of the proper statilliltical analysis.
Last but far from least, I would also like to thank my
wife, Cecilia, and my children, Valerie and Aubrey, for their
patience, love, and support throughout my academic career.
iii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS
iii
LIST OF TABLES.
iv
ABSTRACT . • • • • .
....
INTRODUCTION. .
..
14
METHOD • • •
Subjects. • • • .
Procedure. • • • • .
RESULTS ••
...
....
.......
.
.
c.
D.
E.
21
37
39
APPENDICES.
A.
B.
14
14
25
DISCUSSION
REFERENCES
1
Simulated Jury Study
Summary of Police Files on
Defendant Background.
Summary of Evidence in the Case of
Thomas Longstreet .
Questions -First Set • .
.
Questions - Second Set .
.....
.
. ........
.......
iv
40
41
43
44
LIST OF TABLES
1.
Mean Scores of Unreduced Group Data
31
2.
Multivariate Analysis of Variance of
Unreduced Group Data . .
32
Univariate Analysis of Variance of
Unreduced Group Data Race Dimension . . • . .
33
4.
Mean Scores of Reduced Group Data
34
5.
Multivariate Analysis of Variance of
Reduced Group Data . . .
35
Univariate Analysis of Variance of
Reduced Group Data Race Dimension . . . . .
36
3.
6.
v
ABSTRACT
EFFECTS OF RACE, SOCIO-ECONOMIC STATUS, AND
TYPE OF CRIIVIE ON ATTRIBUTIONS OF GUILT
by
Preston A. Cox
Master of Arts in Psychology
Reports of the experimental treatment of race in a simulated jury trial setting appear infrequently in the literature.
A
rare example is afforded by Gleason and Harris (1975) who found
that, for the crime of murder, defendant socio-economic status,
not defendant race (in this case,
bla~k
affected judgments made by subjects.
or Caucasian), significantly
Kalven and Zeisel (1971 ),
however, present evidence suggesting that viability of race as a cue
for guilt may depend upon the type of crime committed.
For example,
they report that for certain crimes, such as narcotics violations,
blacks represented 60o/o of the defendants on trial.
of murder, 43o/o of the defendants were black.
For the crime
Finally, for white
collar crime and notably for drunken driving, blacks accounted for
Vi
a low number (7o/o) of the defendants on trial.
If these findings
approximate reality, and if the reality is discernible to the population of potential jurors (or subjects), then it is not surprising
that Gleason and Harris (1975) found no race related effects.
Murder is not clearly a race specific crime.
Another crime may
have revealed the race specificity effect.
Because the literature lacks experimental evidence supporting an interaction effect between defendant race and type of
crime, the present study tested for such an effect.
Subjects read
a fictional summary of the trial of either a black or an AmericanIndian defendant charged with either possession of narcotics or
drunken driving.
Subjects rated guilt and severity of sentencing
on a seven-point scale, and assigned a specific sentence.
Based
on the findings of Kalven and Zeisel (1971) described above, and
on the image of the "drunken Indian11 appearing frequently in
popular folklore of the American West, it was predicted that the
American Indian would be favored when the crime was possession
of narcotics, while the black would be favored when charged with
drunken driving.
The present investigation also pursued a clue given by
Gleason and Harris (1975) that the effect of defendant socio-economic status· can be explained by theories used to explain the
effects of defendant physical attractiveness.
vii
The cognitive theory
of physical attractiveness described by Sigall and Ostrove (1975)
was chosen and modified to fit the present case.
The theory
states that subjects make judgments based on a belief of the
likelihood of a defendant to commit future violations, and that
subjects will usually think a high socio-economic status defendant
less likely than a low socio-economic status defendant to commit
future violations.
Therefore, subjects will judge the high socio-
economic status defendant more favorably.
An exception occurs
when the high socio-economic status is an asset to the defendant.
In this case, the defendant will be perceived by the subject as
more likely to be a repeat offender and will be judged at least as
harshly as a low socio-economic status defendant.
In the present study, the defendant was described as being
of either high or low socio-economic status.
Since high socio-
economic status was not clearly an asset to the defendant, it was
predicted that the high socio-economic status defendant would be
judged more favorably on the three dependent variables mentioned
above.
However, to test the cognitive theory, subjects were also
asked to assign probabilities to the likelihood that the defendant
committed the crime and to the likelihood that he might commit
the crime again.
It was predicted that the high socio-economic
status defendant would be assigned lower probabilities on both
items.
viii
The results of the multivariate analysis of variance of the
data failed to support any of the hypotheses.
effect due to race was found (p
<. 03,
Instead a main
for the multivariate analy-
sis) such that subjects recommended harsher sentences for the
American- Indian than for the black defendant (p
univariate analysis).
<. 05, for the
A secondary analysis using data from only
the sixty-one sibjects who correctly identified the race and socioeconomic status of the defendant strengthened these results (p <. 02
and p<. 006, respectively).
It was concluded that the discre-
pancies between the present study and the Gleason and Harris
study (1975) sustain the question of a race by type of crime interaction.
For example, Gleason and Harris (1975) found no race
effect when using a victim crime (murder), while the present
study did find a race effect when using victimless crimes.
The
possibility of an interaction between socio-economic status and
type of crime is also raised by both studies considered together.
Finally, the cognitive explanation for the effect of sicio-economic
status could not be assessed be cause of the failure to obtain an
effect due to .socio-economic status.
ix
Attribution of responsibility has formed the subject of
many reports in the psychology literature.
A popular paradigm
used in many experiments is a simulated jury trial in which factors influencing attirbutions of responsibility (in the form of
attributions of guilt) can be examined.
Some of the factors
studies include sex of the defendant (e. g., Gleason & Harris,
1975; McGlynn, Megas, &
Benson~
1972; Stephan, 1974); amount
of defendant suffering (e. g., Sigall & Landy, 1972 ); vilification of
victim (Shaw, 1972); socio-economic status of defendant (e. g.,
Gleason & Harris, 1975, 1976); attractiveness of defendant (e. g.,
Landy &
Aronson~
1969; Nemeth & Sosis, 1973; Reynolds & San-
ders, 1975; Sigall & Landy, 1972; Sigall & Ostrove, 1975); age of
defendant (Reynolds & Sanders, 1975); group decisions versus
individual decisions (e. g., Gleason & Harris, 1976); perceived
similarity between juror and defendant (e. g., Klinger, Albaum, &
Heatherington, 1964); and the juror's attitude toward the death
penalty (Gleason & Harris, 1975, 1976; Jurow, 1971 ).
