Expert jury selection consulting by psychologists, sociologists, communication specialists, market researchers, and others, has become big business. The largest consultation firms routinely work on multi-million dollar cases, and may charge the lawyers who use them fees upward of $100,000. [FN2]
Despite this boom, the academic community has remained lukewarm in its evaluation of both the ethics and the efficacy of scientific jury selection. While there are a number of studies linking demographic and personality variables to attitudes, there is much less support for the notion that these variables can be linked to juror verdicts. [FN3] In order to show that scientific jury selection works, this link must be established. And perhaps more importantly, particularly in light of the expense of the scientific jury selection method, it must be shown to be superior to the "traditional" system of unassisted attorney jury selection. The purpose of this paper is (1) to recount and evaluate attorney *230

(Cite as: 17 Ohio N.U. L. Rev. 229, *230)

folklore on jury selection techniques; (2) to review the studies which have examined the premises underlying scientific jury selection; and (3) to suggest some new and more fruitful directions for the use of psychological consultants to aid attorneys in jury selection.

ATTORNEY JURY SELECTION FOLKLORE
The publication of jury selection guides by attorneys for the use of other attorneys has a history of over one hundred years, beginning with Donovan's suggestion to the trial lawyers of his time that they should pay attention to the available background information on potential jurors. [FN4] This information included occupation, age, social status, intelligence, and candor. [FN5] Since then, a number of trial practice writers have published rather detailed and sometimes amusing advice, often replete with racist, sexist, ethnic, age, and other forms of bias. In fact, there is so much advice on what kinds of jurors to select that the average practitioner should have no difficulty finding at least a few salient juror "types" in every case. A brief catalog of these "types" follows.

Occupation
Cornelius advised rejecting jurors with the same occupation as the opposing party in a civil case. [FN6] Heyl suggested avoiding jurors with knowledge in areas in which expert witnesses will testify, since they will feel superior to the expert and to other jurors. [FN7] Appleman felt that farmers award larger amounts in civil cases. [FN8] White stated that the unemployed, pensioners, and those on relief tend to be generous, but that clergy, teachers, and lawyers should generally be avoided by plaintiffs in civil cases. [FN9] Biskind and Keeton liked jurors with the same occupation as the client. [FN10] Bodin believed that businessmen and property owners tend to favor civil defendants. [FN11] Shenker argued that in criminal tax cases, artists, farmers, small business owners, presidents of corporations, and other businessmen are good defense jurors, while engineers, accountants, draftsmen, and vice-presidents of banks or *231

(Cite as: 17 Ohio N.U. L. Rev. 229, *231)

corporations should be avoided. [FN12] Belli disagreed with Appleman, suggesting that farmers are good for the defense in civil cases and for the prosecution in criminal cases. [FN13] He also suggested that civil plaintiffs should look for firemen and policemen and their spouses, waiters, and bartenders, and should reject nurses, physicians, accountants, and statisticians. [FN14] Davis and Wiley advocated rejecting jurors with expertise in fields involved in a civil trial, and ex-policemen and investigators, who are inherently skeptical. [FN15] Adkins stated that overdrawn bank customers are hardened and tend to convict more often. [FN16] Campbell agreed with Adkins, but only in robbery and theft cases; he believed that bankers are good criminal defense jurors in white-collar crime cases. [FN17] Bailey and Rothblatt suggested avoiding retired police, military men, and their spouses unless the criminal defendant is a veteran with a good military record. [FN18] Salesmen, actors, artists, and writers are good criminal defense jurors, since they are experienced at forgiving. [FN19] Cartwright suggested that civil plaintiffs avoid precise or technical types, such as bank tellers or accountants, as they require airtight cases. [FN20] In addition, farmers are tough and unsympathetic and should be avoided, as should nurses, police, and others who are inured to suffering. [FN21]
It might be thought that these recommendations are from old sources no longer used by modern lawyers. However, this is not the case. Many recently published trial practice guides take the same approach. For example, Lane suggested artists, musicians, actors, laborers, carpenters, mechanics, salespersons, office workers, and writers as good civil plaintiff jurors. [FN22] Good civil defense jurors include bankers, bank employees, members of management, low-salaried white collar workers, retired police officers, military men, school teachers, clergymen's wives, utility company employees, insurance representatives or adjusters, farmers, accountants, engineers, professional people generally, tool and die makers, cabinet makers, corporate executives, *232

(Cite as: 17 Ohio N.U. L. Rev. 229, *232)

superintendents, and former court officials. [FN23] Jacobs cautions that nurses are not good for civil plaintiffs because they are too intolerant of pain. [FN24] Wagner suggested that civil plaintiffs look for small businessmen, salesmen, students, social workers, teachers, writers, musicians, and persons on welfare. [FN25] Civil defendants are told to look for accountants, engineers, retired military officers, suburban housewives, and utility workers. [FN26]

Gender
Darrow advised lawyers to avoid women in criminal defense cases. [FN27] On the other hand, Goldstein stated that women are sympathetic and "extraordinarily conscientious" jurors. [FN28] Heyl advocated selecting male jurors when representing a female civil plaintiff, and female jurors when opposing a female plaintiff. [FN29] If the plaintiff is male, he advised the defense to select women. [FN30] If the plaintiff is a child, women are to be avoided by the defense. [FN31] White suggested avoiding housewives, as their experience is limited. [FN32] Biskind suggested avoiding women when seeking large damage awards in civil cases, since they are not used to thinking in large sums. [FN33] Bodin called women "unpredictable," since they are influenced by their husbands' experiences. [FN34] Harrington and Dempsey asserted that women are plaintiff-oriented in disfigurement cases. [FN35] Belli stated that women will be lenient with male criminal defendants, but should be avoided if counsel wants to avoid "intuitive and sympathetic thinking." [FN36] Katz and Karcher both agreed that women make good criminal defense jurors except where the defendant is an attractive female. [FN37] Sparling believed that old women who wear too much makeup are unstable and bad for the state. [FN38]
In the more recent publications, Wagner suggested that suburban housewives are conservative on damages and unsympathetic to plaintiffs. [FN39] *233

(Cite as: 17 Ohio N.U. L. Rev. 229, *233)

With working women, attorneys should look to the occupation. [FN40] "Women's liberationist" women may feel antagonism to male plaintiffs or male attorneys (watch for women who appear "strident, self-assertive, or hostile," or who demand to be addressed as Ms. and not Miss). [FN41] In addition, women are envious of "attractive or successful women" and should be avoided where the plaintiff or his wife is young and attractive. [FN42] However, older women may feel "motherly" to the plaintiff and regard her sympathetically. [FN43] If the "femaleness" of the plaintiff is a factor in the case, female jurors are to be sought, since it is likely that they will identify with the plaintiff. [FN44] Lane stated that the civil defendant should prefer women jurors when the plaintiff is a woman, and where the plaintiff is a young female, older females should be sought. [FN45] Men may also be good jurors for the defense, especially neatly dressed men who are meticulous about their persons. [FN46] Men are preferred when the plaintiff is a woman, but women are preferred when the plaintiff is the breadwinner or a child; however, some women "seem to have an adverse reaction to spinsters, especially if they are attractive." [FN47] Brenner stated that a rape defendant should avoid young women as jurors. [FN48] Finally, Wagner stated that homosexual jurors "are generally thought to be sympathetic and thus, pro-plaintiff," particularly if they appear to have "passive personalities." [FN49] An aggressive male homosexual would be a poor juror where the plaintiff is a woman. [FN50]

