How do humans assess large sets of evidence? We start by simplifying and internally summarizing. One of the main ways we do that is to boil the information down to a limited number categories, usually two. For example, let’s say you’re trying to assess the health effects of a glass of wine or two per day. A wealth of research results could be arranged on a spectrum of the degree of harm or benefit that the wine consumption may bring. But chances are, you won’t account for that continuum. Instead, you’ll group those studies into two categories: Those that show moderate drinking is “good for you,” and those that show it’s “bad for you.”
That shortcut is not just a time-saving habit. It is also a feature of human cognition. In a recent study (Fisher & Keil, 2018), researchers from Carnegie Mellon and Yale demonstrate that we have a “binary bias” when it comes to understanding evidence. That means we have a “tendency to impose categorical distinctions on continuous data,” and to “treat evidence as all or none without tracking the differential impact of graded evidence.” In nine studies looking at evidence assessment in several contexts, the researchers documented this bias. For example, creating distributions of data having the same average, they manipulated the distribution to make some top-heavy and others bottom-heavy, and were able to distort participants’ perceptions of the mean. Essentially, people tended to compare the data points above the mid-point to the data points below it, while ignoring most of the other information. The researchers conclude that this binary bias applies in evaluating statistics, but also in a wide variety of other social judgments relating to finance, health, public policy, and law. In this post, I’ll take a look at a few ways this bias is likely to influence a jury’s reaction to evidence.
I see three general ways that the cognitive tendency to sort things into sets of two can influence courtroom persuasion.
Binary Bias in Evaluating Data
Experts often have to present larger data sets. In a products case, for example, an expert might be sharing the failure rates of a number of different products. The jury is inevitably going to be drawn to simplify that data, and knowing that and planning for it can help. For example, since we can expect the jurors to group data like that into buckets of “high” and “low” failure, the expert might help by getting there first, and suggesting the most accurate and helpful way that data could be grouped. That might run counter to the expert’s own training, since aggregating data means throwing away some of the variance. However, not adapting to the jurors’ own tendencies to aggregate means throwing away a chance to influence the ways they group and simplify that data.
Binary Bias in Evaluating Witnesses
To attorneys and those who work with them, there is a great deal of nuance to a witness’s performance on the stand. There is a lot there: some of it helpful, some of it harmful. Within the category of what helps or hurts, some things do so a lot, and other things only a little. The legally trained mind recognizes and thrives on the detail in these distinctions. But to jurors, it is more likely the witness will end up in one of two buckets: “bad” or “good.” The good ones are credible and useful, and the bad ones have something going on — and maybe it is just one thing — that undermines their credibility. I often find when conducting post-verdict interviews of jurors that there will be one dominant reaction they recall from each witness — it could be an attitude or a single line from testimony — that they’ll latch onto and use for that categorization. Knowing the stakes can be unnerving for witnesses and the lawyers preparing them, but it is better to know. Everything matters, and it is critical to not just deliver the testimony, but to be useful and fully credible to the jury.
Binary Bias in Persuasion
I’ve written recently that when making strategic choices, it helps to consider more paths, and to not restrict yourself to just two options. In persuading, however, having fewer options often is better. One example of that might be the “Reptile” perspective that is popular with many plaintiffs’ attorneys. The idea behind the Reptile is that appeals to personal security help in speaking to the supposed reptilian brain that is attuned to these kinds of threats. However, the success of the strategy may be not be so much a matter of “awakening” the Reptile brain as a matter of appealing to people’s “binary bias.” For example, stripping away the nuance of medical decision-making and the standard of care, it is simply easier to think of a physician’s actions as being either “safe” or “unsafe.” As the research authors note, there are evolutionary causes for that kind of thinking. “The root of the binary bias may lie in behavioral control. Many critical behavioral outputs such as flight versus flight and sustenance versus poison are go/no-go (binary) decisions. These survival-relevant processes may shape the sorts of lower-level judgments made in the current studies.”
For litigators, it is helpful to understand and adapt to the ways your jurors or judges are going to be thinking. The binary bias is one of those ways, and reminds advocates to expect simplified decision-making, and to aid that simplification wherever possible.