However, one factor that has not been reported on extensively is race of the defendant. This is somewhat surprising in
view of the social importance of racism.
Joseph A. Califano,
the Secretary of the Department of Health, Education, and Wel1
fare, recently described racism as the most difficult domestic
issue facing the country ("One More Plan to End Fraud in Welfare," 1978).
He said that in spite of the progress of the last
fifteen years,
"we are still plagued in almost every aspect of our
society with issues of race."
Jury trial data collected by Kalven and· Zeisel (1971) suggest the opportunity for potential harm afforded racism.
The
researchers observed that although blacks constitute only one-
•
tenth of the total population, they account for almost one-fourth
of the defendants in jury trials.
If racism does influence attri-
butions, then the courtroom is a good place to study the influence
given the disproportionate number of blacks appearing in jury
trials.
However, most of the information available in the literature
pertaining to the effect of racism in the courtroom consists of
analyses of police and court records.
For example, Hagan (1975)
found that even though all Canadian judges surveyed were strongly
influenced by offense seriousness in assigning a sentence to an
offender, those judges found to be less concerned with the maintenance of law and order were also influenced by the race of the
defendant, his prior record, and the number of charges against
him.
2
Bullock (1961) evaluated data on prisoners in the Texas
State Penitentiary in Huntsville.
The percentage of prisoners re-
ceiving short sentences (less that ten years) or long sentences
was found to vary according to certain non-racial factors such as
the geographic area of the prisoner's origin, whether the prisoner
came from a large., or small city, the type of plea entered (guilty
or not guilty), and the type of offense committed.
Recognizing
that blacks may be exposed to some of these factors to a different extent from whites, Bullock analyzed the data, controlling
for the non-racial factors, and found that in many instances a
significant effect due to race still existed.
For example, of those
prisoners from East Texas Counties who had plead guilty to a
charge of burglary, more whites than blacks received short sentences (p < • 02 ).
The literatu;r-e on the experimental analysis of racism in
the courtroom is scant.
McGlynn, Megas, and Benson (1976)
tested the effect of defendant sex and race on judgments of guilt
by experimental jurors when the defendant was making a plea of
insanity.
The researchers found that black males were judged
guilty more often than white males or females (p< . 06).
The
only significant effect involving race was the length of sentence
imposed by those jurors voting guilty.
White males received
harsher sentences than black males (p< . 03).
3
However, an im-
portant confound introduced in this study concerns the implied
socio-economic status of the defendants.
The blacks were de-
scribed as having committed their crimes in a poor class area
of the city.
The whites committed their crimes in a middle class
area.
Gleason and Harris (1975) studied both race and socioeconomic status.
They found no significant effects due to race
but did find that socio-economic status of the defendant significantly affected judgments made by simulated jurors of guilt and
severity of sentencing.
(This study will be examined more close-
ly below.)
However, it may also be that race is not a particularly
good cue for judging guilt when the defendant has been charged
with murder, which is the type of crime used in the two preceding studies.
Kalven and Zeisel (1971) report that of the mur-
der cases they reviewed, 43o/o of the defendants were black.
Jurors could develop a
11
feeling" for this statistic in the course of
encountering accounts of murder cases reported in the various
mass media.
Since the race of the suspect is often mentioned
in such accounts, it is conceivable that jurors who attend to the
mass media could learn that murder is not race specific, that
blacks and whites go on trial for murder in approximately equal
proportions.
In that event, race would not be a reliable cue for
4
determining the guilt or innocence of a defendant in a murder
trial.
If a particular crime is race specific, then defendant race
could become a viable cue for guilt.
That certain crimes do
seem to be race specific is suggested by evidence presented by
Kalven and Zeisel (1971 ).
Blacks accounted for a high number of
defendants on trial in cases involving the Mann Act (transportation
of women across state lines for immoral purposes, 67o/o of the
defendants were black), prostitution (63o/o), and narcotics violations (60o/o).
Conversely, a low number of blacks were involved
in white collar crimes (5o/o to 8o/o were black) and notably in
drunken driving ( 7%).
If a juror possesses the belief that certain crimes are
race specific, then it would not be unreasonable to predict that
defendant race would interact with the type of crime to affect
judgments of guilt made by jurors.
While the literature lacks
experimental evidence of such an interaction, studies do exist
assessing the interaction of type of crime with other defendant
characteristics.
Richey and Fichter (1969) studied the effect of type of
crime and defendant sex on moralistic judgments.
The types of
crime used were personal misconduct (possession of marijuana
cigarettes) and dishonesty (cheating on an exam).
5
The defendant
was either a male or female student.
Subjects were asked to
express an opinion about the character of a student who would
engage in such behavior if it were proven that the student did.
The subjects were also asked hqxv _severely the student should be
punished if the charges were proven.
teraction effects were found.
No significant main or in-
However, perhaps there is no
reason to believe that women are involved more often than men
in either crimes of personal misconduct or crimes of dishonesty.
Sigall and Ostrove (1975) presented subjects with either a
picture of an attractive defendant, a picture of an unattractive
defendant, or no picture at all (a control condition).
The type of
crime committed was either attractiveness related (swindling) or
attractiveness unrelated (burglary).
The researchers chose
swindling so that the defendant. in this case a female, could use
her physical attractiveness to help deceive a middle-aged bachelor into investing $2,200 in a nonexistent corporation.
ing a sentence, an interaction effect was found (p
In assign-
. 025).
When
the control condition was deleted, the effect was strengthened
(p
. 01 ).
The attractive swindler and the unattractive burglar
received the heaviest sentences (5. 45 and 5. 20 years average,
respectively).
There was no .significant difference between the
unattractive swindler and burglar and their respective controls.
The heaviest and lightest sentences were received by the
6
attractive swindler and the attractive burglar, respectively.
