Skip to content
Search resources Open menu

Jurors and Jury Decision-Making

Mock Jury Research: Methods, Realism, and Limits

Published

Snap Shot

  • The question: How much weight can a court give to mock-jury studies, the main source of experimental evidence about how jurors reason?
  • What the research shows: A 2017 meta-analysis of 53 studies found student and non-student samples did not differ in guilty verdicts, culpability ratings, or damage awards, although a 2026 study found the two kinds of samples differed in attitudes about sexual assault. Participants’ concern about appearing prejudiced can also shape results in studies of race. For some questions realism matters: in the pretrial publicity meta-analysis, effects were larger in more realistic studies, which suggests that, for some questions, low-realism studies can underestimate effects; the direction of any bias depends on the question.
  • What it does not show: Mock studies are strongest at showing that a factor can influence judgments, and weaker at estimating how large or frequent an effect is in a real trial. No mock study can show that a particular juror was influenced or a particular verdict was wrong, and consequentiality cannot be tested experimentally. Real-juror studies are rare, and no published study identified for the paper simulates a court-martial panel.
  • The law: Federal, military, and Arizona no-impeachment rules and 18 U.S.C. § 1508 sharply limit study of real deliberations (Tanner, Warger, Peña-Rodriguez). Ballew relied on jury research; Lockhart noted its limits while assuming the studies were valid.
  • Issues for each side: Prosecutors are likely to encounter convergent findings offered for instructions or educational testimony and single studies cited by the defense; defense counsel are likely to encounter mock-jury findings on venue, voir dire, and limiting instructions and studies offered for educational testimony; for each side the methods behind a study bear on its weight.

Most of what is known about how jurors reason comes from studies in which people who are not serving on a jury decide a case that is not real. Researchers present a trial, in writing, on video, or as a live re-enactment, to students, community members, or online participants, vary one feature of the case, and measure whether verdicts or judgments change. Because deliberations in real cases are closed to observation in almost every American jurisdiction, these mock-jury studies are the main source of experimental evidence about juror behavior, including the research on pretrial publicity, instructions, expert testimony, and bias discussed throughout this hub.

The quality of that evidence reaches court in several ways. Parties cite mock-jury research in motions for change of venue, for expanded voir dire, to exclude or limit evidence, and in support of or against expert testimony about juror beliefs. Appellate courts, including the Supreme Court, have relied on such research or declined to rely on it. A court asked to act on a study needs to know what the study did, how closely its conditions resemble a trial, and what it can and cannot establish.

This paper describes how mock-jury studies are designed, reviews the research on whether their realism affects their results, describes the small body of research on real jurors and the legal rules that restrict it, examines a recent dispute over real-juror and mock-jury evidence as a case study in method, and explains what such research can and cannot show a court. It concludes with the governing federal, military, and Arizona law.

How Mock-Jury Studies Work

Participants

Early mock-jury research relied heavily on undergraduate students, who are inexpensive and available but younger, more educated, and less varied in life experience than jury pools. Researchers increasingly use community samples recruited through advertisements, jury pools, or electoral and voter rolls, and online panels such as Amazon Mechanical Turk, Prolific, and survey firms that build samples to match census characteristics. Online panels permit large samples and careful experimental control at low cost. Their participants self-report jury eligibility, are paid volunteers, may be inattentive, and may differ from jury pools in education and ideology; one recent Prolific sample (N = 1,462) matched census figures on age, sex, and ethnicity but was more educated and more liberal than the population, and its subgroup findings were small-sample findings (Huang, 2025; see the Juror Demographics and Political Identity papers). These are group-level averages from mock-jury studies and do not apply to any individual juror. A replication project that repeated three well-known law-and-psychology experiments on Mechanical Turk, other national online platforms, and in-person laboratory samples partially replicated all three, found broadly similar responses across platforms, and found Mechanical Turk participants, if anything, more attentive than the other samples; the authors also cautioned that materials must be updated for replication to be meaningful (Irvine et al., 2018).