Race/Ethnicity
Darrow advised criminal defense attorneys to keep Irishmen, but to strike Englishmen and Germans. [FN51] Goldstein ranked the emotionality of jurors from highest to lowest as follows: Irish, Jewish, Italian, French, Spanish, and Slavic. [FN52] Nordic, English, Scandinavians, and German jurors should be selected if emotional appeals are to be *234

(Cite as: 17 Ohio N.U. L. Rev. 229, *234)

countered. [FN53] Harrington and Dempsey stated that "warm-blooded" Mediterranean jurors, Jews, and the Irish are good for civil plaintiffs, while the Scotch are not ("No McTavish was ever lavish"). [FN54] Cartwright favored Jewish and Irish jurors for civil plaintiffs. [FN55] Lane noted that civil plaintiffs "seem to prefer" blacks, Hispanics, Irish, Jews, French, Italians, and other Mediterraneans, while Germans, English, and Scandinavians are considered to be best for the civil defense. [FN56] Wagner stated that Jews, blacks, Irish, Italians, Hispanics, Puerto Ricans, and other groups that have experienced oppression are sympathetic to plaintiffs. [FN57] Germans, Norwegians, Swedes, Englishmen, and Orientals are thought by some to be better defense jurors. [FN58] Wagner further stated that there may be some exceptions: Irish jurors "could forgive the plaintiff's intoxication . . . a little more easily than the Italian juror could," [FN59] while Jews "are enamored of the medical profession" and thus might not be good plaintiff's jurors in medical malpractice cases. [FN60] If the plaintiff is Japanese, a Chinese juror might not be sympathetic. [FN61] The same applies to Jews and Arabs or Germans, blacks and Puerto Ricans, Irishmen and Englishmen, and Iranians and Iraqis. [FN62]

Demeanor and Appearance
Cornelius suggested that "weak" jurors who would be frightened by a fight should be selected by plaintiffs in assault cases. [FN63] Darrow and Bodin both recommended selecting jurors who smile, especially at you (the attorney). [FN64] Osborn favored jurors with good-sized mouths--especially those with corners that turn up. [FN65] Biskind advised lawyers to avoid dominating types--they would not change their minds and may sway others. [FN66] Keeton and Adkins both advised taking a "leader" who looks favorable and then trying the case to that juror. [FN67] Davis and Wiley, on the other hand, suggested avoiding dominating jurors and evasive jurors. [FN68] Adkins suggested avoiding jurors who appear either *235

(Cite as: 17 Ohio N.U. L. Rev. 229, *235)

too reluctant or too eager to serve. [FN69] Harrington and Dempsey advised avoiding jurors who feel imposed upon and who resent being called for jury duty. [FN70] Plaintiffs should avoid "meticulously dressed" jurors, and smiling jurors who are trying to disarm you. [FN71] They want to get on the jury and "murder you." [FN72] A "fidgety" juror may give you trouble. [FN73] Katz recounted a large number of non-verbal characteristics to avoid: Jurors with crossed feet or arms, clenched fists, poor posture, cocked head, hands in pockets, a kicking foot, those who talk through their teeth, who stand with their hands on their hips or behind their backs, who drum with their fingers, touch their noses, cover their mouths with their hands, or crack their knuckles. [FN74] Jurors who "respond" to you or imitate you are desirable. [FN75]
Bailey, Rothblatt, and Lane have suggested that a round-faced, jovial, heavy person is more desirable for a criminal defendant than a slight, underweight and delicate type. [FN76] The athletic-looking juror is hard to convince, but once convinced will usually "go all the way for you." [FN77] Persons of superior intelligence should be avoided. [FN78] Shenker, on the other hand, favored "intellectuals" and "thinkers" in criminal tax cases, suggesting that "ignorant" people should be avoided. [FN79] Rothblatt thought that people of "medium" intelligence were best, since it is harder to dislodge preconceived ideas of highly intelligent people, and they will influence the other jurors. [FN80] Cartwright and Wagner advised civil plaintiffs to avoid "kingpins," or strong, dominant men. [FN81] Lane advised plaintiffs to avoid the disabled, who may be "embittered" by their misfortune and thus less likely to sympathize with the plaintiff, who is "no worse off than they are." [FN82]

Wealth and Social Status
Darrow advised that wealthy men will convict unless the defendant is accused of a white-collar crime. [FN83] Biskind and Harrington and Dempsey *236

(Cite as: 17 Ohio N.U. L. Rev. 229, *236)

argued that poor people are not used to thinking in large sums and may be better for a civil defendant. [FN84] Keeton stated that white-collar types identify with businessmen and, though used to big sums, will be less likely to give large awards. [FN85] Belli asserted that the upper class favors civil plaintiffs, [FN86] and Adkins that the poor believe they should stick together when in trouble. [FN87]

Religion
Darrow asserted that for criminal defense, Presbyterians are too cold. [FN88] Baptists are even worse. [FN89] Methodists may be considered, but Lutherans (especially Scandinavians) will convict. [FN90] Jewish, Unitarians, Universalists, Congregationalists, and agnostics are good for the defense. [FN91] Appleman disagreed, arguing that religion is not a helpful indicator. [FN92]

Marital Status
Cornelius suggested avoiding bachelors when representing a woman suing for breach of promise. [FN93] Belli argued that married people are more experienced in life and more forgiving, and thus good for civil plaintiffs and criminal defendants. [FN94] Bailey and Rothblatt generally prefer married jurors to single ones, particularly where the criminal defendant is a younger person. [FN95]

Age
Appleman believed that younger persons are better for the defendant, [FN96] while Adkins argued that the elderly are generally lenient, but sometimes severe in criminal matters. [FN97] Bailey and Rothblatt believe that jurors between the ages of twenty-eight and fifty-five are best for the criminal defense; they will tend to be most alert and receptive to *237

(Cite as: 17 Ohio N.U. L. Rev. 229, *237)

complex defenses. [FN98] Younger persons may be favorable jurors if they have acquired a distrust of police officers. [FN99] Jordan suggested that older persons favor civil plaintiffs since they can identify with the experience of aches and pains, but tend to favor lower awards. [FN100] Lane believed that jurors between the ages of thirty and fifty-five were more favorable to the civil plaintiff, while those twenty-one to thirty and over fifty-five were more favorable to the civil defendant. [FN101] The twenty-one to thirty age group "have had less experience and cannot fully appreciate serious permanent injury as do the older above thirty group." [FN102] Those over fifty-five who live on relatively fixed incomes may hesitate to award large verdicts. [FN103]

EVALUATION OF ATTORNEY JURY SELECTION FOLKLORE
What is one to make of these various and sundry pieces of advice? A practitioner will likely find it impossible to follow the advice given, since it is so often conflicting. Should counsel follow Belli and accept policemen, or reject them as recommended by Davis and Wiley? Should counsel avoid women altogether, as Darrow suggests, or merely use them selectively as according to Heyl, Harrington and Dempsey? Should counsel take the smiling juror as per Darrow and Bodin, or reject him as untrustworthy, as articulated by Harrington, Dempsey and Wagner? As a civil plaintiff, are bankers desirable jurors as Belli and Biskind suggest, or undesirable as according to Keeton? Does religious background make a difference, as argued by Darrow, or does it have no impact, as Applemen suggests?
Despite the conflicts in advice, despite the fact that the advice is often based on racial, sexual, ethnic, or other stereotypes, and despite the fact that the advice is based on the idiosyncratic experiences of the advisors rather than on more reliable forms of data, this kind of advice appears to have enduring currency among practicing trial attorneys. One even occasionally finds such stirring testimonials to attorneys' perspicacity as the one by Begam: "Trial attorneys are acutely attuned to the nuances of human behavior, which enables them to detect the minutest traces of bias or inability to reach an appropriate decision." [FN104] Begam goes on to "explain" why attorneys are (in his view) so accurate: There is so much at stake. [FN105] Some might argue that this merely *238