One of the goals of the present study is to attempt to
demonstrate an interaction effect between race of the defendant
and the type of crime on subjects' assessments of guilt and
assignments of punishment.
be "black" and
The race dimensions to be used will
11
American-Indian."
The type of crime dimension
to be used will be "drunken driving" and "possession of narcotics."
The "black'' dimension is chosen because of Kalven and
Zeisel's (1971) finding that blacks were represented in 60o/o of the
narcotics cases studied.
The "American-Indian" dimension is
chosen because of the image of the "drunken Indian" which appears
throughout much of the popular folklore of the American West.
Also, since blacks appeared in only 7o/o of the drunken driving
cases studied by Kalven and Zeisel (1971), it is expected that
subjects' assessments of blacks and of American-Indians in the
drunken driving condition will differ significantly.
Specifically,
it is expected that in the possession of narcotics condition, blacks
will be found more guilty than American- Indians, will be re commended for harsher sentences, and will actually be given longer
jail sentences.
In the drunken driving condition, however, it is
expected that the American-Indian will score higher on the three
mentioned dependent variables.
As stated previously, Gleason and Harris (1975) found no
7
significant effects due to race but did find significant effects due
to socio-economic status of the defendant on judgments of guilt
(p< . 0005), severity of recommended sentence (p< . 01 ).. and length
of assigned sentence (significance level not reported), with high
socio-economic status defendants being favored on all three dimensions.
No explanation is given for the effect of socio-ec-
onomic status, although Gleason and Harris (1976) suggest a
parallel between socio-economic status and physical attractiveness.
If a parallel does exist between socio-economic status and
physical attractiveness, then it would not be unreasonable to expect that explanations for the effects of physical attractiveness
would also explain the effects of socio-economic status.
There
are at least three popular explanations to be found in the literature for the effect of physical attractiveness.
Sigall and Ostrove (1975) review a reinforcement model of
attraction.
According to this model, physical attractiveness
would have positive reinforcement value and would lead to relatively more positive affective responses toward a physically
attractive person.
Research has indicated (e. g., Landy and
Aronson, 196 9) that liking for a defendant increases leniency.
It would thus be expected that a physically attractive and, therefore, better liked defendant would be punished less than an unattractive defendant.
8
A second explanation, offered by Dion, Berscheid, and
Walster (1972 ). proposes the existence of a physical attractiveness stereotype in which it is expected that physical attractiveness would be correlated with certain inward characteristics.
If
a person is attractive, he will be expected to be more virtuous.
There would be more resistance to the idea of a physically attractive person being capable
o~
committing a crime. thus facili-
tating more leniency toward him because he has a greater _chance
of becoming a more successful person.
A physically unattractive
person is more likely to be capable of committing a crime and
must be punished more severely to compensate for that capability.
However, Sigall and Ostrove (1975) have identified a
situation in which predictions made by the above models no longer
seem reasonable.
They propose a cognitive model which assumes
that a subject has a rational basis for his responses.
If a sub-
ject thinks a defendant is likely to commit future violations, then
it is reasonable to deal more harshly with the defendant to reduce
the probability of subsequent violations.
That is, a subject not
only addresses the probability that a defendant committed the
crime in
question~
but also addresses the probability that the
defendant will commit the crime again in the future.
The cogni-
tive mode 1 asserts that an attractive defendant would be thought
less likely than an unattractive defendant to commit future vio9
lations except in those cases where attractiveness is an asset to
the defendant.
In this case, the attractive defendant has an in-
creased likelihood of committing the crime again and would be
punished more harshly, which is the opposite prediction of the
one afforded by the physical attractiveness as reinforcement
model or the physical attractiveness stereotype model.
Sigall and Ostrove (1975) claim to have found support for
the cognitive model in their study described earlier.
In the
burglary condition, the physically attractive defendant received a
lighter sentence than the unattractive defendant (p < . 025).
How-
ever, in the swindle condition, the physically attractive defendant
received a slightly longer sentence than the unattractive defendant.
although the difference failed to achieve significance.
The re-
searchers concluded that physical attractiveness was an asset to
the attractive swindler.
Therefore, it could have been expected
that the attractive defendant might commit the crime again.
It
was necessary to give the attractive swindler a heavy sentence to
decrease the probability of future repetitions of the crime.
With respect to the variable of socio-economic
status~
the
cognitive model would predict the same results as those obtained
by Gleason and Harris (1975).
However, the cognitive model
would also predict that the low socio-economic status defendant
was judged more guilty and sentenced more heavily than the high
10
-
--
-
-
- - - -·-~··
socio-economic status defendant because the low socio-economic
status defendant could be expected to be more likely to commit
the crime again in the future.
Gleason and Harris (1975, 1976)
were not, of course, testing the cognitive model.
However,
Sigall and Ostrove (1975), while claiming to have found support
for the cognitive model, do not appear to have adequately tested
the probability of committing the crime in the future.
The present investigation will attempt to replicate the
findings of Gleason and Harris (1975, 1976) with respect to socioeconomic status of the defendant while also testing the cognitive
model.
Specifically, it is predicted that for both crimes, drunk
driving and possession of narcotics, the high socio-economic
defendant will be judged less guilty, be recommended for a more
lenient sentence, and be assigned a shorter jail term than the
low socio-economic status defendant.
As a further test of the
cognitive model, it is also predicted that subjects will assign
high probabilities to the likelihood that the low socio-economic
status defendant committed the crime in question and to the
likelihood that the defendant will commit the crime in the future.
)
The corresponding probabilities assigned to the high socio-economic status defendant will be low.
Before proceeding, a comment is appropriate concerning
the simulated jury trial paradigm.
11
The present investigation has
been referred to as a simulated jury trial study.
However, that
name has been attached to this study primarily to reinforce a
similarity with other such studies appearant in the literature.
In
reality, the phrase "simulated jury trial" has been somewhat misleading since most such studies only vaguely resemble an actual
jury trial.
Wilson and Donnerstein (1977) offer the following
appropriate commentary:
It should perhaps be noted at this point that the usual
"simulated" jury study is in fact not a simulation at all.
