Part I: Wording of Questions about Prospective Jurors’ Biases

An abundance of advice suggests how to best phrase questions to prospective jurors in jury surveys, in general or individual voir dire, and in change of venue surveys. Attorneys and judges should want honest answers to questions about personal background, familiarity with a case, relevant experiences, and most of all, potential bias. Legal consultants and writers, based on informed intuitions, on an understanding of the degree to which even small language changes can affect answers, and on their own courtroom experience, have suggested innovative ways to pose bias questions. Several legal experts have proposed that, rather than or in addition to asking people their own opinion about a defendant’s guilt or a company’s liability, probes should be directed toward whether they think their friends or local community think the defendant is guilty or insurance companies are evil. Perhaps prospective jurors project their own opinions onto others, as we are generally wont to do according to decades of research on the false consensus effect (e. g., Baumann & Geher, 2002; Ross, Greene, & House, 1977). A yes answer would thus provide a glimpse into a person’s unconscious biases or biases that they are not willing to admit (e.g., Hamilton et al. 2014 and Morris’s response to the article). Further, the approach might make the prospective juror feel comfortable enough to open up about his or her own opinions. However, neither this nor many other intuitively sensible ideas have been empirically tested—they sound good, but do they really work? We don’t know because only a small fraction of research on the forces that affect people’s opinions has been done in a legal context. For example, the following research topics clearly have applicability to the law but the studies were not run in a legal context: social desirability pressure (Marlow & Crowne, 1961), the inability to recognize one’s own biases (Gilbert & Hixon, 1991; Swann & Read, 1981), a lack of understanding of how much one’s biases affect one’s opinions (Nisbett & Ross, 1980).

Past Research on Variations in the Wording of Bias Questions

One exception is a series of studies by Hamilton and colleagues.[1] Hamilton had suspected, after administering several change of venue (COV) surveys in high profile murder cases, that traditionally worded guilty bias questions underestimate bias (e. g., “In your opinion, is _______ innocent or guilty?”). What is wrong with that question? It turned out that answer patterns for some of the other survey questions (e.g., questions on community perceived opinions, the likelihood of conviction, felt community pressure to find the defendant guilty) suggested that many people who had answered “innocent” most likely believed the defendants had actually committed the crimes. In both lab experiments and several COV surveys, varying the wording of the innocence/guilt bias question confirmed Hamilton’s suspicions. In one COV survey, for example, in comparison to the traditionally worded question, the number of people indicating guilty bias rose by 35% with the “water cooler”[2] alternative version—“If you had to say you lean one way or the other right now about _________’s innocence or guilt, which way would you lean?” For that reason, Hamilton has asked both versions of the guilty bias question in all subsequent COV surveys. She includes the traditional version in case judges or prosecutors challenge the alternative version.

In our opinion, more empirical research is needed in order to aid attorneys, judges, and trial consultants in identifying the best questioning techniques for uncovering bias. Intuition and practical experience are all well and good, but we need to test our ideas scientifically as well. Part II of this article reports an experiment on the effectiveness of question wording concerning two important bias questions commonly asked in COV surveys, jury surveys and voir dire. These questions address putting aside prior information or opinions and assuming innocence until a defendant is proven guilty beyond a reasonable doubt.

Part II: The Put Aside and Assume Experiment

Background: Put Aside, Assume, and Rehabilitation

In high profile criminal cases, potential jurors are often asked whether they would be able to put asideany opinions they have formed based on information they have already heard about the case and whether they could assumethe defendant is innocent until proven guilty beyond a reasonable doubt. We wondered:

  • Does such wording strike potential jurors as a challenge to their capabilities?
  • Do people feel pressure to answer yes in order to appear to be good, thoughtful, and fair?
  • In voir dire, anyone who expresses uncertainty about their ability to put aside opinions or assume innocence can expect to face rehabilitation attempts by the judge and/or the attorneys. If a person answers no, do they perceive the further questioning (and monologues) as pressure to change their answers to yes? Furthermore, how many of them cave to the (perceived or real) pressure?
  • And finally, do prospective jurors who witness others being subjected to rehabilitation attempts try to avoid being bullied into yes answers when they are questioned, by simply answering yes when they mean no?

But wait. If the answers to the above questions are yes or maybe, so what, as long as the phenomenon is rare…

Ubiquity and Prehabilitation

Unfortunately, such questioning is ubiquitous, as documented in a previous TJE article by Hamilton and colleagues (2014). Judges and prosecutors almost always (and even defense attorneys, quite often) strongly push prospective jurors toward yes. For example, one defense attorney said in an individual voir dire interview in a murder case in which Hamilton was a consultant, “You seem like a reasonable person. Do you think you’ll be able to keep an open mind and base your opinion solely on evidence that’s presented in court?”