(Cite as: 17 Ohio N.U. L. Rev. 229, *238)

illustrates the "illusion of control" so often found in other contexts. For example, social-psychological experiments show that people often feel that they have control over random events such as the roll of the dice or a lottery drawing. [FN106]
Another factor that may serve to perpetuate these theories is their simplicity and intuitive appeal. Most of the juror "types"--ethnic origin, for example--can be determined on the basis of information which is readily available to trial attorneys, even in the absence of a more extended voir dire than is customary. Still, as Plutchik and Schwartz wryly noted in their evaluation of attorney jury selection folklore,
the suggestion that the Irish are most desirable to the defense leads inevitably to the conclusion that Ireland, where jurors are monotonously Irish, must be utopian for the practice of criminal law . . . . [However,] an examination of the prisons of Ireland reveals that, as a general proposition, the cells are occupied. [FN107]
Of course, attorney selection accuracy is an empirical question. But given the limited feedback that attorneys receive on the accuracy of their selections, it is quite difficult to imagine that the kind of learning Begam suggests could occur. Attorneys sometimes learn how individual jurors on a jury have voted but they almost never learn how a challenged juror might have voted. [FN108] If they do seek information, it is likely to be confirming rather than disconfirming in nature (i.e., a given juror did in fact vote as predicted). [FN109] This type of feedback allows the attorney to maintain his or her pre- existing jury selection "theory" in the face of its actual incorrectness--and to do so with confidence. [FN110] Negative instances are thus rarely encountered, and when they are, it is probably simply a matter of modifying the selection theory slightly to incorporate the exception. If any learning does take place under these conditions, it is likely to resemble the superstitious behavior of animals reinforced with food at fixed or random intervals. [FN111]

*239

(Cite as: 17 Ohio N.U. L. Rev. 229, *239)

EFFICACY OF ATTORNEY JURY SELECTION STRATEGIES
The efficacy of attorney jury selection strategies has been the subject of a number of empirical studies. Each has yielded interesting findings, but most are plagued with severe methodological difficulties as well. [FN112] The studies have addressed several different questions: (1) What juror characteristics are attorneys interested in? (2) What types of jurors are actually challenged by attorneys? (3) What impact do challenges have on jury composition? (4) Do attorneys exercise their challenges effectively?

Attorney Interest in Juror Characteristics
Hayden, Senna, and Siegel focused on information used by district attorneys in voir dire. [FN113] They presented twenty randomly selected prosecutors from four metropolitan Boston counties with seventeen possible categories of information on jurors to be selected for two hypothetical cases. [FN114] The information was selected one category at a time, and after each selection, the attorney could either decide on the suitability of the juror, or ask for more information. [FN115]
Attorneys requested an average of slightly more than seven categories in both cases, with jurors' age, occupation, demeanor, gender, appearance, and residence the most frequently requested categories. [FN116] The only item to vary significantly in rate of request was race; as suggested by the attorney folklore, prosecuting attorneys examined this variable more frequently in the case with a black defendant. [FN117] There was substantial variability among prosecutors in decisions to challenge or to accept the types of information employed, and the weight given to the information, despite the fact that all of the attorneys examined the same juror "profiles." [FN118] Since no defense attorneys were used, it is not possible to examine defense information use or to compare the "models" of defense and prosecution jury selection. In addition, sample size was small, and only two cases were used.

Jurors Actually Challenged
Hawrish and Tate, and Tate, Hawrish and Clark surveyed a random sample of fifty Canadian attorneys who had conducted trials *240

(Cite as: 17 Ohio N.U. L. Rev. 229, *240)

before the Queen's Bench in Saskatoon and Calgary over a one-year period. [FN119] The attorneys were presented with a list of twenty-seven hypothetical jurors described by age, sex, and occupation in a sort of factorial design, and were asked to rate each on a seven-point scale indicating acceptance or rejection of the juror in each of four general types of trials (rape, murder, fraud, and issuing a fraudulent prospectus). [FN120] After completing the initial ratings, they were shown photographs of the six highest-rated jurors, in one of three pre-tested attires (conservative, moderate, or counter-cultural). [FN121] The principal results were that neither juror characteristics nor attire had much influence on attorney ratings of acceptability, [FN122] although severe methodological problems make it very difficult to draw accurate conclusions about the meaning of the results.
In a study of attorney-conducted voir dire, Padawer-Singer, Singer, and Singer reported that jurors selected with voir dire were similar to non-voir dire jurors on a series of attitudinal questions. [FN123] Voir dire jurors believed less in capital punishment, were more in favor of protecting the rights of the accused and of considering a defendant's past mitigating history, believed that law and order necessitate justice for minorities, and were more likely to follow a law whether or not it was fair even if it meant releasing defendants for insufficient evidence or ignoring trial testimony struck from the record. [FN124] Unfortunately, no data are presented on the challenged jurors, nor were these jurors asked to hear the trial and render their verdicts for comparison purposes. [FN125]
A study of jury challenges in the Huey Newton trial revealed that defense attorney Charles Garry tended to eliminate white and male jurors while prosecutor D. Lowell Jensen eliminated blacks and females. [FN126] Garry also challenged twenty of the twenty-seven jurors from predominantly white, conservative middle class suburbs, while Jensen challenged only three. [FN127] On the other hand, Jensen challenged eighteen of the thirty-five that came from Oakland and Berkeley, while Garry challenged only five. [FN128] However, while interesting and somewhat supportive *241

(Cite as: 17 Ohio N.U. L. Rev. 229, *241 )

of the accuracy of attorney folklore, these data are obviously of limited generalizability.
Finally, Penrod, as part of a larger study comparing attorney and scientific jury selection models, looked at the jury selection strategies of nineteen practicing prosecuting and defense attorneys in the Boston area using a set of thirty-two juror "profiles." [FN129] Attorneys were first interviewed about their backgrounds and their usual jury selection practices. [FN130] They were then asked to sort the profiled jurors on the basis of "juror similarity," to rate them for potential bias in the two cases used, to rate them on a series of bipolar scales, and then asked what they recalled about a subset of the profiled jurors. [FN131] Using a multidimensional scaling analysis, it was possible to examine empirically the dimensions used by the attorneys to make selection decisions. [FN132] Penrod concluded that judgments of juror similarity were based primarily on four characteristics: Attitudes toward legal technicalities, gender, ideological orientation, and age. [FN133] In addition, ratings of juror bias were largely determined by these characteristics, indicating that the attorneys were using relatively simple additive models. [FN134] This is consistent with work on judgments by stockbrokers, [FN135] radiologists, [FN136] sentencing judges, [FN137] and others. The best- remembered juror characteristic was occupation, and the more experienced the attorney, the higher the weight attached to occupation. [FN138] Finally, though 