Studies in which subjects are simply given a booklet and
then told to respond to the information fall into the standard laboratory experiment category. In a true jury
simulation, subjects would be participating in as "real 11 a
trial as possible, typically using a courtroom, court proceedings, and normal jury deliberations, rather than the
artificial paper 'and pencil technique. Such jury simulations have indeed been carried out by some researchers,
but the typical procedure is to administer a booklet of
information, sometimes in a classroom. without any of
the details of a true jury setting being in effect. Many
differences exist between the two techniques and researchers should properly describe the procedure they are
using. Use of the word "simulation" may imply realism
in the experiment that does not truly exist. Perhaps when
using the paper and pencil technique, it would be useful to
call the subjects "laboratory" jurors.
Wilson and Donnerstein (1977) then demonstrate that subjects who believed that their verdicts were going to have real
consequences differed significantly in their verdicts from subjects
who knew they were dealing with a hypothetical situation.
One
class of students was asked to assist the teacher in deciding how
to handle an incident of a student stealing and distributing ques-
12
tions to an exam.
The students were told that their verdicts of
guilty or not guilty and their recommended punishments would
determine how the teacher would handle the matter.
Another
class of students were presented with a similar problem, except
that the problem was presented as an ''experiment dealing with
the student judicial process .... ''
The groups of students differed
on a number of dimensions, but most importantly the ''real consequences" students rendered more guilty verdicts and recommended heavier penalties than did the other students.
With the preceding comments in mind, the present study
was undertaken with the understanding that the laboratory setting
used here only remotely resembles an actual jury trial.
How-
ever, if the findings of Wilson and Donnerstein (1977) can be
supported by future replications, then the possibility arises that
laboratory "verdicts 11 may prove to be a conservative estimate of
what actually occurs in the realm of real jury trial settings.
13
METHOD
Subjects
The subjects were eighty introductory psychology students
who participated in the study for course credit.
subjects were male, four of whom were black.
Thirty- nine
Of the forty-one
female subjects, two were black.
Procedure
The subjects were run in ten groups ranging in size from
five to nine subjects each.
As the subjects arrived, they chose
seats around a large table in a classroom.
The subjects were told that they were to be given a summary of a jury trial to read.
They were told to read everything
carefully. to answer the questions at the end of the booklet, and
to wait quietly for further instructions when completed.
The booklets, which had been randomly arranged previously, were distributed orderly.
When all of the subjects were
finished, the experimenter collected the booklets in the same
order in which they had been distributed.
The subjects were de-
briefed, and the experimental session was concluded.
The book-
lets were coded with a subject number, date, and time.
14
'
The booklet was composed of five pages.
The first page
had the following instructions:
This study is designed to investigate the criteria
which a potential juror, such as yourself, might use to
assess the guilt or innocence of a defendant.
You are going to read a summary of a jury trial.
After you have read the summary, you will be asked
to answer a series of questions. Read the summary carefully and answer the questions as honestly as you can.
Your answers will help provide a valuable insight into the
manner in which people, who might at some time serve on
a jury, evaluate evidence and assess guilt.
In the black, high socio-economic status condition, page
two was labeled "Summary of Police Files on Defendant Backgroundn and contained the following description:
Name:
Age:
Height:
Weight:
Birthplace:
Marital Status:
No. of Children:
Race:
Hair Color:
Hair Texture:
Skin Color:
Distinctive Features:
Birthmarks:
Occupation:
Employer:
Income:
Criminal Record:
Longstreet, Thomas
30 years
6 feet
190 pounds
Omaha, Nebraska
Divorced
0
Negro
Black·
Curly
Dark Brown
None
None visible
Assistant manager-Omaha
Branch Office
Merrill, Lynch, Pierce, Fenner,
and Smith
$2 8, 000 per year
None
In the low socio-economic status condition, the job was
listed as file clerk with income of $7,000 per year.
15
In the
.
American- Indian condition, the race was given as AmericanIndian, the hair texture was given as straight, the skin color was
given as sallow.
In the black, high socio-economic status, narcotics con-
dition, page three read as follows:
SUMMARY OF EVIDENCE IN THE CASE OF
THOMAS LONGSTREET
Charge: ILLEGAL POSSESSION OF NARCOTICS
The following facts were presented by the district
attorney as evidence supporting the guilt of the defendant:
1.
The arresting officer testified that he observed the
defendant driving erratically and had stopped him to
investigate.
2.
The arresting officer, suspecting the defendant of
being intoxicated, asked him to step out of the car
to perform a motor coordination task. Upon leaving
the car, a small plastic bottle fell to the street.
3.
When asked what was in the bottle, the defendant became belligerent, uncooperative, and refused to
answer.
4.
A police laboratory specialist testified that tests performed on the substance in the bag revealed the substance to be methaqualone, a nonbarbiturate hypnotic
drug commonly available under the name 11 Quaaludes. 11
The following facts were presented by the defense
attorney as evidence supporting the innocence of the defendant:
1.
The defendant testified he had been driving erratically because he was, at the time, under the influence of medication for a cold. The medication
makes him drowsy.
2.
The defendant was rude with the arresting officer but
only because he was not feeling well at the time.
3.
The defendant testified that he had never had anything to do with drugs, but that a friend had loaned
him the pills to help deal with insomnia that had
been bothering him lately.
4.
A fellow employee testified that the defendant, a
Negro, had been hired as an assistant manager a
year ago under an affirmative action plan. He had
16
never known the defendant to be involved with drugs
of any kind.
The judge then gave the jury instructions to return a
verdict of either guilty or innocent of the charge of illegal
possession of narcotics.
In the drunken driving condition, the third page read as
follows:
SUMMARY OF EVIDENCE IN THE CASE OF
THOMAS LONGSTREET
Charge: OPERATING A MOTOR VEHICLE WHILE INTOXICATED
The following facts were presented by the district
attorney as evidence supporting the guilt of the defendant:
1.
The arresting officer testified that he had observed the
defendant driving erratically and had stopped him to
investigate.
2.
The arresting officer testified that he had smelled alcohol on the defendant's breath and instructed him to perform a motor coordination task of placing his forefinger
on his nose with his eyes closed. The defendant failed
the test.
When asked to take a breath test, the defendant refused
3.
and was belligerent and uncooperative.