Every recruitment method involves self-selection. Real jurors are summoned and are required to serve, whereas mock jurors volunteer. In a survey of serving jurors in England and Wales, 87% said they would have opted out of service had it been voluntary, which the author argued means volunteers may not represent the people who actually decide cases (Thomas, 2020). Whether that difference changes how people evaluate evidence is an empirical question that has not been directly tested.

Trial Materials and Deliberation

Studies range from written vignettes of a page or two, through audio or video recordings, to filmed trials and live re-enactments using actors, lawyers, and sometimes judges in a courtroom. Shorter materials allow researchers to isolate a single variable and to test many conditions; longer materials include cross-examination, instructions, and the competing presentations that characterize trials. The duration of even the most realistic studies is usually measured in hours, not the days or weeks of many trials. In one online study, 139 Mechanical Turk participants who watched a video or read a transcript of the same trial did not differ in verdicts, but those who watched the video did worse on checks of whether they had attended to the trial (Maeder et al., 2024); the sample was small.

Many studies measure individual judgments without deliberation. Others assemble participants into groups that deliberate to a verdict, sometimes recorded and analyzed for content. Deliberation is costly and reduces statistical power, because the unit of analysis becomes the jury rather than the juror, but it is also the step that most distinguishes jury decisions from individual judgments. Researchers also vary the outcome measured: a dichotomous verdict, a rating of the probability of guilt, a rating of credibility, a sentence or damages award, or reasons written in the participant’s own words. Deliberation can now be studied online as well. In 54 mock juries of community members who watched the same civil trial, virtual and in-person juries did not differ on most measures, although in-person juries interrupted one another more and discussed more topics, and the virtual jurors were younger, wealthier, and more educated (Reed et al., 2026). In every design, participants know their decision has no consequences for a real person.

Does Realism Matter? The Ecological Validity Research

Concern about generalizing from simulations is as old as the method. Reviews in the late 1990s identified the main dimensions of realism: the sample, the medium of presentation, the completeness of the trial, the presence of deliberation, and the consequentiality of the decision (Diamond, 1997; Bornstein, 1999). Bornstein (1999) reviewed the research comparing student with non-student participants and written with more realistic presentations, and analyzed the simulations published in Law and Human Behavior over its first twenty years. Most of those studies used students and written materials, and the methods had become less realistic over time, but few studies comparing samples or media had found differences. Bornstein concluded that the trend was not necessarily cause for concern, while noting how little research had addressed the question. Diamond (1997) took a complementary view: simulations are a substantial improvement over speculation about juries, and more realistic features improve both the quality and the persuasiveness of the results, but the field lacked the theory and data needed to know when the more elaborate and costly features actually change the answers.

A 2017 meta-analysis addressed the sample question with a larger body of research: 53 studies (40 criminal, 13 civil) with 17,716 participants (Bornstein et al., 2017). Student and non-student samples did not differ in guilty verdicts, culpability ratings, or damage awards. Differences appeared for sentences and for liability judgments, but they were small or ran in contradictory directions, and the manner in which the trial was presented was the one moderator with a notable effect. The authors concluded that the results may ease concerns about student samples. Sample type is only one dimension of realism, however, and the meta-analysis could not address consequentiality. A later single study suggests the reassurance may not extend to attitude-dependent topics: when the same sexual assault vignette experiment was run with students and Mechanical Turk workers, the samples differed in data quality, demographics, acceptance of rape myths, and victim blaming, and the authors concluded that results depend on sample source (St George & Osei Arhin, 2026). It examined perceptions in one vignette, not verdicts across cases, and does not overturn the meta-analysis.

A special issue of Behavioral Sciences & the Law in 2011 collected work on these questions. Its editors argued that the realism of a study is a matter of degree rather than a yes-or-no question, that the needed level of realism depends on the question asked, and that the field would benefit from programs of research that test whether findings from simpler designs hold in more realistic ones (Wiener et al., 2011). A later survey of 74 people who conduct, review, and publish jury research found that respondents gave more weight to internal validity than to ecological validity, rated the use of theory as the most important feature of a study, rated jury instructions as the most important trial element to include, and gave the least support to requiring deliberation (Lieberman et al., 2016). The authors suggested that such standards could help courts judge whether jury research methods are generally accepted. The survey describes what researchers value, not what courts have required.