(Cite as: 17 Ohio N.U. L. Rev. 229, *241)

the question of whether the juror characteristics used by the attorneys have any predictive validity could not be answered directly, the relationship between the characteristics used to prepare the juror profiles and the verdict preferences of simulated jurors in another study by Penrod [FN139] showed that none of the individual main effects or two-way interactions accounted *242

(Cite as: 17 Ohio N.U. L. Rev. 229, *242)

for more than about four percent of the variance, and when combined, they accounted for an average of eleven percent of the variance. [FN140]

Impact of Challenges on Jury Composition
In a large-scale study of peremptory challenge practices in New Mexico, Van Dyke found that prosecutors most often challenged jurors who were non- Anglo-European and those between ages thirty and thirty-four. [FN141] On the other hand, defense attorneys most often challenged white jurors and those between ages forty and forty-four. [FN142] There were no differences in the rates of challenging males and females. [FN143] Still, Van Dyke concluded that the net effect of the defense and prosecution challenges on the final composition of juries was minimal, and acknowledged that it is difficult to generalize these findings to other settings (e.g., large cities) or to settings where attorneys have different numbers of challenges available. [FN144]

Efficacy of Challenges
Broeder collected interesting but largely anecdotal reports from interviews of real jurors, focusing on the forms of bias that attorneys failed to detect in voir dire. [FN145] Unfortunately, his findings are not reported in a systematic form, his methods of data analysis are not specified, and his materials are largely anecdotal. However, he did report that, consistent with attorney folklore, "persons with German and British backgrounds were more likely to favor the government, whereas Negroes and persons of Slavic and Italian descent were more likely to vote for acquittal." [FN146] On the other hand, he also concluded that "[v]oir dire was grossly ineffective not only in weeding out 'unfavorable' jurors but even in eliciting the data which would have shown particular jurors as very likely to prove 'unfavorable."' [FN147]
Diamond and Zeisel used a "shadow jury" technique to assess the accuracy of attorneys' peremptory challenges. [FN148] The post-trial juror verdict preferences of groups of peremptorily challenged jurors and *243

(Cite as: 17 Ohio N.U. L. Rev. 229, *243)

randomly selected jurors, all of whom heard actual complete trials along with the real jury, were examined. [FN149] Diamond and Zeisel used a reconstruction technique in order to determine the voting alignments that would have occurred in the "original" jury in the absence of jury challenges, and then used this alignment to predict the likely jury verdict. [FN150] This process was complicated by attrition as well as lack of information in ten of the twelve cases as to individual (rather than overall) distribution of first ballot votes. In the other two, even the overall distribution was unavailable. Although the authors make a relatively convincing case for the accuracy of the reconstruction process, the analysis is built upon so many assumptions that it is difficult to place much confidence in their conclusion. The determination was that non- participating jurors vote in the same proportion as the real jury, and in the two cases in which initial votes are unknown, the initial distribution can be assessed on the basis of the jury's deliberation time. [FN151] Once reconstructed, the verdicts of the original juries were predicted by extrapolating from the relationship between first ballot votes and final verdicts in the 225 juries studied by Kalven and Zeisel. [FN152] Penrod and Hastie have discussed and criticized this process, and offered alternative approaches based on more extensive data. [FN153]
Still, Zeisel and Diamond found very high variability in performance by both defense and prosecuting attorneys across cases. While prosecutors were found to have challenged as many favorably disposed jurors as hostile jurors, defense attorneys challenged somewhat more hostile jurors than favorable ones. [FN154] Overall, challenges based on race were most likely to be correct, with demeanor, residence, age, and occupa less accurate. Gender was judged to be the least accurate indicator. [FN155] Unfortunately, the manner in which these estimates were made was not disclosed, nor was the structure of an optimal selection model or how well such a model might have performed.
In summary, though a number of researchers have made preliminary excursions into the empirical evaluation of attorney voir dire behavior and accuracy, our knowledge in this area is still quite rudimentary. Although most of the studies have qualities that recommend them, many of them also have serious shortcomings--sometimes in method, sometimes in stimulus material, sometimes in data analysis, *244

(Cite as: 17 Ohio N.U. L. Rev. 229, *244)

sometimes in the nature of the sample. [FN156] The great variation in methods and research objectives makes it extremely difficult to collate and generalize from the studies. Broeder's and Zeisel and Diamond's results certainly suggest that attorneys are not very accurate in their selections. [FN157] The studies by Blauner, Padawer-Singer, Singer, and Singer, and Van Dyke suggest that attorneys do exercise their challenges systematically, [FN158] but the results of Hawrish and Tate, Hayden and Penrod suggest that the systematicness may extend only to a few demographic characteristics of limited predictive value. [FN159]

EMPIRICAL TESTS OF THE EFFICACY OF SCIENTIFIC JURY SELECTION
Both direct and indirect examinations of the major assumption of scientific jury selection--that "dispositions and attitudes reliably correlate with juror behavior" [FN160]--have generally indicated that "individual differences are not especially potent predictors of juror verdict preferences." [FN161] Researchers interested in establishing the links between juror verdict preferences and juror attitudes and characteristics have adopted two general strategies. [FN162] Some researchers have surveyed sitting jurors at the conclusion of their jury service and have attempted to relate characteristics to jurors' votes in actual cases. [FN163] Other researchers have adopted an approach in which mock jurors--often experienced or currently sitting jurors-- are presented the same case or cases in an effort to establish links for particular fact patterns. [FN164] Predictive power is typically measured by looking at the percentage of the variance in verdicts that is accounted for by the predictors. [FN165] This is done with a statistical procedure known as multiple regression analysis. [FN166]

*245

(Cite as: 17 Ohio N.U. L. Rev. 229, *245)

Actual Cases
Reed looked at jurors who served in jury trials in East Baton Rouge Parish, Louisiana in 1959-1961. [FN167] Reed contacted 432 jurors, obtaining a fifty- six percent response rate. [FN168] He then looked at, among other things, the relationship between various demographic characteristics and verdict. [FN169] In criminal cases, socioeconomic status, as measured by education and occupational level, was related to verdict: the higher the status, the more likely the juror was to vote "guilty." [FN170] No effect was found for age, marital status, religious preference, or church attendance. [FN171] No significant relationships at all were found for civil cases. [FN172]
Bridgeman and Marlowe looked at a group of sixty-five jurors from Santa Cruz County, California in 1974-75. [FN173] Though their conviction rate was ninety percent, suggesting a potential ceiling effect, and though the sample was small, no demographic correlates of verdict were found. [FN174]
Mills and Bohannon studied a randomly selected sample of 197 criminal jurors in Baltimore. [FN175] They found that guilty verdicts were associated with sex, education, race, and age, with jurors who were female, less educated, black, and older (with some qualifications for males) being generally more conviction- prone. [FN176] They also found that some personality variables derived from Hogan's character-structure view related to conviction (i.e., socialization, empathy, and autonomy). [FN177] A multiple regression analysis showed that between ten and sixteen percent of the variance in verdicts in their cases was predicted by a combination of these variables. [FN178]
Moran and Comfort surveyed 319 felony jurors in Dade County, Florida, looking for links between various demographic and personality variables and verdict. [FN179] They found that sex of the juror was an *246