Even worse, the ubiquity of prehabilitation study showed that attorneys and judges rarely wait until prospective jurors have admitted bias before they start asking leading questions and prodding them toward yes answers. The message prospective jurors receive before they’re even given a chance to explore their own potential biases, is: “If you want to look good to the court, appear competent to your peers, and convince yourself that you are fair-minded, answer all I’ll-be-a-good-juror questions in the affirmative. If you want to feel foolish, look like a bad American, and face more questioning and pressure, say no.” Hamilton and colleagues dubbed this specially timed, lie-promoting, premature rehabilitation prehabilitation.

Social Desirability Pressure

Experimental research by Hamilton and various colleagues[3] shows that such questioning truly does pressure people into underestimating their own bias. It drives their bias underground by exerting social desirability pressure, or pressure to answer questions in socially acceptable ways. Not surprisingly, when people are asked whether they can fulfill their civic duty by assuming innocence or whether they have the ability to put aside news coverage information in order to be fair, they are likely to answer yes. Most people want to be seen and want to see themselves as “able/capable” in a general sense, and more specifically, they want to be as well as appear to be good citizens who understand the American justice system.

But what if people are given the opportunity to admit “legal weaknesses” without appearing foolish or un-American? For this experiment, we crafted what we hoped were new and improved versions of the Put Aside and Assume questions, embedded them in the COV survey for a recent Kentucky murder case, and randomized which version of each question the jury-eligible survey participants were asked. We predicted that the new versions would lead to more honest answers––the answers that psychological research would predict are actually true given the difficulty of ignoring one’s previous knowledge and biases. There is simply too much research evidence that bias goes underground for us to believe that it’s the other way around—that admission of bias could be covering up a lack of bias. For example, Moran & Cutler (1991) found that mock jurors who promise to be impartial and to disregard pretrial publicity are convinced that they have eliminated any bias yet still show bias in their decisions.


Dr. Hamilton was hired to design and analyze the results of a change of venue survey concerning a highly publicized 2014 Kentucky murder case. Thoroughbred Research Group, a Louisville polling company, made calls using random-digit dialing to reach landlines and cell phones. Callers completed telephone interviews of 401 jury-eligible county resident adults aged 18 and above.[4] Results had a margin of error less than five percent. Respondents familiar with the case (311 out of 401) answered one randomized version of each experimental question.

Experimental Questions

Hamilton embedded in the survey various versions of two important bias questions that prospective jurors or survey respondents typically face—the “can you put aside opinions” question (three versions) and the “can you assume Innocence” question (six versions).

Conversion to Scaled Statements

All versions of the questions were improved in the following way—they were revised as statements with which the respondent could agree or disagree to varying degrees on a Likert Scale. Rather than forcing people to choose between a categorical yes (“I’m a good citizen”) and no (I’m a bad citizen”), Likert Scale items allow more nuanced answers by offering a gradation of responses, for example, the gradations we used: “strongly disagree,” “disagree,” “neither agree nor disagree,” “agree,” “strongly agree.” This approach presumably lessens social desirability pressure by allowing people to admit some degree of potential bias as opposed to forcing categorical admission or denial of bias.

Ability vs. Difficulty

To determine whether and to what extent variations in the wording of the statements might influence answers, the callers asked each respondent a randomly assigned version of each experimental question. One version of each item was worded traditionally regarding abilities (“you would be able to,” “you could”…). For the other versions, however, respondents rated statements that prompted them instead to think about how easy or difficult the tasks might be (see exact wordings in Tables 1 and 2). Our thinking was that ability questions may be seen as challenges to one’s competence, whereas wording that normalizes the potential difficulty of a task would not be interpreted that way.


If you were to serve on the jury and had to decide

whether __________ was guilty, you…

…would be able to
put aside
…could easily
put aside
…might have some trouble
putting aside

…opinions you currently have about this crime.

“Innocent Until Proven Guilty Beyond a Reasonable Doubt” (IUPGBRD) vs. Not Guilty

The Assume question varied in a second way. Each version of the ability wording (could, could easily, and trouble/difficulty) was presented in two forms—one containing the traditional “innocent until proven guilty beyond a reasonable doubt” phrase (IUPGBRD) and one with wording we thought might preclude a knee-jerk yes response—yes is the legally required and socially desirable answer in response to a phrase that has been imprinted on our brains from a lifetime of hearing it in civics class, on TV, in the movies, and reading it in crime novels. Half of each version of the Assume statement substituted the customary IUPGBRD phrase with the less problematic “not guilty.”