There is some evidence that realism matters for particular questions. In the pretrial publicity meta-analysis, effects were larger in studies that drew participants from the potential juror pool and used real rather than simulated publicity, and smaller in studies with student participants (Steblay et al., 1999). In one review of research on educating jurors about sexual assault, studies that found no effect of judicial directions or expert testimony tended to be those with lower realism, such as written vignettes, student samples, or no deliberation, while those finding effects tended to be more realistic or to deliver education before the evidence (Leverick et al., 2025); the review’s authors are also on one side of the dispute described below, and realism is one of several differences among the studies. Reviews find small and variable effects overall (Flynn Hudspith et al., 2024; Walters et al., 2025); in the Walters meta-analysis, expert testimony and juror education were more effective than judicial instruction for biases about complainants. These patterns suggest that low-realism studies can underestimate as well as overestimate effects, and that whether a finding generalizes must be assessed question by question.

Realism is not the only threat to validity. In two experiments with 1,018 and 1,253 White online participants, most of those who judged a Black defendant guessed that the study concerned racial bias, and they convicted Black defendants less often than participants who guessed otherwise. In one of the studies, participants first given the chance to acquit a Black defendant in an earlier case were then more likely to convict a Black defendant, a pattern that appeared among participants who identified as conservative but not among those who identified as liberal (Salerno et al., 2023); it is a single study of online participants and does not apply to any individual juror. The authors concluded that concern about appearing prejudiced may inflate favorable judgments of Black defendants in experiments. Laboratory studies of race can therefore understate bias as well as overstate it.

Studying Real Jurors

Research on real jurors is limited by law and by practice. In the 1950s, researchers associated with the University of Chicago Jury Project recorded the deliberations of several civil juries in federal court in Wichita, Kansas, with the consent of the judge and counsel but without the jurors’ knowledge. The disclosure prompted congressional hearings and a federal statute making it a crime to record, listen to, or observe the deliberations of a federal grand or petit jury (18 U.S.C. § 1508). Real-juror research since then has relied on other methods: questionnaires completed by judges, attorneys, and jurors after verdict, as in the judge-jury agreement studies and field studies of evidence strength (Eisenberg et al., 2005; Devine et al., 2009); post-trial interviews, such as a New Zealand study of 121 jurors from 18 sexual violence trials (Tinsley et al., 2021); and surveys of serving jurors, such as the Thomas (2020) survey described below.

The most direct look at real deliberations in the United States came from Arizona. With the approval of the Arizona Supreme Court, at the invitation of the Pima County Superior Court, and with the consent of the parties, attorneys, and jurors, researchers videotaped the jurors’ discussions during trial and their final deliberations in 50 civil trials in Pima County between 1998 and 2001. The project evaluated one of Arizona’s mid-1990s jury reforms, then codified in Arizona Rule of Civil Procedure 39(f), which allowed civil jurors to discuss the evidence among themselves in the jury room before final deliberations; jurors in 37 trials were permitted such discussions and jurors in 13 trials were not (Diamond et al., 2003a, 2003b). The recordings also allowed analysis of the written questions jurors submitted to witnesses, which the researchers described as a direct window into what jurors were trying to understand during trial (Diamond, Rose, Murphy, & Smith, 2006), and of how juries functioned under Arizona’s non-unanimous civil verdict rule (Diamond, Rose, & Murphy, 2006). The researchers reported that jurors discussed the trial extensively, that the discussions often filled gaps in knowledge, corrected misunderstandings, and improved recall of the evidence. They also documented problems: jurors often discussed the case without all members present, contrary to the rule, and some stated early positions on the verdict (Diamond et al., 2003a). The project’s limits are equally important: it involved civil rather than criminal cases, a single county, and jurors who knew they were being recorded.

In England and Wales, where disclosing or soliciting information about deliberations is restricted by statute, researchers have used real jurors in other ways. A Ministry of Justice study had people summoned for jury service at several Crown Courts watch a filmed trial and deliberate, combining the realism of the sample with the control of a simulation, and supplemented this with an analysis of court records and a post-verdict survey (Thomas, 2010, p. i). The design shows that the line between mock and real juries is not sharp: the participants were real jurors, but the case was simulated.

