Keeten v. Garrison Brief in Opposition to Petition for Writ of Certiorari

    No. 84-5187
IN THE

SUPREME COURT OF THE UNITED STATES 
October Tern, 1984

CHARLES BRUCE KEETEN, 
Petitioner,
-against-

SAM GARRISON, Warden of Central 
Prison, and the State of North 

Carolina,
Respondents.

BERNARD AVERY,
Petitioner,
-against-

ROBERT HAMILTON, and RUFUS L. EDMISTEN, 
Attorney General of North Carolina,

Respondents.

BRIEF IN OPPOSITION TO PETITION 
FOR WRIT OF CERTIORARI TO THE 
UNITED STATES COURT OF APPEALS 

FOR THE FOURTH CIRCUIT

LACY H. THORNBURG 
ATTORNEY GENERAL
RICHARD N. LEAGUE
Special Deputy Attorney General
North Carolina Department of Justice
Post Office Box 629
Raleigh, North Carolina 27602
(919) 733-2011

*Remaining Case Caption 
on Inside



1

QUESTIONS PRESENTED

I. SHOULD CERTIORARI ISSUE TO CONSIDER WHETHER THE
EXCLUSION OF PERSONS WHOSE DEATH PENALTY OPPOSITION 
CURRENTLY PREVENTS THEIR INCLUSION ON CAPITAL CASE 
JURIES CREATES AN UNCONSTITUTIONALLY 
CONVICTION-PRONE JURY OR DEPRIVES ACCUSED PERSONS 
OF A FAIR CROSS-SECTION OF THE COMMUNITY ON THEIR 
JURIES?

II. SHOULD CERTIORARI ISSUE TO CONSIDER WHETHER THE WAY
VOIR DIRES ARE CONDUCTED IN CAPITAL CASES CREATES 
AN UNCONSTITUTIONALLY CONVICTION-PRONE JURY?

HI. SHOULD CERTIORARI ISSUE TO CONSIDER IF IT WAS
CONSTITUTIONAL TO EXCUSE A JUROR BECAUSE, WHENEVER 
ASKED, SHE STATED SHE WAS NOT SURE COULD VOTE FOR 
THE DEATH PENALTY EVEN IF THE LAW REQUIRED IT ON 
THE FACTS OF THE CASE INVOLVED?



Pages
QUESTIONS PRESENTED............................................... 1
TABLE OF CASES.................................................... 11
OPINION BELOW...................................................... 2
JURISDICTION....................................................... 2
CONSTITUTIONAL PROVISIONS AND STATUTES............................ 2
STATEMENT OF THE CASE...........................................2
REASONS THE WRIT SHOULD NOT ISSUE.................................. 3
CONCLUSION........................................................ 12
APPENDIX A:....................................................... 13

APPENDIX ..........................................................15
APPENDIX ..........................................................17

TABLE OF CASES

DUNAGIN v. CITY OF OXFORD, MISSISSIPPI,
713 F . 2d 738 , ( 5 Cir. 1983).................................... 7

GRIGSBY v. MABRY, ____  F.2d _____  (8 Cir. 1985)..................2
PEOPLE v. FIELDS, 673 P.2d. 680 (Cal. 1984.)......................2
REASE v. UNITED STATES, 403 F.2d 322 (DCApp. 1979 ).............. 11
RISTAINO v. ROSS, 424 US 589 , 595 ( 1976)......................... 11
SMITH, KLINE AND FRENCH LABORATORIES V.

A. H. ROBINS COMPANY, 61 FRD 24, (ED PA 1973)................ 9
STATE v. AVERY, 299 NC 126 (1980)............................... 10
STATE v. COLVIN, 297 NC 691 (1979)..............................  10
STATE v. WEBB, 364 S02d 984 (La 1978)........................... 11
UNITED STATES v. GENERAL MOTORS CORPORATION,

384 US 127, (1966 ).......................

TABLE OF CONTENTS

7



NO. 84-5187
IN THE

SUPREME COURT OF THE UNITED STATES 
October Term, 1984

CHARLES BRUCE KEETEN, 
Petitioner,
-against-

SAM GARRISON, Warden of Central 
Prison, and the State of North 

Carolina,
Respondents.

BERNARD AVERY,
Petitioner,
-against-

ROBERT HAMILTON, and RUFUS L. EDMISTEN, 
Attorney General of North Carolina,

Respondents.

LARRY DARNELL WILLIAMS, 
Petitioner,
-against-

NATHAN A. RICE, Warden of Central 
Prison, and RUFUS EDMISTEN, 

Attorney General of North Carolina,
Respondents.