A police laboratory specialist testified that a blood
4.
sample taken from the defendant one hour after his
arrest revealed an ethanol level just under 0. 3o/o, but
that it could have been over the legal level at the time
of arrest.
The following facts were presented by the defense attorney as evidence supporting the innocence of the defendant:
The defendant testified he had been driving errati1.
cally because he was, at the time, under the influence of medication for a cold. The medication
sometimes makes him drowsy.
2.
The defendant was rude with the arresting officer
but only because he was not feeling well at the time.
3.
The defendant testified that he had only had one
drink prior to leaving home, and therefore doubted
the accuracy of the blood test.
A fellow employee testified that the defendant, a
4.
Negro. had been hired as an assistant manager a
17
year ago under an affirmative action plan. He had
never known the defendant to drink excessively,
much less drive while intoxicated.
The judge then gave the jury instructions to return a
verdict of either guilty or innocent of the charge of
operating a motor vehicle while intoxicated.
In the American-Indian condition, the fellow employee referred to the defendant as an American- Indian.
In the low socio-
economic status condition, the fellow employee referred to the
defendant's job as a file clerk.
Page four contained four questions based on the reading.
The first question asked the subject to place an "X" on a line
to indicate how guilty or innocent he believed the defendant to be.
The line was an unnumbered seven-point scale marked at equal
intervals.
Doubt. "
The first end point was labeled "Innocent Beyond Any
The second question asked the subject to explain the
reason for his answer to question one.
Question three asked the subject to assume that the majority of the jury had found the defendant to be guilty and that
the subject must now vote on a sentence.
The first part of the
question asked the defendant to place an "X" on the line to indicate with how much leniency or severity the defendant should be
sentenced.
The seven-point, unnumbered line was labeled
"Harshest Sentence Allowed by Law" at one end point, and "Most
Lenient Sentence Allowed by Law 11 at the other end point.
18
The
second part of the question asked the subject to indicate the number of years and/or months the defendant should receive.
Ques-
tion four asked the subject to explain his answer to question three.
Questions two and three were included for a check on the
effectiveness of the experimental manipulations and were not intended to be analyzed quantitatively.
Page five contained instructions to answer the questions
without turning back to the previous pages.
The first question
asked the subject to list, from memory, the defendant's name,
age, birthplace, marital status, race, occupation, and income.
This question was also· intended to be a check on the effectiveness
of the experimental manipulations.
Question two asked the sub-
ject to indicate a number from Oo/o to 1 OOo/o to show what the
subject believed to be the probability that the defendant actually
committed the crime.
Question five asked the subject to indicate
the probability that the defendant might commit the crime in the
future.
A sample booklet is included in the appendix.
The seven-point interval scales in questions one and two
on page four were scored as though each interval was divided
into fourths.
4. 75.
For example, a subject could score 4. 25, 4. 5, or
If an "X" fell between two points, it was scored as the
value of the next higher point.
For example, if an "X" fell be19
tween 4. 5 and 4. 75, it was scored as 4. 75.
The value of one
was given to the end points labeled "Innocent Beyond Any Doubt''
and ''Most Lenient Sentence Allowed by Law,
seven being assigned to the other end points.
20
11
with the value of
RESULTS
The group means on each of the five dependent variables
in each of the eight cells are displayed in Table 1.
The guilt
and punishment dimensions are, of course, unitless.
tence dimension is expressed in units of months.
The sen-
The two pro-
bability dimensions are expressed in terms of percentages and
represent,
respectively~
the probability that the defendant com-
mitted the present crime and the probability that he might commit the same crime in the future.
The data were analyzed by means of a packaged, multivariate analysis of variance computer program (Cohen & Burns,
1977).
For each effect, the program produces the four most
popular statistics for significance testing in the multivariate case,
namely, Roy's largest root, the Hotelling-Lawley trace, Wilks's
likelihood ratio, and Pillai-Bartlett trace.
The program also
computes a univariate F-test for each dependent variable for
each effect.
For the present study, the Pillai-Bartlett trace criterion
will be reported, following Olson's (1976) suggestion that this is
the recommended statistic for general use.
21
For any significant
effects, the univariate F-test will also be reported.
Table 2 lists the results of the multivariate analysis of
variance.
For each of the seven main and interaction effects,
the Pillai- Bartlett statistic and the corresponding significance
level are shown.
The only significant effect was a main effect
for race (p < . 03).
No other main or interaction effects were
significant.
According to Table 3, which lists the univariate F-test for
each of the five dependent vairables on the race dimension, only
the punishment variable achieved significance.
Table 1 permits
the calculation that the black defendant received an average punishment rating of 2. 44 while the American-Indian received an
average punishment rating of 2. 98 (p< . 05).
When asked to list from memory the defendant's name,
age, birthplace, marital status, race, occupation, and income,
95% of the subjects were able to remember correctly the race
and income, and 75o/o were able to remember the race, income,
and occupation.
It appears, then, that the race and socio-econo-
mic status manipulations were effective.
On the other two questions asking the subjects to explain
their answers to the guilt and to the punishment and sentencing
questions, the subjects explained their answers with reference to
the evidence presented.
For example, subjects would refer to
22
the blood alcohol level or to the defendant's admission of possession of drugs.
Several subjects referred to the fact that the
defendant had no previous record.
Other subjects suggested that
in addition to or instead of a jail term, the defendant should be
sentenced to a fine or to classes about drug or alcohol abuse.
No subjects referred to the defendant's race or socio-economic
status.
A second analysis was performed using only data for the
sixty-one subjects (75o/o of the total) who correctly remembered
the defendant's race, income, and occupation.
The group means for each of the five depending variables
in each of the eight cells are reproduced in Table 4.
Since the
number of subjects in each condition is no longer the same, the
number in each ce 11 is also reproduced.
The results of the multivariate analysis of variance are
reproduced in Table 5.
As can be seen, the effect due to race
became more significant (p < . 02) than the effect in the total
sample of eighty subjects (p< . 03).
No. other significant effects
were found.
The results of the ·univariate analysis of variance along
the race dimension, reproduced in Table 6, show that the dependent variable of punishment achieved considerably greater
significance (p < . 006) than achieved in the total sample of eighty
23
subjects (p <. 05).
No other dependent variable achieved signifi-
cance.