A Case Study in Method: The Thomas Debate

A recent dispute over juror beliefs in sexual offense cases illustrates how methodological choices shape conclusions. Thomas (2020), the largest survey of real jurors this review identified, surveyed 771 of 773 jurors (99.7%) on 65 juries in England and Wales immediately after verdict. Few agreed with crude misconceptions about rape, while larger proportions agreed with or were unsure about items concerning demeanor, intoxication, and false allegations. Thomas concluded that claims of widespread juror bias in sexual offense cases were not valid and that mock-jury studies and public polls cannot stand in for real jurors.

Researchers who rely on mock-jury and deliberation studies responded. Chalmers, Leverick, and Munro (2021) argued that the conclusions were broader than the data and defended mock-jury research against the claim that participants behave differently from real jurors. A commentary by eleven researchers argued that the survey used unvalidated items with a three-point scale and no measure of social desirability, that a post-verdict attitude survey still measures abstract beliefs rather than reasoning in deliberation, and that a near-complete response rate inside a court may reflect institutional pressure (Daly et al., 2023). They pointed to mock-jury research in which participants who rejected misconceptions in the abstract applied them when discussing a concrete case. A counter-critique argued that the claim of overwhelming evidence of juror prejudice rests on mock-jury methods with recognized limits and on attitude scales that are themselves contestable (Waiton, 2024).

Each side identifies real limitations of the other’s method. The real-juror survey had an unusually representative sample and near-complete participation, but it did not observe deliberations or connect beliefs to verdicts. The mock-jury studies observe deliberations and connect attitudes to verdicts, but with volunteers, shortened trials, and no real consequences. Researchers have tried to narrow the gap: a live three-hour re-enactment with participants drawn by random sampling from the electoral register found that attitude scores predicted verdicts before and after deliberation, consistent with less realistic studies (Willmott et al., 2026). A video recreation of a genuine contested rape trial, decided by 13 twelve-person mock juries (156 participants), likewise found acceptance of rape myths associated with not-guilty votes before and after deliberation (Willmott & Woodhams, 2026); the participants were still volunteers deciding a case with no consequences. The most recent meta-analysis found that such attitudes were strongly associated with blaming complainants but only weakly with verdicts (Bouillaut et al., 2026). Several authors of the critique (Daly et al., 2023) also conducted some of the mock-jury studies and reviews under discussion, and Thomas’s conclusions rest on her own survey; readers weighing the competing claims are best served by the primary studies. The substance of the debate is addressed in the Counterintuitive Testimony resource; its methodological lesson is that no single method answers all of the questions a court may ask.

What Mock Studies Can and Cannot Show a Court

Mock-jury studies are strongest at establishing that a factor can influence judgments under controlled conditions. Random assignment allows causal conclusions that field studies cannot support, and a body of consistent findings across samples and designs, such as the meta-analyses of pretrial publicity and instructions to disregard inadmissible evidence (Steblay et al., 1999; Steblay et al., 2006), provides a reasonable basis for concluding that an effect is real. Mock studies are also the only way to compare procedures, such as alternative instructions, before they are adopted.

They are weaker at estimating how large an effect will be in a real trial, how often it occurs, or whether it is offset by cross-examination, instructions, and deliberation. Effect sizes obtained from vignettes without deliberation cannot be assumed to apply to a full trial. And no mock study can show that a particular juror in a particular case was influenced by a particular factor, or that a particular verdict was wrong.

Courts have taken both approaches. In Ballew v. Georgia (1978), Justice Blackmun’s lead opinion relied extensively on empirical research, including mock-trial studies comparing six- and twelve-person groups, in holding that a five-person criminal jury is inadequate; it also gave little weight to two simulations it found methodologically weak, and Justice Powell’s concurrence objected to the opinion’s heavy reliance on statistical studies. In Lockhart v. McCree (1986), the Court examined studies offered to show that death-qualified juries are more prone to convict and identified several limitations: only some measured guilt judgments at all, the more recent studies used members of the public rather than people sworn to decide a real case, some did not include deliberation, and most did not account for jurors who could not fairly decide guilt. The Court then assumed, for purposes of its opinion, that the studies were valid, and held that death qualification does not violate the Constitution in any event. The two decisions illustrate that the weight a court gives to jury research depends both on its methods and on how the legal question is framed.