(Cite as: 17 Ohio N.U. L. Rev. 229, *246)

important moderator of the relationship between personality and demographic variables and post- deliberation verdicts. [FN180] For males, guilty verdicts were associated with lower income, more children, and lower social desirability scores. [FN181] The variance in verdict that was accounted for by these variables was 10.7 percent. [FN182] For females, guilty verdicts were associated only with higher Just World Scale scores, which predicted only 5.9 percent of the variance in verdict. [FN183] Both males and females had higher authoritarianism scores, though this variable did not appear in the regression equation for statistical reasons. [FN184] Moran and Comfort noted, as argued later, that such dispositional and personality variables may have more subtle effects on juror behavior, such as degree of participation in jury deliberation, amount of perceived influence in deliberation, and so forth (though we will suggest cognitive processing of evidence as a more likely locus). [FN185]
Though the data are anecdotal, Tybor discussed an actual case in which Dr. Hans Zeisel was hired by the plaintiff in a civil case as a jury selection consultant. [FN186] A pretrial survey was conducted with a sample of 713 respondents. [FN187] Just before the trial began, but after jury selection, the case settled for $5.5 million. [FN188] Without telling the jury the amount, Judge Alan Morrill asked the jury to try the case anyway, with Zeisel observing. [FN189] The jury deliberated and awarded the plaintiff between three and four million dollars. [FN190] Zeisel stated that the predictions in the jury profile "were quite good" and in particular that the predictions as to which jurors would vote for higher and lower awards was also "good." [FN191]
Moran and Comfort examined attitudes and demographics of over 600 jurors in two studies. [FN192] Overall, juror death penalty attitudes correlated with individual pre-deliberation verdicts. [FN193] Furthermore, a variety of demographic characteristics and attitudes were correlated *247

(Cite as: 17 Ohio N.U. L. Rev. 229, *247)

with death penalty attitudes. [FN194] The death-qualification process in jury selection was challenged in Lockhart v. McCree, [FN195] and despite the fact that the defense was able to present fifteen studies demonstrating a link between death penalty attitudes and conviction-proneness, the Supreme Court was not persuaded that "death-qualified" juries could not impartially decide capital cases. [FN196]
In a study focused on juror decision-making in sexual assault cases, Visher interviewed 331 jurors who had served on thirty-eight different sexual assault cases (a seventy percent response rate). [FN197] Visher assessed jurors' characteristics and attitudes and defendant and victim characteristics, and evaluated the trial evidence. [FN198] She found modest but significant correlations between pre-deliberation verdict preferences and race, education, occupational status, attitudes on crime, and tendency to blame victims. [FN199] In a hierarchical regression analysis, Visher found that evidence accounted for thirty-four percent of the variance in jury verdicts, victim and defendant characteristics eight percent, [FN200] and juror characteristics (when entered last into the analysis) two percent. Unfortunately, this analytic strategy may reverse the causal relationships, at least in part. Because Visher relied upon jurors' characterizations of the parties and the evidence, it is entirely plausible that pre-existing juror differences substantially affected these jurors' perceptions. From this perspective, one might argue that juror characteristics ought to enter the analysis at the first step, rather than the final step. Unfortunately, this cannot be done with the statistics available in the Visher article.

Mock Trial Studies
Simon examined the relationship between juror characteristics and verdicts in two insanity cases. [FN201] She found no significant relationship for juror occupation, gender, income, religion, or age. [FN202] The only difference found was for race. [FN203] Blacks voted to acquit more than jurors of other races, and jurors with less than a high school education voted to acquit more than jurors with a college education. [FN204] These *248

(Cite as: 17 Ohio N.U. L. Rev. 229, *248)

differences would aid jury selection somewhat (and be in line with attorney folklore, at least for race), but education accounted for less than five percent of the variance in verdicts, [FN205] and since education and race are correlated, the predictive power of the two variables cannot be simply added together.
Saks looked at twenty-seven predictor variables, [FN206] and found that the best predictor (an attitudinal question regarding the belief that crime is a product of "bad people" or "bad social conditions") accounted for only nine percent of the variance in juror verdicts, [FN207] and did so in the "wrong" direction (i.e., those who believed that crime was a product of bad social conditions were more likely to vote guilty). [FN208] The four best predictors combined to explain thirteen percent of the variance, [FN209] and none of the others added more than minimally to the variance explained. [FN210]
In interviews of 305 people who had been chosen for jury duty in St. Louis County, Missouri, Hepburn obtained demographic and attitudinal information. [FN211] He then asked them to consider a hypothetical murder case. [FN212] He found verdict associated only with age and prior military service. [FN213] Race, education, military status, and prior criminal victimization were unrelated to verdict. [FN214] The nine variables together explained only eight percent of the variation in verdict. [FN215] Perceived strength of prosecution and defense evidence, however, appeared to be critical mediator variables, [FN216] and Hepburn was able to construct a path analysis linking the attitudinal and demographic variables to the strength of evidence variables, [FN217] and these latter variables in turn to verdict. [FN218] The fact that attitudinal and demographic variables were linked to verdict only indirectly, through cognitive variables, is important, and suggests that an approach focusing on juror cognitions as mediators of verdict (such as that proposed below) could be fruitful.
In their study of the effects of decision rules on jury decision- making, Hastie, Penrod, and Pennington found modest links between *249

(Cite as: 17 Ohio N.U. L. Rev. 229, *249)

demographics and attitudes and verdict preferences. [FN219] Jurors, while on jury duty, participated in the study by viewing a videotaped re- enactment of an actual murder trial. [FN220] Using limited data from the full sample of 828 jurors, Hastie found correlations between verdict preference and employment status, gender, prior criminal jury service, and prior jury service, both criminal and civil. [FN221] A total of 3.2 percent of the variance in verdict preference was accounted for by these variables. [FN222] A subsample of 269 jurors completed more extensive questionnaires and these produced other variables with significant relationships to verdict preference, including wealth of residential area, attitudes regarding punishment of people who take another person's life, marital status, and newspaper preference. [FN223] When considered as a group, these variables accounted for eleven percent of the variance in juror verdict preferences. [FN224]
Cowan, Thompson, and Ellsworth presented a videotaped murder trial to a representative cross-section of adults, including 258 death-qualified and thirty "excludable" jurors (i.e., those whose opposition to the death penalty is strong enough to disqualify them from jury service). [FN225] Death-qualified subjects were more likely to vote guilty, both in the pre-and post-deliberation phases. [FN226] Excludables were more likely to be female and Catholic, [FN227] but in general juror characteristics were uncorrelated with verdict; verdict severity correlated negatively with level of education and prior jury panel experience, and positively with age, sex, and view of the level of subjective certainty that would justify a guilty verdict. [FN228] Unemployed subjects and those who had actually served on a jury were marginally more likely to give harsher verdicts. [FN229] Attitude toward the death penalty and scores on "legal authoritarianism" were the most successful predictors of verdict severity. [FN230]
Penrod assessed various attitudinal and demographic characteristics of 367 actual jurors on jury duty in Boston, and looked at their *250