BRIEF IN OPPOSITION TO PETITION FOR 
WRIT OF CERTIORARI TO THE 

UNITED STATES COURT OF APPEALS 
FOR THE FOURTH CIRCUIT



BRIEF IN OPPOSITION TO PETITION FOR 
WRIT OF CERTIORARI TO THE 

UNITED STATES COURT OF APPEALS 
FOR THE FOURTH CIRCUIT

Respondents oppose a grant of certiorari in these 
cases. Although there is a conflict in the circuits on the 
conviction-prone jury issue, respondents feel the court will have 
no trouble summarily reversing the Eighth Circuit in Grigsby v.
Mabry, ______  F . 2d ________ (8 Cir. 1985) in which certiorari will
soon be sought by the State of Arkansas. In addition, the 
California Supreme Court seems to have recently moved away from 
its Hovey decision in People v. Fields, 673 P.2d. 680 (Cal. 1984) 
so that this remaining pillar of support for Petitioners may yet 
fall on its own. The method-of-voir-dire issue is based only on 
a single study and therefore is too undeveloped for meaningful 
court review. Finally, the single juror exclusion issue in 
Petitioner Williams' case is clearly covered by Wainwright v.
Witt, ____  US ______  ( 1985) .

OPINION BELOW 
JURISDICTION

CONSTITUTIONAL 
PROVISIONS AND STATUTES
STATEMENT OF THE CASE

Pursuant to Rule 34.2 of the Supreme Court, these items 
are omitted.

2



REASONS THE WRIT SHOULD NOT ISSUE
I. THE CONVICTION-PRONE JURY ISSUE SHOULD

NOT BE REVIEWED IN THIS CASE BECAUSE THE 
FACTUAL PREDICATES FOR FINAL DECISION 
ARE PROBABLY MISSING.

This case has been approached by the Petitioners as 
purely an evidence case and, as such, is too mundane for the 
court's consideration. They have tried to put it into both a 
directed-verdict mold, claiming all the evidence is in and on 
their side; and a trial-court-findings-bind-appellate-courts 
mold, which in view of the endorsement of the studies by the 
district court might automatically win for them. Neither 
approach, however, is correct; each will clutter the court's 
final decision on this issue, to be made most likely in Grigsby; 
and Petitioners have not shown so far (or even sought to show) 
that any of the excluded jurors in their cases could have been 
fair and impartial on the issue of guilt, despite their death 
penalty opposition. This was not expressly covered in voir dire 
and it is crucial to Petitioners because they say inclusion of 
such jurors is their only aim, not inclusion of all current 
Witherspoon excludables (WEs).

A.
Petitioner's first argument —  that the evidence is 

exclusively favorable to them —  is wrong; and not only that, the 
deficiencies noted in Witherspoon have not been remedied. If 
taken at face value, the studies Petitioners rely on are 
consistent with the idea that the stronger one's opposition to 
the death penalty, the likelier one is to have a somewhat 
jaundiced view of the criminal justice system; and that in mock

3



trials without such things as an oath or deliberations 
beforehand, the likelier they are to acquit. This, however, is 
exactly what the proof showed in Witherspoon, albeit more 
modestly. What was not shown then, or now, is (i) that general 
pre-trial attitudes about the criminal justice system will affect 
real juror behavior in concrete cases in the face of voir dire, 
oath, evidence, instructions, and deliberations;  ̂ (ii) that 
nullification is not a real problem or that it can be adequately 
handled if WEs sit on capital case juries; (iii) that if there is 
a true phenomenon of conviction—proneness, it has the substantial 
effect Witherspoon requires; and (iv) that mock jury trials 
predict the outcome of real jury cases. With these deficiencies 
in the predicate factual proof, the legal issue of whether 
conviction proneness in juries is constitutional cannot be 
reached.

Putting aside the probability of nullification and the

1The district court claimed that this was just common 
sense but it is an issue on which the general field of psychology 
is split into four groups —  attitudes cause behavior, behaviors 
cause attitudes, each cause the other, neither operates on the 
other, "Attitudes Cause Behaviors" (ST), Journal of Personality 
and Social Psychology, 37: 315 (March 1979). The quantification 
of the attitude —  behavior relationship frequently has been 
noted to be only about 10 percent at most. Put another way, 
differences in attitude account for (not necessarily cause) 10 
percent of the variation in behavior. Dr. Allen Wicker found 
this generally true in the 1959 survey of research on this 
relationship up until that date, "Attitudes Verses Actions 
(ST) , "Journal of Social Issues, Autumn 1969 :4 . Dr. Gerald S’nure 
testified to this in the Grigsby hearing; and Dr. Michael Saks 
found this to be true in the single-attitude variable in jurors 
he tested, Saks and Hastie, Social Psychology in Court (New York 
1978) 65. The only close relationship between attitude and 
action appears when the former is closely linked to the 
possibility of the latter, such as attitudes favoring political 
candidates causing a person to speak out in their behalf or 
strong religious beliefs causing a persons to attend church.