The total mean rating for each race on the punishment
variable can be determined from Table 4.
After taking into
account the different number of subjects per cell, it can be seen
that the average punishment rating for the black,,,·~efendant was
2. 08 while the average punishment rating for the American- Indian
was 2. 85.
Thus, the direction of the effect is the same as it
was using the total sample of eighty subjects.
A harsher sen-
tence was recommended for the American- Indian than for the
black (p < . 006 ).
24
DISCUSSION
No support was found for either the hypothesized main
effect due to socio-economic status or the interaction effect between race and type of crime.
A significant main effect due to
race was found (p< . 03) but only on the punishment dependent
variable.
Subjects recommended harsher sentences for the
American-Indian than for the black (p < • 05).
It would also be reasonable to expect the length of sentence assigned by the subjects to be longer for the AmericanIndian.
cant.
However, the difference on this variable was not signifiSince some of the subjects recommended a sentence other
than a jail term, perhaps the device used to measure this variable was not sensitive enough to pick up actual differences.
It
is also possible that the subjects did not have a uniform concept
of exactly what a harsh or lenient sentence should be.
It would
be of interest to test the effect on the results of an instruction
from the judge as to how long or short a sentence the law permits.
No significant difference existed between the two races on
the guilt dimension..
The results reviewed thus far indicate that
25
judgments made by jurors are affected by defendant race when
the issue is how severely the defendant should be punished but
not when the issue is guilt of the defendant.
If a similar effect
occurs in real jury trials, then the race effect could be overcome by allowing jurors to assess guilt and not allowing them to
determine punishment.
However, in many instances real jurors
are permitted to dete;t:'mine punishment indirectly by determining
the seriousness of the crime committed by the defendant.
For
example, in a murder trial, jurors may not only have the option
of determining guilt or innocence, but may also have the option
of determining if the defendant was guilty of first; second, or
third degree murder.
Each level of crime implies a different
range of sentences available for punishing the defendant.
Such
an optional method for determining guilt is, of course, valuable
when the defendant may be guilty of a lesser crime than the one
with which he was charged.
Another consideration of importance here is that race may
not affect jurors 1 judgments in all possible jury trials.
In the
laboratory study of Gleason and Harris (1975), no significant
main effect due to race was found.
significant race effect.
The present study did find a
The contradictory findings could be a
result of the differences between the two studies.
For instance,
the Gleason and Harris (1975) study used the crime of murder,
26
a crime which involves a victim.
The present study used the
crimes of possession of narcotics and drunken driving. both of
which are "victimless" crimes.
It could be inferred from the
findings of the two studies that the race effect may exist in some
situations. but not in others.
If this is true of real jury trials.
then no prescription can now be made for preventing the race
effect. since it is not known when the race effect will or will not
occur.
Another interesting point concerning the race effect is that
the effect became more pronounced when the data was reanalyzed
using only the data from the subjects who. could recall the race
and socio-economic status cues.
Since no subject related data
were collected. nothing can be said about why some subjects recalled the race and socio-economic cues and why others did not.
Likewise, nothing can be said about why the race effect strengthened when considering only the data from those subjects who recalled both cues.
If the subjects who recalled both cues could be
labelled as "racists,
11
then it could be said the race effect was
stronger among them because of their increased sensitivity to
racial cues.
However, there are no data to support the con-
tention that such subjects are more race sensitive.
The present study did _not adequately test the cognitive
explanation for the effect of socio-economic status on judgments
27
made by laboratory jurors.
The test of this explanation was not
adequate since the socio-economic status effect failed to occur.
The socio-economic status effect may have failed to occur
for the same reasons that the race effect did occur.
That is,
perhaps the socio-economic effect occurs in some situations, but
not ln others.
The effect occurred in the Gleason and Harris
(1975, 1976) studies but not in the present study.
The notion that race and socio-economic effects may vary
•
as the situation varies seems plausible given the findings of the
present study and the Gleason and Harris (1975, 1976) studies.
However, with respect to race and type of crime committed, the
present study examined the same notion.
That is, it was pre-
dieted that the subjects would tend to find the black defendant
guilty of possession of narcotics and the American- Indian defendant guilty of drunken driving.
action failed to materialize.
The race by type of crime interThe failure to find the interaction
could indicate that the subjects do not discriminate differently
between minorities as the type of crime varies.
Per~ps
the
interaction might still occur if a white defendant is used as one
of the· defendants rather than using a minority race for each of
the defendants.
Perhaps also the subjects associate certain types
of crimes rather than specific crimes with different races.
As
mentioned earlier, crimes may vary along the dimension of in-
28
volving a victim.
collar 11 or
11
Some crimes may also be considered as "white
blue collar" crimes.
Subjects may form expectations
for different groups of defendants on the basis of differences
among crimes other than those chosen for study here.
In summary, the present study raises questions of importance to, future studies of}he effect of jurors 1 prejudice on
courtroom judgments.
First, it cannot now be said, as Gleason
and Harris (1975) said, that race is not an important cue influencing judgments made by jurors.
The race effect may occur in
some situations and not others, in which case the problem becomes one of determining the parameters controlling when the
race effect will or will not occur.
The type of crime committed
may still be a viable cue given the results of the Gleason and
Harris (1975) study which used a crime against a victim, and
the results of the present study which used
11
victimless 11 crimes.
Perhaps the consequences of the crime may be a factor in determining whether or not a race effect will occur.
Second, subject characteristics may affect the occurrence
of the race effect.
In the present study, the race effect was
strengthened when considering only the data from those subjects
who recalled the race and socio-economic cues.
real jurors are attentive to such cues.
Perhaps not all
If only some jurors
attend to such cues, would the race effect be strengthened among
29
such jurors ?
Finally, nothing can be said about the cognitive explanation
for the effect of socio-economic status.
In fact. the possibility
arises that the socio-economic status effect may occur only in
some situations.
30
TABLE 1
MEAN SCORES OF UNREDUCED GROUP DATA
-a
s
Q)
C)
rJl
Q)
s::
...c::
,ti
•.-I
;::!
Group
d
:,:...,
:,:...,
+>
......
......
•.-I
+-'
......
Q)
•.-I
s::::::$
-a
!l.t
rn.