Limits of the Evidence

The research on research has its own limits. Studies comparing samples and media are fewer than the studies they evaluate, and they can test only the dimensions of realism that researchers can vary; consequentiality, the knowledge that a real person’s liberty depends on the verdict, cannot be tested experimentally. Real-juror studies are rare, often decades old, mostly civil or non-American, and rely on self-report or on jurors who know they are being observed. Online panels are relatively new, change over time, and may differ from jury pools in ways that are not captured by demographic matching. Publication practices may favor studies that find effects. Very few mock studies use military participants, and no published study identified for this paper simulates a court-martial panel with members selected under Article 25.

Questions to Ask About This Research

When the 2017 meta-analysis is cited to show that student samples decide cases like others (Bornstein et al., 2017), ask what exactly was counted: across 53 studies and 17,716 participants, samples did not differ in guilty verdicts, culpability ratings, or damage awards, but differences appeared for sentences and liability judgments, and sample type is only one dimension of realism; a later study found student and online samples differed on attitudes about sexual assault (St George & Osei Arhin, 2026). When a single online or vignette study is offered for a venue change or expanded voir dire, ask who was studied and how much evidence it rests on, as with a Prolific sample that matched census figures on age, sex, and ethnicity but was more educated and more liberal than the population (Huang, 2025). When the survey of 771 real jurors is offered against mock-jury findings (Thomas, 2020), ask whether the conclusion follows from the design, since it measured beliefs without observing deliberation or connecting beliefs to verdicts. When the publicity meta-analysis is cited to show that realistic studies find larger effects (Steblay et al., 1999), ask whether it applies to the question at hand, because generalization must be assessed question by question. These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.

Mock Jury Research and the Law

Federal and Military Standards

Federal Rule of Evidence 606(b) bars a juror, during an inquiry into the validity of a verdict, from testifying about statements made during deliberations, the effect of anything on any juror’s vote, or any juror’s mental processes concerning the verdict. It allows testimony about whether extraneous prejudicial information was improperly brought to the jury’s attention, whether an outside influence was improperly brought to bear on any juror, and whether there was a mistake in entering the verdict on the verdict form. In Tanner v. United States (1987), the Supreme Court, by a five-to-four vote on this issue, held that the rule barred juror testimony that other jurors had used alcohol and drugs during trial, reasoning that intoxication is not an outside influence and that the jury system is protected by other safeguards, including voir dire, observation by the court and counsel during trial, jurors’ ability to report misconduct before verdict, and non-juror evidence. In Warger v. Shauers (2014), the Court held that the rule applies to evidence offered to show that a juror lied during voir dire. In Peña-Rodriguez v. Colorado (2017), the Court recognized a Sixth Amendment exception when a juror makes a clear statement indicating reliance on racial stereotypes or animus to convict.

These rules, together with 18 U.S.C. § 1508, are the principal reason research on American juries relies on simulations. They protect the finality of verdicts, the candor of deliberation, and jurors from harassment, and they also mean that the evidence courts receive about how jurors think comes mostly from mock studies.

Military Rule of Evidence 606(b) follows the federal rule, applied to inquiries into the validity of a finding or sentence, but its outside-influence exception expressly includes unlawful command influence, reflecting a concern specific to the military justice system; it also permits testimony about a mistake in entering the finding or sentence on the forms. When research on jurors is offered through an expert, the expert’s testimony is subject to Federal Rule of Evidence 702 and Military Rule of Evidence 702, and the features of study design described above bear on whether the research is reliable and fits the question before the court. Ballew and Lockhart illustrate how courts weigh such research in deciding legal questions.