4



lack of substantial impact on which there is virtually nothing 
favorable to Petitioners and focusing on the first and fourth of 
the above since they involve the inferences Petitioners want to 
be drawn from their two main groups of studies, there are two 
pertinent adverse studies unmentioned by them, one by Dr. Steven 
Penrod, another by Dr. Hans Zeisel. Neither study appears to 
have been cited to the courts in either the Hovey or Grigsby 
cases.

Dr. Penrod, a well-known jury researcher, did a study 
for his doctoral thesis to determine if 347 persons in answering 
attitude questions about the criminal justice system and in 
voting in four mock cases (three criminal) showed any correlation 
between the several attitudes, between the several verdicts, or 
between the attitudes and the verdicts. This research is 
especially important because none of Petitioners studies test 
these; and the attitudes checked included whether a person not 
testifying probably had something to hide, a person being 
prosecuted was probably guilty by virtue of that fact, and guilty 
persons often escaped on technicalities, all similar to tnose in 
the studies Petitioners rely on. Dr. Penrod's tables and a 
rearrangement of his conclusions appear in Appendix A and he 
found virtually no correlations. In the Fourth Circuit, 
petitioners said only that Dr. Penrod had changed his mind due to 
a recent collaboration with Dr. Hastie, Petitioner s witness in 
Griqsby, and a statement appearing in their book, (a copy of 
which is attached as Appendix B for the court's own determination 
on this); disclaimed reliance on attitudes affecting behavior (an

5



assertion that immediately made their six surveys and about half 
of their testimony irrelevant); and asserted Dr. Penrod's study 
had no bearing because it did not differentiate between death 
penalty proponents and opponents, even though this fact showed 
that whatever group one was in, the attitudes and verdicts did 
not correlate. These skimpy arguments do not warrant the Court's 
rev iew.

Dr. Zeisel's study on "felt responsibility" in
simulations is set out in detail as Appendix C and its impact on
the issues cannot be minimized. After comparing the results of
10 real juries and two sets of "shadow" juries hearing the same
cases and being treated in the same way, Zeisel concluded that
the real juries' convictions in only one-half of the cases
compared with one set of shadow juries' convictions in 3 of 10
cases and the others' convictions in all 10 cases indicated that:

...despite all efforts to equate experiences, 
jurors on the real jury had a different 
standard for the meaning of "guilty beyond a 
reasonable doubt" than those whose decision 
was not binding on the defendant, whose 
conviction could not send an individual to 
prison.

Against this in the Fourth Circuit, however, Petitioners only 
claimed'that Zeisel was their witness, which was true despite 
this study which he did not mention in his testimony; that 
Zeisel's 1955 survey was not considered on this point, although 
it did not deal with felt responsibility at all and Zeisel 
himself described it as being "full of such weak points" (Moore 
Tp 83); and that Zeisel had changed his orginial conclusions in

6



"The Effect of Peremptory Challenges" (ST) 30 Stanford L. Rev. 
194, 498, 511-513 (1978) although this was true with only one set 
of the shadow jurors and was caused these jurors having been 
subject to voir dire, something missing in all the mock jury 
tests on which Petitioners relyl These arguments are also 
undeserving of this Court's time.

B.
Petitioners' other evidentiary-based argument —  that 

the district court's factual findings bound the circuit court, 
thereby automatically winning for them —  is both legally and 
factually incorrect also.

First, in a "documents case," as this is, appellate 
courts historically have taken more leeway in reviewing district 
court decisions, see e.g. United States v. General Motors 
Corporation, 384 US 127, 144 (1966) fn. 16 even within the 
confines of this rule. Moreover, as social science literature 
and testimony is involved,strict use of this principle of review 
would permit different results district-wide in this country, not 
just in individual cases. Only decisions on 'natters of law 
reviewable on appeal normally can have this effect. Therefore, 
this sort of review of similar evidence has been rejected 
recently in Dunagin v. City of Oxford, Mississippi, 718 F.2d 738, 
748 (5 Cir. 1983) fn.8. Were this not so, however, the district 
court's decision would remain incorrect for three reasons.
First, in the district court's opinion at p 1171-1177, 
it finds only that the studies presented by Petitioners are as 
they appear to be in terms of their procedures, calculations and

7



outcome. It does not find the criticism of those studies untrue 
or that Respondents studies and evidence are not true but only 
that they do not persuade it of the ultimate legal conclusion 
Respondents argue for. Therefore, no impediment to use of 
Respondents' evidence on appellate review exists.