Q)
4-!
......
......
i::..O
Q)
cd
rn..o
Q)
0
$..<
$..<
!l.t!l.t
Q)~
$.-<..0
.B;::!
0
$..<
f:r.I.P-t
High Ses
Black
Alcohol
_Drug
59;'50d
51.50
31. ood
39.50
3.70
3.65
53.50
53.70
45.00
22.10
2.00
2.85
1. 47
11.82
38.50
72.40
23.00
38.10
3. 25
3.05
3.78
8.80
64.40
60.50
50.50
45.50
4. 28a
4.20
2.68b
2. 23
3. 83
3.60
3.19
2.45
3. 58
5.15
4.80
4.30
3.40c
10.47
Indian
Alcohol
Drug
Low Ses
Black
Alcohol
Drug
Indian
Alcohol
Drug
a
1
= Innocent
Beyond Any Doubt, 7
= Guilty Beyond Any
Doubt
b 1 = Most Lenient Sentence Allowed By Law, 7 = Harshest Sentence Allowed By Law
c Unit is Year
d Expressed as a percentage
31
TABLE 2
MULTNARIATE ANALYSIS OF VARIANCE
OF UNREDUCED GROUP DATA
PillaiBartletta
Approximate
Ses
• 04123
. 58477
Race
.16092
2.60818'"
Crime
. 10331
1. 56682
Ses x Race
. 05814
. 83949
Ses x Crime
. 06653
. 96927
Race x Crime
.12300
1. 90746
Ses x Race x Crime
. 05344
. 76776
Source
Fb
..,_
a~
= 1,
~
= lt,
n - 33.
b with df = 5 and 68 •
. 03.
32
TABLE 3
UNIVARIATE ANALYSIS OF VARIANCE OF
UNREDUCED GROUP DATA
RACE DIMENSION
~)
Fa
Variable
Guilt
• 32025
_,_
3.83654'~'
Punishment
Sentence
• 57381
Present Probability
.12629
Future Probability
a with df
p
==
2.16810
1 and 72 •
. 05.
33
TABLE 4
MEAN SCORES OF REDUCED GROUP DATA
-aa>
s
..t::
~
.....
~
Group
d
.....rn
:>-,
:>-,
+>
a>
CJ
~
a>
~
:::1
1:!
~
V1
a>
+>
......
·-<
.....
,-I
,-I
+> ......
~..0
a> ..0
a> ro
rn..o
a> 0
+>
..0
0
P-tP-t
·~
ilt
l-<
~--<
~
~
l-<
ro
l-<
n
High Ses
Black
Alcohol
Drug
3. 93a
4.04
1. 96b
1. 96
2. 00 c
3.82
51. 29d
52.14
25. 71 d
45.00
3.83
3.60
2.29
2.45
• 67
3. 65
41. 67
33. 70
34.17
22.10
6
10
3.84
4. 71
2.03
2.39
1. 09
3. 71
45.63
60.57
26.25
31.57
8
7
4.80
4.80
3.25
3.42
3.78
11. 67
64.40
60.00
50.50
40.83
10
7
7
Indian
Alcohol
Drug
Low Ses
Black
Alcohol
Drug
Indian
Alcohol
Drug
a 1 = Innocent Beyond Any Doubt. 7 = Guilty Beyond Any Doubt
. b 1 = Most Lenient Sentence Allowed By Law, 7
tence Allowed By Law
c
d
Unit is Year
.
Expressed as Percentage
34
= Harshest
Sen-
6
TABLE 5
MUL TIV ARIA TE ANALYSIS OF VARIANCE
OF REDUCED GROUP DATA
PillaiBartletta
Source
Approximate
Fb
Ses
.12262
1. 36961
Race
. 21765:.:'
2.72629*
Crime
. 08329
. 89045
Ses x Race
.12434
1. 39159
Ses x Crime
.03385
1. 04053
Ses x Race x Crime
.02969
.29989
a~
= 1,
~
= 1!.
n = 23!.
b with df = 5 and 49 .
. 02.
35
TABLE 6
UNIVARIATE ANALYSIS OF VARIANCE OF
REDUCED GROUP DATA
RACE DIMENSION
Variable
Guilt
• 16161
Punishment
8. 04794>!<
Sentence
1. 43774
Present Probability
• 07464
Future Probability
• 63898
a df
p
=
1
and 53
• 006
36
REFERENCES
Bullock, H. A. Significance of the racial factor in the length
of prison sentences, Journal of Criminal Law, Criminology, and Police Sciences, 1961, 52, 411-417.
'1
Cohen, E. and Burns, P. SPSS-MANOVA-Multivariate analysis
of variance and covariance (Document No. 413). Evanston, Illinois: Northwestern University, Vogelback Computing Center, June 1977.
Dion, K., Berscheid, E. and Walster, E. What is beautiful is
good, Journal of Personality and Social Psychology, 1972,
24, 285-290.
Gleason, J. M. and Harris, V. A. Race, socio-economic status,
and perceived similarity as determinants of judgments by
simulated jurors, Social Behavior and Personality, 1975,
3, 175-180.
Gleason, J. M. and Harris, V. A. Group discussion and defendants' socio-economic status as determinants of judgments
by simulated jurors, Journal of Applied Social Psychology,
1976, 6, 186-191.
Jurow, G. L. New data on the effect of a "death qualified" jury
on the guilt determination process, Harvard Law Review,
1971, 84, 567-611.
Kalven, H. and Zeisel, H. The American Jury.
University of Chicago Press, 1971.
Chicago, The
Klinger, E., Albaum, A. and Hetherington, M. Factors influencing the severity of moral judgments, The Journal of
Social Psychology, 1964, 64, 319-326.
Landy, D. and Aronson, E. The influence of the character of
the criminal and his victim on the decisions of simulated
jurors, Journal of Experimental Soctal Psychology, 1969,
5, 141-152.
37
McGlynn, R. P., Megas, J. C. and Benson, D. H. Sex and
race as factors affecting the attribution of insanity in a
murder trial, The Journal of Psychology, 1976, 93, 9399.
Nemeth, C. and Sosis, R. H. A simulated jury study: Characteristics of the defendant and the jurors, The Journal
of Social Psychology, 1973, 90, 221-229.