Arizona

Arizona divides the subject between its evidence and procedure rules. As currently published, Arizona Rule of Evidence 606(b) is framed for inquiries into the validity of a verdict in a civil case and tracks the federal rule, barring juror testimony about deliberations, the effect of anything on a juror’s vote, and jurors’ mental processes, subject to the same three exceptions. In criminal cases, the governing provision is Rule 24.1(d) of the Arizona Rules of Criminal Procedure, which permits testimony about the conduct of jurors, court officials, or third persons but not about the subjective motives or mental processes that led a juror to agree or disagree with the verdict. The Arizona Jury Project shows that the state’s courts have permitted, by special approval and with consent, research that would otherwise be impossible; the project was conducted to evaluate a reform in Arizona’s civil rules permitting jurors to discuss evidence during trial (former Ariz. R. Civ. P. 39(f)). That research remains among the few direct observations of American jury deliberations.

Special Issues in Legal Practice

The principles described above apply with equal force to every party in a legal proceeding. The science operates the same way regardless of who is asking the questions, and the research summarized here does not by itself favor either side. What differs between the parties is not the science but the vantage point — the particular places where these issues tend to surface, and the particular errors each party is best positioned to guard against. The two summaries below identify those recurring issues. They are offered for orientation, not as litigation strategy; in every case the controlling question is what the evidence in that case will, and will not, support, and the expert’s role is to inform the trier of fact rather than to advocate for either side.

Special Issues for Prosecutors

The ecological validity research supports reliance on well-replicated mock-jury findings for limited purposes, and the reviews and meta-analysis comparing student with community samples found few differences, and those small, in guilt judgments. Mock studies also supply evidence that some juror misconceptions persist into deliberation and that some forms of education or instruction can change how jurors discuss evidence. Prosecutors are likely to encounter convergent findings across methods when instructions or educational expert testimony are proposed, and claims remain confined to what the studies measured.

The research also counsels caution. When the defense cites a single online or vignette study to support a venue change, an expanded voir dire, or the exclusion of evidence, the appropriate response is to examine its sample, materials, and measures rather than to dismiss mock-jury research generally, since the same methods underlie research on which both parties rely. Real-juror data such as the Thomas (2020) survey are also limited, measuring beliefs but not deliberation or verdicts, and should not be overstated in either direction.

Special Issues for Defense

The research supports reliance on mock-jury findings to show that factors such as pretrial publicity, inadmissible evidence, and identity cues can influence judgments, and Ballew shows that courts will consider such research in assessing trial procedures. The meta-analytic evidence that publicity effects were larger in more realistic studies suggests that simplified studies may underestimate as well as overestimate the effect for that question. Defense counsel are likely to encounter this research in motions concerning venue, voir dire, and limiting instructions.

The research also counsels caution. Mock studies cannot show that a particular juror was biased or that a particular verdict was affected, and Lockhart noted limits of studies that did not involve actual sworn jurors but assumed the studies were valid and held that the Constitution did not require a different result. When studies are offered in support of educational testimony about juror beliefs, either side may examine whether they used students, vignettes, or no deliberation, and whether real-juror data point in the same direction. Rule 606(b) and its Arizona and military counterparts sharply limit post-verdict inquiry, so concerns about juror bias are usually best addressed through voir dire before trial.

This paper is for educational purposes only and is not legal advice. The author is a psychologist, not an attorney; summaries of law are general and must be verified against current authority or with a licensed attorney.

References

Bornstein, B. H. (1999). The ecological validity of jury simulations: Is the jury still out? Law and Human Behavior, 23(1), 75–91. https://doi.org/10.1023/A:1022326807441

Bornstein, B. H., Golding, J. M., Neuschatz, J., Kimbrough, C., Reed, K., Magyarics, C., & Luecht, K. (2017). Mock juror sampling issues in jury simulation research: A meta-analysis. Law and Human Behavior, 41(1), 13–28. https://doi.org/10.1037/lhb0000223

Bouillaut, K., Hinton, W., Fitton, L., & Beazley, P. (2026). Rape myths and juror decision-making in sexual assault cases: An updated systematic review and meta-analysis. International Journal of Law, Crime and Justice, 86, 100887. https://doi.org/10.1016/j.ijlcj.2026.100887

Chalmers, J., Leverick, F., & Munro, V. E. (2021). Why the jury is, and should still be, out on rape deliberation. Criminal Law Review, 2021(9), 753–771.