Second, even under the standards of Zenith Radio 
Corporation v. Hazeltine Research, Inc., 395 US 100 (1969), the 
District Court's findings do not pass muster as they are not 
supported by substantial evidence, especially in light of the 
Penrod and Zeisel studies previously mentioned. Finally, the 
District Court misapplied the law and rested its decision on 
assumptions about attitudes affecting behavior, nullification 
being no problem, mock jury results predicting real world 
results, and substantiality of impact which are not proven. 
Therefore, the court need not wrestle with the transcript 
—  searching exercise Petitioners have in mind for it on this 
issue.

II. THE METHOD—OF-VOIR-PIRE-ISSUE , COMMON 
TO ALL THREE CASES, SHOULD NOT BE 
REVIEWED AS IT IS BASED ON ONLY A 
SINGLE PIECE OF SOCIAL SCIENCE 
RESEARCH YET ARGUES FOR A DRASTIC 
PROCEDURAL CHANGE

One of Petitioner's studies is a singular one done by 
Dr. Craig Haney of the University of California at Santa Cruz. 
Rather than dealing with conviction-proneness because of 
pre-existing attitudes about the death penalty, its thesis is 
that the way voir dire is held may have a biasing effect, mainly 
in that death penalty opponents are excluded (denigrated) while

8



others in saying they will consider the death penalty are 
accepted (favored) and have made a public (reinforced) commitment 
which increases their likelihood of imposing the death sentence.

This is a separate issue because a claim for relief is 
the aggregate of operative facts which give rise to enforceable 
right, Smith, Kline and French Laboratories v. A. H. Robins 
Company, 61 FRD 24, 28 (ED Pa 1973); and the method claim alone 
meets this description. Whether a jury is conviction-prone 
because of attitudes on the criminal justice system or the death 
penalty is simply not the same as whether it is conviction-prone 
because of the way the court holds a voir dire. To satisfy the 
Petitioners on the latter point all that needs occur is an 
individual voir dire and a single death penalty question, as 
proposed by Dr. Haney. However, to satisfy Petitioners on the 
first point requires inclusion of those persons currently 
excluded under the Witherspoon. If either of these were done 
alone, they would not ameliorate the other complaint.

Respondents have argued so far that the test was poorly 
done, to which there has been no substantial answer by 
Petitioners in either the District Court or the Fourth Circuit. 
However, even were Dr. Haney's study the best possible one, 
Respondents doubt that the court would make Dr. Haney a one man 
constitutional convention by requiring as a constitutional 
imperative the voir dire remedies his single study proposes 
which, as noted above, are a single question on the death 
penalty and an individual, sequestered voir dire.

In light of the above, certiorari should not be granted 
to review this issue.

9



III. THE INDIVIDUAL JUROR ISSUE IN 
WILLIAMS' CASE —  EXCLUSION OF 
MS. MELTON FOR HER OPPOSITION 
TO THE DEATH PENALTY —  SHOULD 
NOT BE FURTHER REVIEWED AS IT 
IS ADEQUATELY COVERED BY 
WAINWRIGHT V. WITT.

Ms. Melton, while stating generally she could do her 
duty as a juror, nevertheless hedged on three occasions when 
asked if her death penalty views would stand in the way of her 
following the law and the court's instructions if that would lead 
to the death penalty, saying she was not sure she could return a 
death sentence. She was excused for cause and no rehabilitation 
was attempted, although objection was then made by the defense 
and the North Carolina Supreme Court reviewed the issue on its 
merits. The Witherspoon, Adams and Witt cases cited by 
Petitioners all prevent relief on this issue and do this so 
clearly that no further exposition of the law need be made.

First, under Witt, Judge Snepp's voir dire and his 
decision to strike are the same sort of findings that Witt held 
entitled to acceptance under 28 USC §2254, absent circumstances 
not present here. This being so, his decision all but ends the 
review as he certainly knew the law, having been on the bench as 
a Superior Court judge since at least 1963, 273 NC iv; thereafter 
presiding over several capital case voir dires, see e.g. State v 
Colvin, 297 NC 691 (1979), State v. Avery, 299 NC 126 (1980) and 
having questioned Ms. Melton in language consistent with 
Witherspoon, Adams and Witt.