If you were to serve on the jury and had to decide whether [the defendant] was guilty, you…

…could assume


…could easily assume


…would have some difficulty assuming


…could assume

he is

…could assume

he is

might have some trouble assuming

he is


  1. Put Aside: We predicted that a higher percentage of people would admit to potential bias in response to the Put Aside statement if we changed the wording to “you could easily put aside opinions you currently have…” It seemed to us it would be easier for most people to admit they could not do something easily than to imply that they could not do it at all.
  2. Put Aside: We predicted, for similar reasons, that switching from the traditional wording to “might have some trouble putting aside” would lead to more admissions of potential bias. Again, saying that one might have some trouble doing something should be more comfortable than admitting that one cannot do it at all.
  3. Assume: We hypothesized, just as for Put Aside, that the word “easily” would increase admissions of potential bias in both the Not Guilty and IUPGBRD conditions.
  4. Assume: We also hypothesized again that the Trouble/Difficulty wordings would increase admissions of potential bias.
  5. Assume: Finally, we predicted that the Not Guilty versions of the assume statement would lead to more admissions of potential bias than would the traditional version.

Results and Discussion

Put Aside

As apparent in Table 3, Version 3 of the Assume statement (“you might have some trouble … IUPGBRD”) performed as expected, leading to much more bias acknowledgement than did the traditional version (“you would be able to”). In fact, people who heard the trouble version were two-and-a-half times as likely (52% vs. 21%) to express doubt about putting aside pretrial opinions, a dramatic and statistically significant difference.[5]

Table 3


Note: The percentages represent the proportion of people who acknowledged in some way that putting aside opinions might present challenges, regardless of the direction of question wording—that is, whether agreement indicated a denial (V1, V2) or admission (V3) of bias.

Furthermore, the difference between the two percentages (52% – 21% = 31%),suggeststhat when the put aside question is posed in the traditional way in the courtroom, it is likely that about one-third (31%) of prospective jurors will misrepresent themselves, not admitting they would struggle to put aside prior opinions when in fact they would. That is a frightening, though not surprising, prospect.

Contrary to our prediction, adding the word “easily” to the traditional wording did not raise the likelihood of people’s admitting bias—the easily version of Put Aside also led to a low rate of admission—16%.


For the Assume question, both changing the focus to difficulty or trouble and using not guilty elicited dramatically more honest and realistic appraisals of potential bias. In response to the traditionally worded question (Version 1), the admission rate was only 4%. Four of the other five statement versions (Version 3 through Version 6) had a powerful effect on the rates of admitting potential bias, raising them to an average of 27%[6](range, 22%-30%). Using one of the four alternative versions, in other words, bought us an average improvement over the traditional version of 23% (27% – 4% = 23%), which means that nearly one-fourth of the respondents who hear a traditionally worded version of the Assume question in court may well be hiding any problems they might have with assuming innocence. Clearly the traditional wording triggers a legalistic, automatic, I’m-a-good-and-knowledgeable-American-citizen response.

Interestingly, for the Assume statement, unlike for Put Aside, just adding the word “easily” to the IUPGBRD statement (Version 2) nearly tripled the percentage of people who acknowledged their imperfection as potential jurors, raising it from 4% to 11%. On the other hand, while adding “easily” to “not guilty” (Version 4 vs. Version 5) appears to have increased bias admissions (from 22% to 27%), that increase was not statistically significant.

Table 4


Part III: Suggestions

The results of our experimental COV survey suggest that making it more comfortable for people to express doubts about their ability to ignore pretrial publicity and to presume innocence increases the likelihood that the justice system will get honest answers from potential jurors. This should in turn increase the likelihood of seating a fair jury. In particular: (1) It is helpful to switch from ability challenges to wordings that allow people to rate the difficulty of a task. (2) Adding the word “easily” to traditional wording may or may not help—our results were inconclusive on that variable. In one case, adding easily helped significantly; in another, it appeared to help according to the “Hamilton Eyeball Test” but not according to significance testing; and in the third case it definitely did not increase bias admissions. (3) Replacing the legalistic, overly familiar “innocent until proven guilty beyond a reasonable doubt” phrase with the simpler and less evocative expression “not guilty” drastically increases admissions of bias.

Below are some specific suggestions about how to use our results and how related suggestions from others fit in.