Daly, E., Smith, O., Bows, H., Brown, J., Chalmers, J., Cowan, S., Horvath, M., Leverick, F., Lovett, J., Munro, V., & Willmott, D. (2023). Myths about myths? A commentary on Thomas (2020) and the question of jury rape myth acceptance. Journal of Gender-Based Violence, 7(1), 189–200. https://doi.org/10.1332/239868021X16371459419254

Devine, D. J., Buddenbaum, J., Houp, S., Studebaker, N., & Stolle, D. P. (2009). Strength of evidence, extraevidentiary influence, and the liberation hypothesis: Data from the field. Law and Human Behavior, 33(2), 136–148. https://doi.org/10.1007/s10979-008-9144-x

Diamond, S. S. (1997). Illuminations and shadows from jury simulations. Law and Human Behavior, 21(5), 561–571. https://doi.org/10.1023/A:1024831908377

Diamond, S. S., Rose, M. R., & Murphy, B. (2006). Revisiting the unanimity requirement: The behavior of the non-unanimous civil jury. Northwestern University Law Review, 100(1), 201–230.

Diamond, S. S., Rose, M. R., Murphy, B., & Smith, S. (2006). Juror questions during trial: A window into juror thinking. Vanderbilt Law Review, 59(6), 1925–1972.

Diamond, S. S., Vidmar, N., Rose, M. R., Ellis, L., & Murphy, B. (2003a). Inside the jury room: Evaluating juror discussions during trial. Judicature, 87(2), 54–58.

Diamond, S. S., Vidmar, N., Rose, M. R., Ellis, L., & Murphy, B. (2003b). Juror discussions during civil trials: Studying an Arizona innovation. Arizona Law Review, 45(1), 1–81.

Eisenberg, T., Hannaford-Agor, P. L., Hans, V. P., Waters, N. L., Munsterman, G. T., Schwab, S. J., & Wells, M. T. (2005). Judge-jury agreement in criminal cases: A partial replication of Kalven and Zeisel’s The American Jury. Journal of Empirical Legal Studies, 2(1), 171–207. https://doi.org/10.1111/j.1740-1461.2005.00035.x

Flynn Hudspith, L., Wager, N., Willmott, D., & Gallagher, B. (2024). The impact of rape myth education on jury decision-making: A systematic review. Trauma, Violence, & Abuse, 25(5), 4062–4077. https://doi.org/10.1177/15248380241270082

Huang, L. (2025). Xenophobia in juror decision-making. Law & Social Inquiry, 50(1), 170–194. https://doi.org/10.1017/lsi.2024.36

Irvine, K., Hoffman, D. A., & Wilkinson-Ryan, T. (2018). Law and psychology grows up, goes online, and replicates. Journal of Empirical Legal Studies, 15(2), 320–355. https://doi.org/10.1111/jels.12180

Leverick, F., Chalmers, J., Keane, E., & Kinghan, J. (2025). Rape myths and jury decision making: An evidence review. University of Glasgow.

Lieberman, J. D., Krauss, D. A., Heen, M., & Sakiyama, M. (2016). The good, the bad, and the ugly: Professional perceptions of jury decision-making research practices. Behavioral Sciences & the Law, 34(4), 495–514. https://doi.org/10.1002/bsl.2246

Maeder, E. M., Yamamoto, S., & Ewanation, L. (2024). Quality-checking the new normal: Trial modality in online jury decision-making research. Journal of Experimental Criminology, 20(4), 1079–1098. https://doi.org/10.1007/s11292-023-09570-0

Reed, K., Hans, V. P., Rotenstein, V. N., McKendall, P., Rodriguez, A., Helm, R. K., & Reyna, V. F. (2026). Virtually the same? A comparison between in-person and virtual mock jury deliberations. Law and Human Behavior, 50(3), 225–239. https://doi.org/10.1037/lhb0000643

Salerno, J. M., Kulak, K., Smalarz, L., Eerdmans, R. E., Lawrence, M. L., & Dao, T. (2023). The role of social desirability and establishing nonracist credentials on mock juror decisions about Black defendants. Law and Human Behavior, 47(1), 100–118. https://doi.org/10.1037/lhb0000496