Second, while there are semantic differences between 
the language of two the prospective jurors, the Witt standard

10



easily covers Ms. Melton's responses. Uncertainty and the lack 
of a commitment before an oath obviously compromise and invite 
its demeaning if it is ultimately taken, so as to prevent or 
substantially impair the performance of juror duties. Also, the 
situation noted in Witt seems to be present here —  that bias may 
not be made clear by some people's spoken language, yet be 
revealed by their overall appearance. This is why the usual 
approach is to commit the decision on a juror to a trial court's 
discretion in an ambiguous situation, see e.g. Ristaino v. Ross, 
424 US 589, 595 (1976), Rease v. United States, 403 F.2d 322 
(DCApp. 1979), State v. Webb, 364 S02d 984 (La 1978).

Third, Judge Snepp's decision is bolstered by 
recollection that it is a general and unremarkable principle of 
law that while at the outset of voir dire, all persons are 
presumed competent to be jurors, if voir dire reveals an 
uncertainty about this due to views on punishment, the juror is 
disqualified, United States v. Gonzalez, 483 F.2d 223 (2 Cir. 
1973), rather than making the litigants take a chance on this.
Any other system would be absurd and Respondents cannot believe 
that this raises a constitutional question.

In light of the above, certiorari should not be granted
to review this issue.

11



CONCLUSION

court will 
certiorari

For the reasons above, Respondents pray that the 
deny the Petitioners application for a writ of

l —
This /3 rd day of March, 1985 .

LACY H. THORNBURG 
ATTORNEY GENERAL

Richard N. League
Special Deputy Attorney General
Post Office Box 629 
Raleigh, NC 27602 
(919) 733-2011

12



APPENDIX A: THE PENROD STUDY

Bearing in mind that in social science, a statistical

correlation of .30 is low, Downie 

Methods (New York, 1974) Chapter

for only 9% of any variance found

this case , Dr. Pen rod found: (a)

and Health, Basic Statistical 

7 and that a .30 would account 

according to the evidence in 

on attitude verdict correlation:

CORRELATIONS BETWEEN ATTITUDE VARIABLES 
TABLE 12 AND VERDICT PREFERENCES

Cases
Predictor Attitude

Variables Harder Rape Robbery Nellis Attitudi

lA Harsh Sentences .0 3 .02 .03 -,0d .25'
2* Punish Causers of Death -.01 .15 -.01 -.on ,n8'-
3* WeaDon Carriers Trouble .07 .08 -.02 -.05 .60 '

Minority Trials Unfair .01 -.on .06 .on • .16 '
5* No Testr.nony=Hldlnit -.03 .07 .03 -.06 .51 .
6 Require Rape Resistance .22 .26 -.02 -.01 .02
7 Drunlc Defendant Affect Verd .02 -.03 .02 .09 ' .01
8* Escape on Technicalities .09 .00 .02 -.11 .ns »
9* Prosecuted Probably Guilty .on .13 .06 .09 ! .36 -r
10 Collect on Pain & Suffering .01 .06 -.05 -.01 -. m
11 Awards Cause Suits .13 .01 .00 -.15 .25
12 Consider Defendant's Wealth .08 .05 -.07 -.01 .on

Attitude Scale (Sun •) l • o -1 .16 .06 -.06

This caused him to conclude:

The pattern of correlations clearly^ shows that most 
of the attitude variables bear no relationship to the 
verdict perferences (with the notable exception of 
the rape resistance question and the rape verdict):
23 of the 48 relevant correlations are negative.... 
The ultimate resolution of this problem was to 
construct a sing~le_new_,^t titude_ scale based on the 
summed normalized responses of the jurors on the 
seven attitude variables concerned with crimina1 
behavior',' 'inferences of ~de fendant' s guilt.and 
punitiveness.... The results for the scale .are 
even weaker. In each instance the scale correlates 
in the wrong direction wich the case verdicts.

(The text above was rearranged for purposes of emphasis; emphasis 

was supplied. )

13



(b) on inter-attitude correlation:

The intercorrelations between the attitude variables 
were rather low - only the "weapons" with "no 
testimony" (r=.25), the "harsh sentence" with 
"minority trial fairness" (r=.23), the "harsh 
sentence" with "drunk defendant" (r=.21) and the 
"consider wealth" with "awards cause suits" (r=.2) 
exceeded r= .2. twenty—two other correlations 
ranged between .10 and .20 and 40 others were under .10.

(c) and on correlation between verdicts:

TABLE II CORRELATIONS AMONG VERDICTS

,c a us in g

Murder Raoe Robberv

Rape Verdicts -.131

Robbery Verdicts .040 .021

Negligence Verdicts .030 .036 -.025
him to conclude:

These verdict preferences were not highly interre­
lated. As shown in Table II the highest correlation 
(-.13) is between the rape and murder verdicts (p <.02). 
None of the other correlations are statistically 
significant. The low correlations suggest there are 
no general conviction- or acquitta-l-prone juror 
"types" (e.g. authoritarians or dogmatists) whose 
behavior can be predicted across cases.

