As jury consultants, one of the questions we hear most often is, “What kinds of jurors do I want on my jury?” Related to that, we’re frequently asked, “Do I want men or women on my jury?” “Do you think older jurors will be better for me than younger jurors?”

Jury Selection Is Actually De-Selection: Find Your Riskiest Jurors

Because it is so important to identify the riskiest jurors for cause strikes and peremptory challenges, we think of the process as jury de-selection. It is very tempting to want to identify jurors you like for your case – it makes you feel more confident that you’ll have someone on your side in deliberations. It can be pretty difficult to put that urge aside and instead talk with jurors about negative things that put your case at risk. It isn’t always very comfortable. Still, the most important thing to remember is that if you identify jurors favorable to your case, those same jurors are bad for your opponent and you’ve done opposing counsel a big favor by identifying and prioritizing their cause and peremptory strikes.

The better question to ask is: “Which jurors pose the greatest danger to my case?”

Knowing Your Verdict Goal Is a Good Start

In every case, the goals are different. For some cases, your goals are attaining a full defense verdict and zero dollars of damages. In others, liability has been decided and the defense case is focused on reducing damages. Knowing the goal of voir dire is key because relevant questions and risk factors can vary greatly between damages cases and liability cases.

Demographics Are Not Sole Predictors of Juror Verdicts

Unless you are trying a discrimination case or a case directly related to a specific set of demographics, such characteristics in general predict only a small proportion of a juror’s verdict leaning. Every woman is different and has different experiences. Every African American is different and has different experiences. And so on – for every single demographic out there. If that’s the case, what matters? A lot of other important things, including:

  • Attitudes related to your case.
  • Attitudes toward jury damage awards.
  • Attitudes toward government regulation.
  • Personal experiences and those of close others, specifically ones related to your case.
  • Feelings of victimization or fatalistic views (i.e., feeling no personal control over one’s outcome and at the mercy of the hands of fate).
  • Type of job – does the juror work in a helping, service, science or management position? What jurors focus on every day for work can offer insights to their interests, attitudes and experiences.

A Mock Trial or Other Jury Research Can Create a Statistical Juror Profile, Predictive Factors to Identify Riskiest Jurors

Although there are a whole host of great reasons to conduct jury research – honing one’s case story and themes being a big one – the data collection that takes place during such research can lead to a statistical profile of the riskiest jurors for a specific case. This statistical profile can inform the development of a Supplemental Juror Questionnaire (SJQ), voir dire questions and a strike/cause profile for use in jury de-selection.

In some cases, this quantitative data collection can be quite powerful, offering a whole host of risk factors to target in voir dire and ultimately jury de-selection. Those distinguishing risk factors vary case to case, but with enough data[1] for statistical analysis, can have reliability and predictability in a particular case. In serial cases on the same issue with combined datasets over time, the risk factors and validity of the data solidify and offer long-term guidance on which jurors are dangerous for your client’s case.

While a qualitative reviews of deliberations or other juror data can be conducted to determine a jury profile and risk analysis, one can strengthen that analysis by identifying the strongest pro-plaintiff (and often pro-damages) jurors and the most ardent pro-defense jurors, pitting them against each other in statistical analyses that narrow the most important factors via the data. Consultants can then take this valuable information and look at it in its entirety, offering true insight into the constellation of key attitudes and experiences that will identify our greatest risks in jury de-selection.

At times, with a large enough dataset,1 we are able to run deeper and more complex statistical analyses (i.e., regression models) to not only winnow down the risk factors to just a few, but also use those statistics to develop a Risk Score that can be applied to each juror during voir dire based on their responses to direct questioning.


Jury de-selection is more than just demographics; it is a constellation of demographics, attitudes and juror experiences to develop the most robust juror profile for deselecting the riskiest jurors for your case.

With over 11 years of trial consulting experience, Jill Leibold has applied her expertise in juror decision-making to hundreds of cases across all genres of litigation. Clients rely on her skills in preparing challenging witnesses for deposition and trial, and on her extensive experience in jury selection for both civil and criminal cases. She specializes in developing statistically based, juror risk profiles to identify jurors for cause and peremptory strikes, and also applies the qualitative analyses to develop case stories and themes. Jill frequently presents at national legal conferences and writes for legal trade publications about juror attitudes, implicit bias and jury selection in the areas of environmental and toxic torts, personal injury, asbestos, insurance bad faith, patent and trademark, product liability, fraud and criminal cases in her work at Litigation Insights.

[1] To run a regression analysis, the dataset should have 15 respondents per variable entered. So, for a regression with only four predictors, the sample size should be a minimum of 60. The smaller the sample size, the more “noise” the data will bring with it, so the ultimate predictiveness will be lower with the bare minimum sample size as well. More typically, these analyses are conducted on a larger combined sample of multiple mock trials with at least three deliberating juries per mock trial. However, as with anything relating to human behavior, nothing can be calculated to guarantee a degree of statistical certainty.