Steblay, N., Hosch, H. M., Culhane, S. E., & McWethy, A. (2006). The impact on juror verdicts of judicial instruction to disregard inadmissible evidence: A meta-analysis. Law and Human Behavior, 30(4), 469–492. https://doi.org/10.1007/s10979-006-9039-7

Steblay, N. M., Besirevic, J., Fulero, S. M., & Jimenez-Lorente, B. (1999). The effects of pretrial publicity on juror verdicts: A meta-analytic review. Law and Human Behavior, 23(2), 219–235. https://doi.org/10.1023/A:1022325019080

St George, S., & Osei Arhin, E. (2026). Comparing the convenience, data quality, generalizability, and outcomes of student- and MTurk-generated data in an experimental vignette rape perception study. Behavioral Sciences & the Law, 44(2), 249–264. https://doi.org/10.1002/bsl.70039

Thomas, C. (2010). Are juries fair? (Ministry of Justice Research Series 1/10). Ministry of Justice.

Thomas, C. (2020). The 21st century jury: Contempt, bias and the impact of jury service. Criminal Law Review, 2020(11), 987–1011.

Tinsley, Y., Baylis, C., & Young, W. (2021). “I think she’s learnt her lesson”: Juror use of cultural misconceptions in sexual violence trials. Victoria University of Wellington Law Review, 52(2), 463–486. https://doi.org/10.26686/vuwlr.v52i2.7128

Waiton, S. (2024). What do we know about ‘rape myth’ research and the claim that there is ‘overwhelming evidence’ that juries are prejudiced in rape trials? International Journal of Evidence & Proof, 28(2), 154–177. https://doi.org/10.1177/13657127231217510

Walters, J., Nguyen, L., Liu, Y., Monreal Ijurco, S., Evans, S., Chacos, N., Duran, M., & Smith, C. (2025). Justice without bias: A systematic review and meta-analysis of interventions aimed at reducing jury bias in rape and sexual assault trials. Journal of Criminal Justice, 100, 102491. https://doi.org/10.1016/j.jcrimjus.2025.102491

Wiener, R. L., Krauss, D. A., & Lieberman, J. D. (2011). Mock jury research: Where do we go from here? Behavioral Sciences & the Law, 29(3), 467–479. https://doi.org/10.1002/bsl.989

Willmott, D., Boduszek, D., & Booth, N. (2026). Profiling jury bias in rape trials: Examining the link between juror characteristics and trial decisions. In E. Dowds, S. Weare, & S. Leahy (Eds.), Researching rape trials: Methods, ethics and empirical perspectives (pp. 127–144). Routledge. https://doi.org/10.4324/9781003613015-10

Willmott, D., & Woodhams, R. (2026). Stranger rape or impromptu consensual sex? Investigating mock juror decision-making in a genuine contested rape trial. Behavioral Sciences & the Law, 44(2), 169–180. https://doi.org/10.1002/bsl.70032

Legal Authorities

Ballew v. Georgia, 435 U.S. 223 (1978).

Lockhart v. McCree, 476 U.S. 162 (1986).

Peña-Rodriguez v. Colorado, 580 U.S. 206 (2017).

Tanner v. United States, 483 U.S. 107 (1987).

Warger v. Shauers, 574 U.S. 40 (2014).

U.S. Const. amend. VI.

18 U.S.C. § 1508.

Fed. R. Evid. 606(b).

Fed. R. Evid. 702.

Mil. R. Evid. 606(b).

Mil. R. Evid. 702.

Ariz. R. Crim. P. 24.1(d).

Ariz. R. Civ. P. 39(f) (pre-2017 numbering).

Ariz. R. Evid. 606(b).

Related topic papers: Introduction to Juror Decision-Making · Pretrial Publicity, Social Media, and Juror Misconduct · Limiting Instructions, Inadmissible Evidence, and Instructions to Disregard · Juror Beliefs in Sexual Assault and Child Abuse Trials · Deliberation, Unanimity, and Court-Martial Verdicts

Related topics from Arizona Forensic Services: Counterintuitive Testimony · Courts-Martial