Likert Scales in Real Life

For all of the questions, offering a scale of disagree strongly to agree strongly allows for subtleties, not forcing someone to flat-out confess to being a legally incompetent poor excuse for an American. Gradations in answer options may reduce the potential shame involved in expressing doubt about one’s likelihood of being fair. So how can we use Likert scales or other rating scales in our consulting practices?

Scaled answers are always an option in a COV survey given that the questions are in the consultants’ and attorneys’ control. Some judges may not allow jury surveys to include scaled answers, but if your judge does, use them. Finally, although orally delivered scale-based questions might be difficult for listeners to understand and a bit awkward and unwieldy in voir dire, it is possible to utilize them in that situation if the judge will allow it. To make orally delivered scaled-answer questions clear, careful construction with an eye to simplicity and clarity is necessary.

More than just the Put Aside and Assume questions can be asked on a scale. One could ask the “lean guilty” question in jury selection as “If you had to say you lean one way or another right now about the guilt or innocence of ___________, would you say you lean heavily toward guilt, somewhat toward guilt, somewhat toward innocence, or heavily toward innocence?” To avoid being accused of leading prospective jurors toward a guilt answer, it is important to randomize the option order, that is, start with the guilt options half the time and use the innocence options first for the other half.

Another consideration: answer options to the lean question and to the questions in the experiment reported here reflect forced choice scales—they had an even number of options. There was no “neither innocent nor guilty” option in the middle, no 3 on a 1-5 scale. The forced choice approach makes people pick a side, which can be helpful. Although they can still volunteer a “neither” response, there’s no sense making it easier for them to cop out.

The Put Aside Question in Real Life

For the Put Aside question, the vital ingredient for our outcome was altering the traditional focus on the ability to do what is required of a juror. In Version 3, we switched to an emphasis on potential difficulty with putting aside prior opinions. Admissions of potential bias more than doubled.

How can we help our clients, knowing that result? In all three situations, COV surveys, jury surveys, and voir dire, we should avoid words like “able,”and phrases like “can you?.”Instead, we should ask whether someone might have some (or a little or any…) difficulty/trouble (or might struggle a bit…) putting aside opinions (or disregarding news coverage or not succumbing to community pressure…). Then, in voir dire in particular, if the person has admitted potential bias, the attorney should probe them about what biases they have and why rather than going straight into (useless) rehabilitation. That just leads prospective jurors by the nose until they give up and say, “Yeah, okay, I guess I can do that.” Shockingly, even defense attorneys make this mistake. In six of the eleven cases Hamilton and colleagues analyzed across the country and in Canada, even defense attorneys engaged not only in rehabilitative questioning but also in prehabilitative questioning (Hamilton et al., 2014, TJE).

The Assume Question in Real Life

Advice for asking this question is similar to the advice above, with one important addition. Avoid focusing on abilities, but also, avoid triggering people’s “I’ve watched a million court shows and I know I’m supposed to assume innocent until proven guilty beyond a reasonable doubt” responses. Instead, ask, “Do you imagine it might be difficult for you to go into trial with the belief that the defendant is not guilty?” or “How easy or hard do you feel it would be to assume the defendant is not guilty, on a 1-10 scale?” (à la Morris’s consultant response to Hamilton et al., 2014).

Part IV: Conclusions

Our results showed that had we asked only the traditional, leading, prehabilitatively worded Assume and Put Aside questions in the Kentucky COV survey, we would have severely underestimated the degree of anti-defendant bias in the community. We would have concluded that only about one in five doubted their ability to put aside opinions when in fact over half of them did. Results for the Assume Innocence question were even more drastic—had we used only the traditionally worded Assume statement we would have believed that only one in twenty-five people (4%) doubted their ability to assume innocence, whereas in actuality it was one in every three to four people!

We believe that the alternative wordings we used, and likely, similar rewordings, have the potential to promote far more careful, honest answers concerning bias. Many other writers have made suggestions about the wording of bias questions, some of which concern further probing after it becomes apparent that a prospective juror harbors some doubts. Among these writings are the Morris and Wiley responses to the Hamilton et al. 2014 TJE article, Broda-Bahm’s blog about the article, and Broda-Bahm (2011).[7] We strongly recommend that such alternative questioning techniques be used by jury consultants, judges, and attorneys in their attempt to seat unbiased juries.

In sum, when you are seeking realistic answers, how you ask the question makes all the difference.


Mykol C. Hamilton, PhD [] is the H.W. Stodghill, Jr. and Adele H. Stodghill Professor of Psychology at Centre College in Danville, Kentucky. Her research interests are in the social psychology of jury selection, change of venue issues, and the effects of linguistic choices in voir dire. She does change of venue and other jury-related consultation. You may find more information about Dr. Hamilton at her college webpage, and at her consulting webpage [].