While Dr. Penrod's subjects were not divided by death 
penalty views into DQs and WEs and therefore his study does not permit 
a comparison between the groups, the fact that in recent years DQs 
have greatly outnumbered WEs, coupled with the fact that the correlations 
were virtually non-existent appear to mean that because DQs' attitudes 
on the matters inquired about do not relate to their vote, these persons 
are not conviction-prone to any significant degree on account of their views.

14



APPENDIX B: HASTIE & PENROD
INSIDE THE JURY (CAMBRIDGE, 1983)

INDIVIDUAL DIFFERENCES AM ONG JURORS

In fact. estimates of the method's efficacy arc not precise. An accuracy
rate of TO percent has been claimed (Kairvs et al.. 1975: Silver, 1978). How­
ever. if half of all jurors are expected to vote for conviction, simple guessing 
or coin-flijipmg would yield a 50 percent accuracy rate. And il 80 percent 
of the jurors preier conviction, either in a single case or across cases, an 80 
percent accuracy rate can be achieved by always predicting conviction.

Efforts to select jurors on the basis of their demographic, personality, or 
attitudinal characteristics also have not enjoyed much success (Berman 6c 
Sales, 1977). The few relationships observed do not provide significant ad­
vantages in jury selection. In one study 780 jurors viewed the same video­
taped burglarv trial and then deliberated, as pines, to a verdict (Saks. 1977 . 
Demographic and attitudinal information from the jurors was used to pre­
dict their votes. The single best predictor, a question on whether crime was 
mainly the product of "bad people” or of "bad social conditions,” ac­
counted for only 9 percent of the variance, and the four best predictors to­
gether accounted for less than 13 percent.

The relationships between juror behavior and juror attitudes, personahtv. 
and demographic characteristics are not well understood. More than i60 
jurv studies provide little systematic evidence' that personality variables, 
such as authoritarianism, locus of control, ana legal attitudes, provide tne 
predictive power needed to detect and challenge biased jurors, even assum­
ing that requisite information on prospective jurors is available in voir dire 
(Berg & Vidmar. 1975; Buckhout. 1973; Boehm. 1968; Jurow. 1971; Buckh- 
out ct al., 1979; Sosis. 197-1; Kauffman 6c Rvekman. 1979). Furthermore, de­
liberation may operate to nullify biases that exist before deliberation t Kap­
lan 6c Miller, 1978). On the whole, this implies low efficacy for jury selection 
strategies based on personality characteristics.

But attitudes about crime in general, particular crimes, or particular 
cases are a fruitful source of information on juror bias. Public opinion pods 
find that jurors who arc strongly opposed to the death penalty tend to be 
less conviction prone than jurors who are not strongly opposed t Bronson. 
1970, 1980; Harris. 1971; Ellsworth 6c Fitzgerald. 1983). Data Irom posttrjjij 
interviews with actual jurors and simulated juror decision-making studies 
are also consistent with the conclusion that those strongly opposed to the 
death penalty are likelier to vote tor acquittal (Zeisel. 1969; Guldlierg. 
1970; jurow. 1971; Ellsworth. Thompson. 6c Cowan. i983). Researchers 
found that iiirors' altitudes regarding rape are also related to their decisions 
in rajie cases iFcild. 1973). Attitudes on punishment were toimd to tie the

127

15



a civil case—which included upcninc, and closing arguments, witness lesti- 
mon>. and instructions by a judge iPenrod. 19791. Analysis of the jurors al­
titudinal and demographic characteristics showed that no variable had a 
strong relationship to actual verdict preferences, the highest correlation 
-being +.I-S. lutiddition. multiple regression techniques showed an average 
of 11 percent explained variance for the verdict preferences in the four 
casev with the highest R* =  .16 for the rape case. This implies that anyone 
■who is aware of the relationships between the attitudes, demographic fac­
tors. and verdict preferences in the rape ease can. in principle, do better 
than chance in predicting juror preferences. If half the jurors in a jury ve­
nire vote to convict at the end of the trial, those relationships would accu­
rately predict the verdict preferences of TO percent of the jury venire, as 
opposed to the 50 percent accuracy rate obtained by coin flipping. Of 
course, this assumes that all the appropriate attitudinal and demographic 
data arc available to the jury selection team. It is quite unlikely, however, 
that the usual systematic methods of jury selection will yield a criterion as 
reliable as the verdict preferences obtained in the Penrod study.