(Cite as: 17 Ohio N.U. L. Rev. 229, *250)

verdicts in four different simulated trials. [FN231] Penrod regressed twenty-one attitudinal and demographic items over each verdict separately, [FN232] and found that the overall variance explained in verdict by these variables ranged from 4.9 to 14.1 percent. [FN233] No single predictor worked in even as many as three of the cases, [FN234] and the verdicts themselves were only weakly related; [FN235] the highest correlation between any two verdicts was not highly significant. [FN236]
In summary, the studies on the efficacy of attempts to link demographic and personality variables directly to verdict appear to find modest relationships, with the variance explained in verdict preferences appearing to range from approximately five to fifteen percent. It is interesting to note that the mock trial studies, which focus on only one type of case (e.g., murder or rape) and which therefore control for the "noise" introduced by this variable, typically find stronger relationships between demographic and attitudinal variables and verdicts. This suggests that case-specific relationships found in scientific jury selection analyses may not generalize across cases, and may explain why the field studies are somewhat less sensitive. [FN237]
While the obtained estimate of the variance in verdict explained appears low on its face, consider the possible impact of such information in an example given by Penrod and Cutler: [FN238] An attorney operating on a completely random basis with a fifty percent favorable and fifty percent unfavorable jury pool would correctly classify fifty percent of the jurors. [FN239] However, if a jury survey detected a reliable relationship in which five percent of the variance in verdict was accounted for by attitudinal and personality measures, successful use of that information would increase the attorney's performance to sixty-one percent correct classifications. [FN240] With fifteen percent of the variance accounted for, performance would increase to sixty-nine percent correct. [FN241] Clearly, although the percentage of variance explained may be small, the potential improvement in selection performance is not insignificant. If a defendant has his life or millions of dollars at stake, *251

(Cite as: 17 Ohio N.U. L. Rev. 229, *251)

the jury selection advantages conferred by scientific jury selection techniques may well be worth the investment. [FN242]

New Directions in Jury Selection Consultation
It is likely that jury consultants will refine and improve their jury prediction skills, as research methods and the theoretical underpinnings of this enterprise continue to advance. A variety of critiques may be leveled against this enterprise. [FN243] From one point of view, one of the major drawbacks to the commercialization of jury studies is that otherwise sound research may be disappearing into the vaults of research consultants who feel a need to protect trade secrets and other commercial interests. We would like to encourage consultants to bring their research into the public domain in order to enrich our theoretical understanding of jury decision-making, and to provide a sound basis for formulating policies that will ensure equitable application of social science methods. Finally, non-commercial jury research and theorizing remain active and may contain practical applications for legal practitioners.
The more recent published works on jury selection directed at practitioners are psychologically sophisticated and strongly advocate the use of "trial consultants," [FN244] as they are now called (as opposed to the older term "jury selection experts"). The major theme of these recent publications is emphasis on the fact that attorneys cannot simply go from demographic variables directly to predictions of verdict, as the folklore encourages them to do. [FN245] Instead, juror attitudes, personality variables, nonverbal and verbal cues, and jury interactional variables might also be profitably assessed. [FN246] More importantly, there are clear cautions that the "case-specific approach" we discuss above must be taken, as findings regarding predictor variables may not apply across *252

(Cite as: 17 Ohio N.U. L. Rev. 229, *252)

cases or jurisdictions. [FN247] Trial simulations with mock juries or focus groups are particularly endorsed, along with consultation on the formulation of a case theme and voir dire questions. [FN248] Attorney and client dress, demeanor, and speech are also scrutinized and suggestions made for change. [FN249] Almost all of these are considerably beyond the scope of the conventional attorney folklore focus on the demographics of jurors (though the evidence for the efficacy of some of these remains weak). Their use suggests that a broader "case presentation" approach is now being taken by consultants, rather than the narrower "jury selection" approach.
More recently, Pennington has offered a tantalizing view of a new and promising approach to case analysis that, though not fully developed, might be useful in jury selection consulting. [FN250] Pennington has generated persuasive research evidence in support of the "story model" of juror cognitions. [FN251] The story model postulates that in coming to a decision on verdict, jurors first construct a plausible account or scenario of the events in question based on the available evidence, and then match that account to the verdict alternative that best fits. [FN252] The story model has been extended from the murder case originally used by Pennington to a rape case by Olsen- Fulero, Fulero and Wulff. [FN253] In the latter study, forty-eight subjects who had previously completed a battery of attitude and personality scales viewed a videotaped rape trial and completed a post-trial questionnaire asking for various verdict and attributional judgments. [FN254] Using a modified version of Pennington's juror interview in which subjects were asked to arrive at a verdict by discussing out loud what they thought had taken place, Olsen-Fulero, Fulero and Wulff were able to construct coherent "guilty" and "not guilty" stories from subject accounts. [FN255] Interestingly, several pieces of evidence were common to both stories, suggesting that the same piece of evidence may be interpreted differently by jurors. [FN256] In addition, the *253

(Cite as: 17 Ohio N.U. L. Rev. 229, *253)

study found that while the attitude and personality measures did not correlate with verdict directly, several of these measures did correlate with story, which in turn correlated with verdict. [FN257]
The story model offers a promising analytic opportunity to jury selection consultants and to attorneys, as it suggests a case-specific approach to understanding the manner in which demographic, personality, and attitudinal characteristics predispose jurors to process trial-related information and come to verdict. The model further suggests that the "focus group" technique currently used by trial consultants could be modified to include a recounting by each mock juror of the perceived scenario of events that led him to his verdict, and a construction of the two (or more) most plausible verdict stories. Not only do the stories offer a link between the usual jury selection variables that might be detected through public opinion surveys and verdicts, but they also offer a fruitful means of consultation on such important litigation areas as case theme, voir dire questioning, and evidence presentation. While this line of research and application to jury consultation needs refinement, the approach is empirically grounded, and is clearly consistent with the two major and justified cautions previously sounded by critics of the scientific jury selection approach: First, that the nature and strength of the evidence in a given case is the critical variable in predicting verdict, rather than simple demographics, and second, that a case-specific approach is necessary for each trial.
 

FN1. Reprinted with permission of Hemisphere Pub. Corp., N.Y., from Forensic Reports, Vol. 3, p.233 (1990). The article has been revised by the authors from its original publication in Forensic Reports.

 

FNa Professor of Psychology, Sinclair College, Dayton, Ohio. B.A. 1973, University of Maryland; M.A. 1975, University of Oregon; Ph.D. 1979, University of Oregon; J.D. 1979, University of Oregon. Dr. Fulero also maintains private practices in law and psychology.

 

FNaa Professor of Law, University of Minnesota College of Law; B.A. 1969, Yale; J.D. 1974, Harvard University; Ph.D. 1979, Harvard University.

FN2. See Adler, Litigation Science: Consultants Dope Out the Mysteries of Jurors for Clients Being Sued, Wall St. J., Oct. 24, 1989, at 1, col. 1.

 

FN3. See generally M. SAKS & R. HASTIE, SOCIAL PSYCHOLOGY IN COURT 47-71 (1978).

 

FN4. Mariani, Peremptory Challenge--Divining Rod for a Sympathetic Jury, 21 CATH. LAW. 56, 61 (1975) (citing J. DONOVAN, MODERN JURY TRIALS AND ADVOCATES 227 (1887)).

 

FN5. Id. at 61 n.29.

 

FN6. See A. CORNELIUS, TRIAL TACTICS 128 (1932).

 

FN7. Heyl, Selection of the Jury, 40 ILL. B.J. 328, 340 (1952).

 

FN8. See SUCCESSFUL JURY TRIALS: A SYMPOSIUM 148 (J. Appleman ed. 1952) [hereinafter SUCCESSFUL JURY TRIALS].

 

FN9. White, Selecting the Jury, in SUCCESSFUL JURY TRIALS, supra note 8.

FN10. See E. BISKIND, HOW TO PREPARE A CASE FOR TRIAL (1954); R. KEETON, TRIAL TACTICS AND METHODS 251 (2d ed. 1973).

 

FN11. See H. BODIN, SELECTING A JURY 40-41 (1954).