Kate Zephyrhawke has a degree in English Literature from UC Berkeley and master’s degrees in psychology, mass communications (abt), and rhetoric and composition. She teaches English at Hillsborough Community College in Tampa, Florida, and does psychology of law research with Mykol Hamilton Hamilton Zephyrhawke Trial Consulting.


Bauman, K. P., & Geher, G. (2002). We think you agree: The detrimental impact of the false consensus effect on behavior. Current Psychology, 21(4), 293-318.

Broda-Bahm, K. (2011, January 24). In jury selection, pay all kinds of attention to the man behind the curtain. Persuasive Litigator,

Cox, M., & Tanford, S. (1989). Effects of evidence and instructions in civil trials: An experimental investigation of rules of admissibility, Social Behavior, 4(1), 31-55.

Gilbert, D. T., & Hixon, J. G. (1991). The trouble of thinking: Activation and application of stereotypic beliefs. Journal of Personality and Social Psychology, 50, 509-517.

Marlow, D., & Crowne, D. P. (1961). Social desirability and response to perceived situational demands. Journal of Consulting Psychology, 25(2), 109-115.

Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, NJ: Prentice Hall.

Hamilton, M. C., Augustus, A. N., & Melloan, J. (2011, March). Voir dire, change of venue, and “legal desirability”: Hidden jury bias in high PTP cases. Poster presented at the American Psychology-Law Society Conference, Miami, FL.

Hamilton, M. C., & Henize, H. (2013a, March). Prehabilitation, individual voir dire, and detecting prospective juror bias: The Trayvon Martin case. Poster presented at the American Psychology-Law Society Conference, Portland, OR.

Hamilton, M. C., & Phipps, C. (2013b, May). Voir Dire and Social Desirability: Does “Prehabilitation” Reveal or Conceal Juror Bias? Poster presented at the American Psychological Science Conference, Washington, D.C.

Hamilton, M.C., Linden, E., Pitt, M., & Robbins, E. (2014, August). The ubiquity of premature rehabilitation: How “prehabilitation” leads prospective jurors to hide their biases, The Jury Expert, 26(3), 48-65.

Hamilton, M. C., & McGinley, K. (2009, March). Lies, damn lies, and voir dire: Prospective jurors’ opinions of guilt/innocence in highly publicized cases. Presented at the American Psychology-Law Society Conference, San Antonio, TX.

Moran, G., & Cutler, B. L. (1991). The prejudicial impact of pretrial publicity.

Journal of Applied Social Psychology, 21, 345-367.

Pronin, E., Lin, D. Y., and Ross, L. (2002). The bias blind spot: Perceptions of bias

in self versus others. Personality and Social Psychology Bulletin, 28(3), 369-381.

Ross, L., Greene, D., & House, P. (1977). The “false consensus effect”: An egocentric bias in social perception and attribution processes. Journal of Experimental Psychology, 13, 279-301.

Siler, H., & Hamilton, M. C. (2010, April). The role of context and wording in prospective juror questioning. Presented at the conference of the Southwestern Psychological Association, Dallas, TX.

Swann, W. B., & Read, S. J. (1981) Acquiring self-knowledge: The search for feedback that fits. Journal of Personality and Social Psychology, 41, 1119-1128.

[1] For example, Hamilton & Henize (2013a); Hamilton & McGinley (2009).

[2] The alternative version was an attempt to approximate the following relaxed, low social desirability pressure approach, but with slightly more formal wording appropriate to a survey. “If you were standing around the water cooler at work what would you say?”

[3] For example, Hamilton & Henize (2013a); Hamilton, Linden, Pitt & Robbins (2014); Siler & Hamilton (2010).

[4] The sample population was representative of the jury-eligible community by gender, race, age, income, and education.

[5] X2 (1) = 46.07, p < .0001

[6] X2 (1) = 38.74, p < .0001

[7] The blog response is accessible under Comments at the end of the TJE article or at

Charli Morris has been a trial consultant for 23 years and works nationwide on civil and criminal cases. She is co-author of The Persuasive Edge, a regular reader of The Jury Expert, and she can be reached directly by e-mail to

#WordsMatter: Helping Jurors Feel Better About Acknowledging Bias

Authors Hamilton and Zephyrhawke know that I am standing in their corner when it comes to conducting empirical research that will support the assertions trial consulting social scientists have made for years about the trouble with how many judges and attorneys craft voir dire questions. More importantly, the findings in their “Put Aside and Assume” Experiment are highly generalizable to all cases – and all forms of bias – so they represent a step forward for us all.