AID analyses were also performed in Penrod’s study. The results were 
similar to the regression results: in none of the four cases was it possible to 
predict verdict preference at high levels of accuracy. Although the AID 
analyses could sometimes account for as much as 25% of the variance in 
verdicts, subsequent tests demonstrated that much of the variance ac­
counted for" was actually error variance. In effect, no research lias pro­
vided evidence that social scientific methods can be a powerful aid to at­
torneys in the task of detecting juror bias. However, attitudes, particularly 
case-relevant attitudes, such as toward the death penalty or rape, appear;Jo 
be the most powerful individual difference predictors of verdict preference

peM&sii that have been studied to date.
■ One purpose of the present mock jury study was to evaluate jury selec­

tion techniques. The 825 jurors provided information on their personal 
backgrounds in a postdeliberation questionnaire, including age. gender, oc­
cupation. residence, education, political party, ideology, marital status, in­
come. race, number of previous cases heard as a juror, and number of previ­
ous criminal cases heard as a juror. This information was entered into a 
step-wise multiple regression, with variables such as occupation and resi­
dence included to reflect socioeconomic status, and using predelilieration 
verdict preference as the dependent variable, assuming equal interval 
values ranging from one lor first degree murder to lour for self-tleleiise. 
Only four variables produced significant /•"’s (overall /•' [5. i-12| =  5.22). htn- 
plovcd versus nonemployed. including retired, entered first with the highest 
correlation r =  .119. Juror gender ir =  —.073) brought the multiple H to 
.115. number of previous criminal eases as jurors (r =  —.059) raised R to . 15.»

128



APPENDIX C: ZEISEL
STUDY

l } • >

► :• ■

iC

.t*iA

1 (J . J:

Proceedings
of the irJ: '...

Division of Personality.
and ; " •

Social Psychology,. 1974

*♦ ..

1 *' ; ’ ‘ T

- t

U a;u  ~ii f, u l*v** p,*! p< \* vy .J  !:• h '/ y lV  ft tj „ t... y m i  ij t/ r# yc.f M 1 M j--g O  a H • •Livi'E.A.-a LJJ ...
; ** /

• -i' . IVti-•*’.< * ; - •• ■
•VjJ V  ;i.fj* .* ••. v.-a'.: . ■* ■ *

. I' t| •V ‘ ■■'I1 • i: .•••.'•'•:• *. ••• i•i.. vu-\

17



Circle

276
A Courtroom Experiment on 

Juror Selection and Decision-Making
Shari Seidman Diamond, University of Illinois, Chic;

Hans Zeisel, University of Chicago Law School

— r. la£t several years psychologists have begun to turn their attention
_of their investigations have centered on the jury,j^deciliw*.

, iacr several vears psychologists have begun to cum m e n
In the las y investigations have centered on the jury. Apart iT

f  !°socIoUg « l ^ d f b  hKalven and Zeisel (1967) comparing Judge and Jury decis^. 
large sociol B , a the laboratory with mechanically reproduced trial
» » «  of the »ort ha, boon -  *  “ “ " . l l a a y s  . haunting suspicion that lnfot-
frequently using stud.nt. “  ™"'ndsubject populations will not apply »

of their case. This experiment was planned to test tne accuracy
aactclsing. such challcngas.^ ^  federal tt?rles cases in the b S. CM*,,
for the northern District of Illinois uas tried ‘- the presence o f t h r e } ” “ J; ...

r -S  p ec^ u /h y
not°learn ^interviewing indicated that this
system had successfully avoided juror awareness). , ;th were present i» l*
" During the trial, all three juries were coi^ro^ihen the lawyers
courtroom throughout the trial, they were rem 4Uror fee by the Courts, and their
wished the Jury absent, they were paid the standard Juror f b y ^  ^  gf the trial, e»‘ 
employers were notified they were involved in jury ser . A  ^  evidence for inspec*
jury went to its own deliberation room, receiv|£ cop*“  lunchtime occurred during 
tion, and deliberated until a verdict was reached ^ tion on a point of U« ~
deliberations, free meals were supplied. If any Jury " exocJilnental Juries were treated 
ing deliberation, the Judge answered it. In short, the T  f*rth Jhe judge impressed
ms nearly as possible as though they were real juries. A case began. The
pon them the importance of their Job in a special sessionh / their verdict would «»* de ______ m  rh* exnerimental Jurors knew that cneir v«=‘ ___

ir Job in a special session betore m e  -e.
ii; differences' were: (1) Che experimental Jurors knew Chet the rver Q)„,
ide the case; (2) they sat in the spectators' seats ^ r" ions Indicating hi.