 

FN12. Shenker, How to Try a Criminal Tax Case, 1 CRIM. L. BULL. 27, 30 (Jul. - Aug. 1965).

 

FN13. See generally M. BELLI, MODERN TRIALS (1954).

 

FN14. Id.

 

FN15. See Davis & Wiley, Forty-Nine Thoughts on Jury Selection, 34 DIST. COLUM. L.J. 15, 19 (1967).

 

FN16. Adkins, Jury Selection: An Art? A Science? Or Luck? 5 TRIAL 37, 37 (Dec. - Jan. 1968-1969).

 

FN17. Campbell, The Multiple Function of the Criminal Defense Voir Dire in Texas, 1 AM. J. CRIM. L. 255, 262 (1972).

FN18. See generally F. BAILEY & H. ROTHBLATT, FUNDAMENTALS OF CRIMINAL ADVOCACY (1974) [hereinafter F. BAILEY & H. ROTHBLATT, CRIMINAL ADVOCACY].

 

FN19. Id.

 

FN20. Cartwright, Jury Selection, 13 TRIAL 28, 30 (Dec. 1977).

 

FN21. Id.

 

FN22. See F. LANE, LANE'S GOLDSTEIN TRIAL TECHNIQUES (3d ed. 1984).

 

FN23. Id.

 

FN24. Jacobs, Jury Selection Tips, CAL. TRIAL LAW. at 344, 345 (Dec. 1983).

 

FN25. See W. WAGNER, ART OF ADVOCACY: JURY SELECTION (1989).

 

FN26. See id.

 

FN27. See Darrow, Attorney for the Defense, 8 ESQUIRE 35 (May 1936).

FN28. J. GOLDSTEIN, TRIAL TECHNIQUES (1935).

 

FN29. See Heyl, supra note 7, at 340.

 

FN30. Id.

 

FN31. Id.

 

FN32. White, supra note 9, at 123-24.

 

FN33. See E. BISKIND, supra note 10.

 

FN34. H. BODIN, supra note 11, at 21.

 

FN35. Harrington & Dempsey, Psychological Factors in Jury Selection, 37 TENN. L. REV. 173, 174 (1969).

 

FN36. See M. BELLI, supra note 13.

 

FN37. Karcher, Importance of Voir Dire, 15 PRAC. LAW. 59, 62 (Dec. 1969); Katz, The Twelve Man Jury, 5 TRIAL 39, 39 (Dec.-Jan. 1968-69).

FN38. See Sparling, Jury Selection in a Criminal Case, Tex. Observer, (May 11, 1983).

 

FN39. See W. WAGNER, supra note 25.

 

FN40. See notes 6-26 and accompanying text.

 

FN41. See W. WAGNER, supra note 25, at I-24.

 

FN42. Id.

 

FN43. Id.

 

FN44. Id. at I-25.

 

FN45. See F. LANE, supra note 22.

 

FN46. See id.

 

FN47. Id. at 56.

FN48. See Brenner, Voir Dire and Jury Selection, in CRIMINAL DEFENSE TECHNIQUES (S. Allen, I. Rosen, D. Winston & L. Belfiore eds. 1989).

 

FN49. W. WAGNER, supra note 25. at I-25.

 

FN50. Id.

 

FN51. See generally Darrow, supra note 27.

 

FN52. See I. GOLDSTEIN, supra note 28.

 

FN53. Id.

 

FN54. Harrington & Dempsey, supra note 35, at 175.

 

FN55. See Cartwright, supra note 20.

 

FN56. See F. LANE, supra note 22.

 

FN57. See W. WAGNER, supra note 25.

FN58. See id.

 

FN59. Id. at I-24.

 

FN60. Id.

 

FN61. See generally id.

 

FN62. See generally id.

 

FN63. See generally A. CORNELIUS, supra note 6.

 

FN64. See Darrow, supra note 27. See also H. BODIN, supra note 11, at 47.

 

FN65. See A. OSBORN, THE MIND OF THE JUROR 111-12 (1937).

 

FN66. E. BISKIND, supra note 10.

 

FN67. See R. KEETON, supra note 10, at 251; Adkins, supra note 16, at 39.

FN68. See Davis & Wiley, supra note 15, at 19.

 

FN69. Adkins, supra note 16, at 37.

 

FN70. Harrington & Dempsey, supra note 35, at 178.

 

FN71. Id. at 175-78.

 

FN72. Id. See WAGNER, supra note 25.

 

FN73. Harrington & Dempsey, supra note 35, at 178.

 

FN74. Katz, supra note 37, at 40, 42.

 

FN75. Id. at 42.

 

FN76. See generally F. BAILEY & H. ROTHBLATT, CRIMINAL ADVOCACY, supra note 18. See also F. BAILEY & H. ROTHBLATT, SUCCESSFUL TECHNIQUES FOR CRIMINAL TRIALS (2d ed. 1985) [hereinafter, F. BAILEY & H. ROTHBLATT, SUCCESSFUL TECHNIQUES]; F. LANE, supra note 22.

FN77. See supra note 76.

 

FN78. Id.

 

FN79. See Shenker, supra note 12.

 

FN80. Rothblatt, Techniques for Jury Selection, 2 CRIM. L. BULL. 14, 19 (May 1966).

 

FN81. See Cartwright, supra note 20.

 

FN82. F. LANE, supra note 22, at 56.

 

FN83. See Darrow, supra note 27.

 

FN84. See E. BISKIND, supra note 10; Harrington & Dempsey, supra note 35.

 

FN85. R. KEETON, supra note 10, at 252.

 

FN86. See M. BELLI, supra note 13.

FN87. Adkins, supra note 16, at 37.

 

FN88. See Darrow, supra note 27.

 

FN89. See id.

 

FN90. See id.

 

FN91. See id. For additional advice on the Jewish religion, see supra notes 51- 62 and accompanying text.

 

FN92. SUCCESSFUL JURY TRIALS, supra note 8, at 128.

 

FN93. A. CORNELIUS, supra note 6, at 127.

 

FN94. See M. BELLI, supra note 13.

 

FN95. See F. BAILEY & H. ROTHBLATT, CRIMINAL ADVOCACY, supra note 18; F. BAILEY & H. ROTHBLATT, SUCCESSFUL TECHNIQUES, supra note 76.

 

FN96. SUCCESSFUL JURY TRIALS, supra note 8, at 127.

FN97. Adkins, supra note 16, at 37.

 

FN98. See supra note 95.

 

FN99. Id.

 

FN100. See W. JORDAN, JURY SELECTION (1980).

 

FN101. See F. LANE, supra note 22.

 

FN102. Id. at 57.

 

FN103. See id.

 

FN104. Begam, Voir Dire: The Attorney's Job, 13 TRIAL 3, 3 (Mar. 1977).

 

FN105. Id.

 

FN106. See E. LANGER, THE PSYCHOLOGY OF CONTROL 59-63 (1983).

FN107. Plutchik & Schwartz, Jury Selection: Folklore or Science? 1 CRIM. L. BULL. 3, 5 (May 1965).

 

FN108. See Diamond & Zeisel, A Courtroom Experiment on Juror Selection and Decision-Making, 1 PERSONALITY & SOC. PSYCHOLOGY BULL. 276 (1974); Zeisel & Diamond, The Effect of Peremptory Challenges on Jury and Verdict: An Experiment in a Federal District Court, 30 STAN. L. REV. 491, 502 (1978).