Previously I’ve written a response to the aptly-named construct Hamilton calls “prehabilitation,” in which I suggested additional follow-up questions that flow from the items she tested.[1] I agree with how we can measure the ease or difficulty of doing a thing to provoke more candid responses, as well as strategies like asking for bias projection and using scaled-response questions. And I agree wholeheartedly with Diane Wiley’s suggestion (in her commentary to the same article) that “HOW” follow up questions are perhaps more important than any question about “if” or “whether” (e.g., How will you overcome the opinion you say you’ve held closely for many years?).

What we all seem to agree on is that words matter. I’m even tempted to put a hashtag in front of that to demonstrate how universal the idea really is: I see it working as a tool for repairing relationships, an anti-bullying campaign, a session conducted by the Human Resources department, or a lesson in contract negotiations. It’s just so true.

But another of the words that I come across in Hamilton, et al.’s current article is “assume” so I want more research to explore that too. The authors pit two phrases against each other on the topic of guilt or innocence: “Innocent Until Proven Guilty Beyond a Reasonable Doubt”[2] (legalistic and formal and likely to provoke responses that mask bias) versus “Not Guilty” (which I think the authors suggest may be less likely to mask bias although I’m not fully convinced)[3]. They leave out an important verb that typically precedes the longer phrase which is “Presume,” as in Presumption of Innocence. But they add the word “Assume” to their scaled response item. And that stuck with me even after I was persuaded by the results.

I’d like research to tell us whether there is a qualitative or connotative difference for most jurors between the words presume and assume. I found a great comparison of the two words by way of a Google™ search on It explains that presume means you suppose or believe something without any proof, but you do so based on probability. For example: if someone is banging on your door late at night you might assume it is your crazy neighbor; but if this is something she does regularly you aren’t just guessing when you presume it is the crazy neighbor because she has a habit for banging on your door late at night.

Is there a difference between presume and assume that matters to jurors? Maybe not, but we have all been taught to break down the word ass-u-me into three smaller parts to illustrate a point about how your bias might just come back to bite you in your assumption (or someone else’s).

For that reason, I wonder if Hamilton and others will be interested in refining the language of the questions further in future research? After all, being asked to assume something – in and of itself – is presumably a socially undesirable thing to do. What if the ease or difficulty the prospective jurors report is not with the question of guilt or innocence (as the question intends), but instead with the proposition that he or she is being asked to assume something without having heard a shred of evidence? For me this doesn’t necessarily diminish the importance of their research, but it does raise another question about how word choice can and does influence responses.

We clearly need different and better ways to measure pre-existing bias. Hamilton and her co-authors have presented us with empirical proof that the way we word those questions can produce meaningful differences. Most importantly, better word choices can help us identify bias that is harmful to the fundamental tenant of impartiality in civil and criminal jury trials.

I hope more lawyers will use new questions to elicit and identify bias, and I hope more judges will be informed by this research. Let’s get the word out. All the words. Because they matter.


[2] Note taken that Diane Wiley is also correct to assert that the fair and proper “U” word should be Unless, not Until, but sadly this is not a substitution our courts regularly adopt. Defense lawyers must and should.

[3] “Not Guilty” has the potential to be just as loaded a phrase, in my view, as “IUPGBRD.” I imagine that for some, at least, Not Guilty is just as absolute – conceptually – as is Guilty and therefore may be subject to some form of hidden bias. I defer that debate for another article or commentary.

Christina Marinakis, J.D., Psy.D. is a Senior Consultant with Litigation Insights. In addition to conducting pre-trial jury research, she specializes in drafting voir dire and juror questionnaires, frequently assisting counsel in court during jury selection. She has worked on a variety of civil and criminal matters in jurisdictions throughout the United States.

Bias is in the Eye of the Beholder

Dr. Hamilton and Ms. Zephyrhawke address the importance of a considering the nuance of language and how it can influence jurors’ responses to survey, questionnaire, and voir dire questions, and they offer sound advice for practitioners who are seeking to elicit admissions of juror bias. While they raise interesting propositions and recommendations regarding the importance of language choices, in offering their “do’s and don’ts” of jury questioning, they unfortunately do not take into consideration some of the realities of our adversarial system. Yes, the court’s goal should be to sit a fair and impartial jury, but in reality, that is achieved by competing parties who are seeking a jury biased in their favor – or at least forcing their opponent to use a strike on a juror that is biased in their favor. Thus, in some instances, attorneys will aim to suppress bias (in favor of their case or client) as much as they aim to elicit bias (against their case or client). Therefore, to make blanket statements about avoiding prehabilitation and encouraging jurors to admit bias warrants a caveat that it will depend on the attorney’s goals and the subject at issue.