T5r“ i”^?s- •« -
tTen"criminal'cases lasting f ™  several

:rom draft evasion to conspiracy and extortion, before the jury returned it*
:hree separate jury verdicts were produced p^thermore dccision would h.«
/evdict, the judge in the case indicated on a questionnaire wn
>een.

Cuilty
Judge

9
Real Jury 

5
EnCJisn juiv 

10 6

Hot Guilty 1 5 0 2
------oiatti"--.. i- vice upre related and the tour occision-maac...------  Iince the cells frequencies , fcipoel 1956) The obtaintd !ets of jud^ents, Cochran's Q was used to analyze the data (Siegel, 1956). T  ̂ j

W3S The^pat tern5of these results suggests that, despite all efforts to equate t*- ,|,r* ,

18



277
jjperience, jurors on the real jury had p different standard for the meaning of "guilty 
^yond 4 reasonable doubt" than those whose decision was not binding on the defendant,

conviction could not send an individual to prison. While such a pattern also may 
jjve resulted from the differences in jury composition, an important point.argues against 
thi»* The En6li-sh or random jury is really a composite of the two other juries, those 
cicused plus those retained, yet the random jury convicted in every-case with a sub- 
,:intial majority favoring conviction on the first vote. Clearly, the only way to 
pliably test the effect of real decision-making in this situation, would be to get 
,t:orneys in a series of cases to select two juries. Both would hear the case, and the >
fJly difference between them would be that one jury would know that it would be deciding 

ictual verdict and the other would be aware that its decision would not affect the 
rt*l outcome of the case.

With this important caveat on the interpretation of this study, the predeliberation 
billots of the jurors on the challenged Jury were compared with the source of their 
iitusal. It- seven of the ten cases, defense attorneys successfully eliminated more Jurors 
¥we favored a guilty verdict than a not guilty verdict, while in two cases equal numbers 
ef friendly and unfriendly jurors were removed, and in one case errors exceeded accurate 
rc,0vals. Despite the generally greater willingness of experimental Jurors to convict, 
t».e prosecutors did almost as well in their jury selection. In five cases they elimi- 
y.cd more jurors who voted for acquittal than for conviction, in two cases their removals 
balanced, and in three others the excused jurors were more likely to vote for conviction, 
ifter the juror selection in each case, attorneys filled out a brief questionnaire 
jfldieating the reasons for each excusal. The msot frequently mentioned reasons were 
■‘cr-canor, occupation, residence, sex, age, and race. Excuses motivated (at least in 
ft) by race were most likely to be correct (89%), followed by demeanor (76%), residence 

(JO'’.), age (67%) and occupation (62%). Sex was least likely to be a successful predictor
{-»)•All conclusions drawn from this study, however, are subject to the same potential 
.'oblcrs of interpretation surrounding the conviction pattern. We do not know the extent 
to which studies of jury behavior are distorted, not merely on likelihood of conviction, 
i. j on other variables and interactions with other variables as well. The differences 
|3 percept ion between real and simulation decision-makers may affect the structures of 
^liberations as well as their outcomes. Yet true field experiments with juries and 
ether decision-makers of interest, although theoretically possible (2eisel and Diamond, 1974), 
jjv be limited by legal, ethical, or practical constraints.

References .
filven, H. 6 Zeisel, The American Jury. Boston: Little, Brown 6 Co., 1967.
Sicgd, S. Konparnmetric Statistics for the Bchaviora1 Sciences, New York: McGraw-Hill, p

1956.
♦tiseli H. 6 Diamond, S. S. "Convincing Empirical Evidence", University of Chicago Law 

Review, Spring, 1974.

>1

19



CERTIFICATE OF SERVICE
A copy of the foregoing has been deposited in

the United States Mail, postage prepaid and addressed as 
follows:

Mr. Jack Boger 
Attorney at Law 
99 Hudson Street 
New York, NY 10019
Mr. Anthony Amsterdam 
40 Washington SQ. S 
New York, NY 10012
Mr. Samuel Gross 
886 Richardson Court 
Palo Alto, CA 94303
Mr. James C. Fuller, Jr.
Attorney at Law 
P. 0. Box 470 
Raleigh, NC 27602
Mr. James E. Ferguson, II 
Mr. Thomas M. Stern 
Attorneys at Law 
951 S. Independence Blvd.
Charlotte, NC 28202
Mr. Adam Stein 
Ms. Ann B. Petersen 
Attorneys at Law 
P. O. Box 1070 
Raleigh, NC 27602
This / day of March______ , 1985.

Special Deputy Attorney General