 

FN109. Wason, On the Failure to Eliminate Hypotheses in a Conceptual Task, 12 Q.J. EXPERIMENTAL PSYCHOLOGY 129, 129 (1960).

 

FN110. Lord, Ross & Lepper, Biased Assimilation and Attitude Polarization: The Effects of Prior Theories on Subsequently Considered Evidence, 37 J. PERSONALITY & SOC. PSYCHOLOGY 2098, 2108 (1979).

 

FN111. See Skinner, Superstition in the Pigeon, 38 J. EXPERIMENTAL PSYCHOLOGY 168 (1948).

 

FN112. See Plutchik & Schwartz, supra note 107.

 

FN113. See Hayden, Senna, & Siegel, Prosecutorial Discretion in Peremptory

Challenges: An Empirical Investigation of Information Use in the Massachusetts Jury Selection Process, 13 NEW ENG. L. REV. 768 (1978).

 

FN114. Id. at 781-82.

 

FN115. Id. at 783.

 

FN116. Id. at 784.

 

FN117. Id. at 785.

 

FN118. Id.

 

FN119. Tate, Hawrish & Clark, Communication Variables in Jury Selection, 24 J. COMM. 130, 135 (1974).

 

FN120. Id. at 136.

 

FN121. Id.

 

FN122. Id. at 137.

FN123. See generally Padawer-Singer, Singer & Singer, Voir Dire by Two Lawyers: An Essential Safeguard, 57 JUDICATURE 386 (1974).

 

FN124. Id. at 391.

 

FN125. See id.

 

FN126. See A. GINGER, JURY SELECTION IN CRIMINAL TRIALS (1975).

 

FN127. See id.

 

FN128. See id.

 

FN129. S. Penrod, Study of Attorney and "Scientific" Jury Selection Models (1979) (unpublished doctoral dissertation, available at Harvard University) [hereinafter S. Penrod, Jury Selection Models].

 

FN130. See id.

 

FN131. See id.

FN132. See generally Carroll & Chang, Analysis of Individual Differences in Multidimensional Scaling Via an N-Way Generalization of "Eckart-Young" Decomposition, 35 PSYCHOMETRIKA 283 (Sept. 1970).

 

FN133. See S. Penrod, Jury Selection Models, supra note 129.

 

FN134. See id.

 

FN135. See Slovic, Analyzing the Expert Judge: A Descriptive Study of a Stockbroker's Decision Processes, 53 J. APPLIED PSYCHOLOGY 255 (1969).

 

FN136. See Hoffman, Slovic, & Rorer, An Analysis-of-Variance Model for the Assessment of Configural Cue Utilization in Clinical Judgment, 69 PSYCHOLOGICAL BULL. 338 (1968).

 

FN137. See Ebbesen & Konecni, Decision Making and Information Integration in the Courts: The Setting of Bail, 32 J. PERSONALITY & SOC. PSYCHOLOGY 805 (1975).

 

FN138. See generally id.

FN139. See Penrod, Predictors of Jury Decision Making in Criminal and Civil Cases: A Field Experiment, 3 FORENSIC REP. 261 (1990).

 

FN140. See generally id.

 

FN141. J. VAN DYKE, JURY SELECTION PROCEDURES 157 (1977).

 

FN142. Id.

 

FN143. Id.

 

FN144. Id.

 

FN145. See Broeder, The University of Chicago Jury Project, 38 NEB. L. REV. 144 (1959) [hereinafter Broeder, Jury Project]; Broeder, Voir Dire Examination: An Empirical Study, 38 S. CAL. L. REV. 503 (1965) [hereinafter Broeder, Empirical Study].

 

FN146. Broeder, Jury Project, supra note 145, at 748.

 

FN147. Broeder, Empirical Study, supra note 145, at 505.

FN148. See Zeisel & Diamond, supra note 108, at 498. See also Diamond & Zeisel, supra note 108.

 

FN149. Diamond & Zeisel, supra note 108.

 

FN150. See generally id.

 

FN151. See generally id.

 

FN152. H. KALVEN, JR. & H. ZEISEL, THE AMERICAN JURY 487-89 (1966).

 

FN153. See Penrod & Hastie, Models of Jury Decision-Making: A Critical Review, 86 PSYCHOLOGICAL BULL. 462 (1979).

 

FN154. Zeisel & Diamond, supra note 108, at 513-18.

 

FN155. See generally id.

 

FN156. See generally id.

FN157. Broeder, Empirical Study, supra note 145, at 505; Zeisel & Diamond, supra note 108, at 517.

 

FN158. See J. VAN DYKE, supra note 141, at 146; Padawer-Singer, Singer & Singer, supra note 123, at 386. See also A. GINGER, supra note 126.

 

FN159. See Hawrish & Tate, Determinants of Jury Selection, 39 SASK. L. REV. 285 (1974-1975); Hayden, Denna & Siegel, supra note 113, at 777-78; S. Penrod, Jury Selection Models, supra note 129.

 

FN160. See Penrod & Cutler, Assessing the Competence of Juries, in HANDBOOK OF FORENSIC PSYCHOLOGY 297 (I. Weiner & A. Hess eds. 1985) [hereinafter HANDBOOK OF FORENSIC PSYCHOLOGY].

 

FN161. Id. at 298. See Davis, Bray & Holt, The Empirical Study of Decision Processes in Juries, A Critical Review, in LAW, JUSTICE, AND THE INDIVIDUAL IN SOCIETY: PSYCHOLOGICAL AND LEGAL ISSUES 326-61 (J. Tapp & F. Levine eds. 1977); Hans & Vidmar, Jury Selection, in THE PSYCHOLOGY OF THE COURTROOM 47 (N. Kerr & R. Bray eds. 1982); Saks, The Limits of Scientific Jury Selection: Ethical and Empirical, 17 JURIMETRICS J. 3 (1976).

FN162. HANDBOOK OF FORENSIC PSYCHOLOGY, supra note 160, at 293.

 

FN163. Id. at 296.

 

FN164. Id. at 297-98.

 

FN165. See, e.g., id. at 298-302.

 

FN166. Saks, supra note 161, at 7.

 

FN167. Reed, Jury Deliberation, Voting and Verdict Trends, 45 SW. SOC. SCI. Q. 361, 362 (1965).

 

FN168. Id.

 

FN169. See id. at 365-70.

 

FN170. Id. at 366.

 

FN171. Id. at 367.

FN172. Id.

 

FN173. Bridgeman & Marlowe, Jury Decision Making: An Empirical Study Based on Actual Felony Trials, 64 J. APPLIED PSYCHOLOGY 91, 91-92 (1979).

 

FN174. Id. at 94.

 

FN175. Mills & Bohannon, Juror Characteristics: To What Extent Are They Related To Jury Verdicts?, 64 JUDICATURE 22, 26 (1980).

 

FN176. Id. at 27.

 

FN177. See R. HOGAN, PERSONALITY THEORY: THE PERSONOLOGICAL TRADITION (1976).

 

FN178. Mills & Bohannon, supra note 175, at 29-31.

 

FN179. Moran & Comfort, Scientific Juror Selection: Sex as a Moderator of Demographic and Personality Predictors of Impaneled Felony Jury Behavior, 41 J. PERSONALITY & SOC. PSYCHOLOGY 1052 (1982) [hereinafter, Moran & Comfort, Scientific Juror Selection].

FN180. See id.

 

FN181. See id.

 

FN182. See id.

 

FN183. See id.

 

FN184. See id.