Prehabiliation, for instance, is particularly useful and important for prosecutors and plaintiffs, given that the defense will always have the last word and plaintiffs’ counsel will not have the opportunity to rehabilitate a juror who expresses favor to their side. In their article, Dr. Hamilton and Ms. Zephyrhawke frequently cite the example of pre-trial publicity bias against a criminal defendant. While their recommendations are well-suited for the defense attorney in that matter, for the prosecutor working on that case, the goal would be to suppress or minimize that bias. Now, suppose a juror on the panel has a family member who was wrongfully accused of a crime. In that instance, would the criminal defense attorney want to suggest that the juror “might have difficulty putting that aside?” Of course not. Rather, in this scenario, the defense attorney should revert to the “ability to set it aside” framing. In every case – criminal and civil – there will be some topics where counsel wants to elicit bias responses, and other topics where they want to minimize it. Thus, while the article offers important considerations for trial counsel, it is just as important to qualify that advice by noting that the language used by counsel should vary according to the goals of the party posing the question.

I also have some reservations about the authors’ characterization of the jurors’ admission to bias as an “honest” response. While there is certainly plenty of empirical evidence showing that people often cannot recognize or admit their own biases (i.e., implicit bias), I think it is a leap to describe an admission of bias as “honest” and a denial of bias as “dishonest.” Indeed, there will be many issues for which a juror truly does not harbor strong feelings or is able to logically separate their own experiences from those of the parties in the case. For instance, not everyone who reads about a criminal defendant in the newspaper will believe what they have read. Likewise, I have run across many good defense jurors in toxic tort cases who have personally had cancer, but are able to separate their own experience from the plaintiff’s cancer – perhaps because it is a different kind of cancer, or cancer runs in their family. In both of these examples, the manner in which the jurors respond is far more telling in whether they have a bias than the actual response they give. If a juror hems and haws, hesitates, frowns, furrows their brow, or is fighting back tears, I am much more inclined to believe they are harboring bias that cannot be set aside, as opposed to a juror who quickly, unemotionally, and affirmatively denies bias. Now certainly, there will be some jurors who are oblivious to their own bias, and that bias may influence them at some point during the trial, but to describe only those who admit being biased as “honest,” and to suggest that those who deny bias are just trying to avoid being seen as a “bad American” is troublesome. Rather, the authors would garner greater credibility if their article was void of these judgmental phrases and focused instead on “admission of bias” or “denial of bias.” Whether the juror’s response is honest or dishonest is not for us to know or judge.

That said, I often encourage counsel to make adjustments in the phrasing of their questions to normalize bias and increase the likelihood of eliciting an admission of bias from certain undesirable jurors, so it is encouraging to now be able to point to some empirical evidence on the position. In addition to avoiding talk about a juror’s “ability or inability,” which the authors suggest, I have also noticed that jurors seem reluctant to admit they “have a problem.” Too often, I hear attorneys ask questions such as, “Would anyone have a problem awarding substantial damages?” Yet, how often do we like to admit we have a problem? It seems to me – and the study’s findings may support this – that language such as, “difficulty with,” “struggle with,” and “concerns about” is more likely to elicit an admission of bias than “have a problem with” and “unable to.” Another method of normalizing bias is to use the “some people believe X, while others believe Y; which one is closer to how you feel?” format. This technique can be used to normalize bias while also previewing your case themes.

I also agree with the authors that questions about a juror’s “leaning” or about what their friends or people in their community believe may be helpful to gauge a juror’s potential bias and assist in exercising peremptory strikes. However, I disagree that a juror’s leaning necessarily indicates a bias. Rather, it evidences the lens through which the juror will listen to and filter the evidence. We all have these lenses and leanings on various issues, so labeling any juror who admits to such as “biased” would mean there could never truly be a “fair and impartial jury.”

It is worth noting that counsel will inevitably have to ask those “unable to set aside” and “unable to be fair and impartial” categorical questions in order to be successful in removing the juror for cause. While the article alludes to other publications that discuss this reality, the authors would benefit from suggesting that their techniques be used to begin the discussion with jurors about their potential biases. Then, once jurors voice their opinions and are anchored into their responses, counsel can lead them down the rabbit hole until the juror agrees with the court’s “magic words” necessary to secure a cause challenge.

Though the article focuses on high-publicity criminal trials, the potential for juror bias exists in nearly every case, and the authors’ conclusions can be applied in a variety of settings. The study’s results supports the techniques that trial consultants often suggest to maximize the value of jury questionnaires and voir dire for their clients.