Olson, C. L. On choosing a test statistic in multivariate analysis of variance, Psychological Bulletin, 1976, 83, 579One More Plan to End Fraud in Welfare, U.S. News and World
Report, January 9, 1978, 41-44.
Reynolds, D. E. and Sanders, M. S. Effect of defendant attractiveness, age, and injury on severity of sentence
given by simulated jurors, The Journal of Social PsycholoEX.• 1975, 96, 149-150.
Richey, M. H. and Fichter, J. J. Sex differences in moralism
and punitiveness, Psychonomic Science, 1969, 16, 185186.
Shaw, J. I. Reactions to victims and defendants of varying degrees of attractiveness, Psychonomic Science, 1972, 27,
329-330.
Sigall, H. and Landy, D. Effects of the defendant's character
and suffering on juridic judgment: A replication and
clarification, The Journal of Social Psychology, 1972,
88, 149-150.
Sigall, H. and Ostrove, N. Beautiful but dangerous: Effects of
offender attractiveness and nature of the crime on juridic
judgment, Journal of Personality and Social Psychology,
1975, 31, 410-414.
Stephan, C. Sex prejudice in jury simulation, The Journal of
Psychology, 1974, 88, 305-312.
Wilson, D. W. and Donnerstein, E. Guilty or not guilty? A
look at the "simulated" jury paradigm, Journal of Applied
Social Psychology, 1977, 7, 175-190.
38
APPENDIX A
SIMULATED JURY STUDY
This study is designed to investigate the criteria which a
potential juror, such as yourself, might use to assess the guilt
or innocence of a defendant.
You are going to read a summary of a jury trial.
After you have read the summary, you will be asked to
answer a series of questions.
Read the summary carefully and
answer the questions as honestly as you can.
Your answers
will help provide a valuable insight into the manner in which
people, who might at some time serve on a jury, evaluate evidence and assess guilt.
39
APPENDIX B
SUMMARY OF POLICE FILES ON DEFENDANT BACKGROUND
Name:
Longstreet, Thomas
Age:
30 years
Height:
6 feet
Weight:
1~0 pounds
Birthplace:
Omaha, Nebraska
Marital Status:
Divorced
No. of Children:
0
Race:
Arne rican- Indian
Hair Color:
Black
Hair Texture:
Straight
Skin:
Sallow
Distinctive Features:
None
Birthmarks:
None visible
Occupation:
File clerk--Omaha Branch Office
Employe:::
Merrill, Lynch, Pierce, Ferner,
and Smith
Income:
$7, 000 per year
Criminal Record:
None
40
APPENDIX C
SUMMARY OF EVIDENCE IN THE CASE OF
THOMAS LONGSTREET
Charge: OPERATING A MOTOR VEHICLE WHILE INTOXICATED
The following facts were presented by the district attorney
as evidence supporting the guilt of the defendant:
1.
The arresting officer testified that he had observed the defendant driving erratically and had stopped him to investigate.
2.
The arresting officer testified that he had smelled alcohol
on the defendant's breath and instructed him to perform a
motor coordination task of placing his forefinger on his
nose with his eyes closed. The defendant failed the test.
3.
When asked to take a breath test, the defendant refused
and was belligerent and uncooperative.
4.
A police laboratory specialist testified that a blood sample
taken from the defendant one hour after his arrest revealed an ethanol level just under 0. 3o/o, but that it could
have been over the legal level at the time of arrest.
The following facts were presented by the defense attorney
as evidence supporting the innocence of the defendant:
1.
The defendant testified he had been driving erratically because he was, at the time, under the influence of medication for a cold. The medication sometimes makes him
drowsy.
2.
The defendant was rude with the arresting officer but only
because he was not feeling well at the time.
3.
The defendant testified that he_"had only had one drink prior
to leaving home, and therefore doubted the accuracy of the
41
p .
blood test.
4.
A fellow employee testified that the defendant, an American- Indian, had been hired as a file clerk a year ago
under an affirmative action plan. He had never known
the defendant to drink excessively, much less drive while
intoxicated.
The judge then gave the jury instructions to return a
verdict of either guilty or innocent of the charge of operating a
motor vehicle while intoxicated.
42
APPENDIX D
QUESTIONS - FIRST SET
read.
1.
Answer the following questions based on what you have just
YOU MAY refer back to previous pages, if you wish.
Place an °X11 on the line be low to indicate how guilty or innocent you believe the defendant to be.
Innocent
Beyond
Any Doubt
Guilty
Beyond
Any Doubt
2.
Explain the reason for your answer to Question 1.
3.
Assume that the majority of the jury found the defendant to
be guilty and that you must now vote on a sentence.
A) Place an 11 X 11 on the line below to indicate with how
much leniency or severity the defendant should be sentenced.
Most Lenient
Sentence
Allowed By Law
Harshest
Sentence
Allowed By Law
B) Indicate in years and/ or months the length of sentence
you believe the defendant should receive.
______YEARS, _______MONTHS
4.
Explain your answers to Question 3.
Please be sure ALL of the above questions are answered before proceeding to the next set of questions.
43
APPENDIX E
QUESTIONS- SECOND SET
Answer the following questions.
previous pages.
1.
DO NOT turn back to any
Based on your memory, fill in as many of the blanks as you
can.
NAME
----------------------------------------------------------------------------------BIRTHPLACE
---------------------------------MARITAL STATUS
-----------------------------RACE
-----------------------------------------OCCUPATION
---------------------------------INCOME
---------------------------------------
AGE
2.
What do you believe is the probability that the defendant actually committed the crime?
Indicate a number from 0 to 1 00.
O% means you are certain the defendant DID NOT commit the
crime.
100% means you are certain the defendant DID commit the
crime.
------%
3.
What do you believe is the probability that the defendant
might commit the crime in the future?
Indicate a number from 0 to 100.
O% means you are certain the defendant WILL NOT· commit
the crime in the future.
100% means you are certain the defendant WILL commit the
crime in the future.
o/o
When you are finished, close ..,..,th.-l. .s....,..b_o_o..,..k
.
...le-.t and please wait quietly for
further instructions.
44
© Copyright 2026 Paperzz