Author Response to Dr. Christine Marinakis

Dr. Marinakis agrees that the court’s goal should be to seat a fair and impartial jury. Yet she believes that our best practice recommendations do not take into consideration some of the realities of the adversarial system. We would argue that we do indeed take those realities into consideration. We do so with the intention of finding ways to improve the odds of seating a truly fair jury. Dr. Marinakis believes “[impartiality] is achieved by competing parties who are seeking a jury biased in their favor.” We contend that this practice (including the usual ways of probing for bias) is acquiescing to an imperfect system that approximates fairness, yet may be, more often than not, stacking the deck against the defendant, particularly (though not exclusively) in high-publicity criminal cases where emotions run high. When moral outrage is evoked, as it always is in these cases, people become more punitive (Jonathan Haidt and others, TJE article by Liana Peter-Hagene, Alexander C. Jay and Jessica M. Salerno). It is unrealistic to expect people to ignore their emotions.

In addition to the problem of emotion coloring and often overriding reason, someone being arrested and charged with a crime causes people to lean towards an assumption of guilt, a fact that is well known in the criminal justice system. Since the law requires a presumption of innocence, and if impartial juries are always seated by competing counsel as Dr. Marinakis states, why is it that jury acquittals are rare in both state and federal trials? In 2010, nationwide, the ratio of guilty verdicts to acquittals was 143 to 1. Are the police always right? Is the defense always incompetent? Or could there be something else going on? Juries are reputed to be “unpredictable.” Ninety percent of criminal cases that go to trial result in a guilty verdict. Knowing these facts, many defense attorneys – even when they believe their clients are innocent – recommend a plea bargain; over ninety percent of criminal cases never go to trial. Moreover, juries get the verdict wrong sometimes, for a variety of reasons, and innocent people go to jail. Jurors who unwittingly misrepresent their ability to be impartial – our survey shows that to be one third of potential jurors – could at least partially explain the enormous disparity between convictions and acquittals.

Since no device, formula, or algorithm can reliably determine whether someone is or is not capable of being impartial, the best we can do is give people the opportunity to acknowledge, without embarrassment, that they might have trouble putting aside their opinions and feelings. If prosecutors are intentionally looking for jurors who will likely decide in their favor, they will instinctively avoid following our suggestions and simply continue the prehabilitation/rehabilitation status quo for voir dire. Dr. Marinakis’s caveat is unnecessary. Though certainly not our intention, our findings could, in fact, encourage some attorneys to hone their prehabilitation skills.

Dr. Marinakis also states: “it is a leap to describe an admission of bias as ‘honest’ and a denial of bias as ‘dishonest.’” Yet the word dishonest, with its connotation of intentionality (and synonyms such as fraudulent, cunning, devious, and unethical), does not appear anywhere in our paper. We refer to honest answers on juror surveys as more realistically assessing a respondent’s ability to put aside prior information and opinions. The appropriate contrasting term to describe their answers, in context, would be self-deceptive. While anyone, attorney or bystander, would correctly identify a potential juror in voir dire who “hems and haws, hesitates, frowns, furrows their brow, or is fighting back tears” as biased, the absence of such displays does not constitute impartiality, particularly when potential jurors feel pressured to give the “correct” and socially desirable answer. We make it clear in the article that people do not just want to avoid appearing to be bad citizens; they sincerely want to be fair and impartial, and thereby convince themselves they can. Dr. Marinakis suggests we strip the article of “judgmental” terms that trouble her, such as “bad American,” and focus on admission or denial of bias. But that entirely misses the point of our research: social desirability, which we characterize by those “judgmental” phrases, influences juror response. Whether a juror’s answers to questions about bias, as well as their ability to be impartial, are honest or “dishonest” (i.e, realistic or illusory) is most clearly and emphatically in our interest – not just as attorneys and trial consultants, but as Americans – to know and judge.

Dr. Marinakis is correct when she concludes that “the potential for juror bias exists in nearly every case.” While our research may not be earth shaking, it is ground breaking in that it suggests juror bias runs deeper and is more resistant to rehabilitation than has always been assumed.

If our recommendations were widely accepted and jurors who could safely admit they might have trouble being fair and impartial were eliminated, perhaps some wrongful convictions would be avoided. If juries had less of a propensity towards a guilty verdict, plea bargains for innocent people might not be the norm. If attorneys and judges are willing to adapt to new information, some flaws in the system will be corrected, and we will be one step closer to achieving the goal of trial by jury, which is not just winning, but justice.