It should have been big victory for social science in the courtroom: Noting that lay jurors tend to give eyewitness testimony in criminal cases far more weight than it deserves, psychological researchers weighed in with recommendations, and remarkably, judges actually listened. In July of 2012, New Jersey's judiciary adopted new jury instructions that sought to educate jurors on the problems of human memory in the context of eyewitness identification. The new instructions were designed to function as a kind of tutorial on how to use and to value the testimony without exaggerating its veracity.  Specifically, the expanded instructions aimed at disabusing jurors of the notion that memory is "like a video recording," emphasizing that police identification procedures can be intentionally or unintentionally misleading, and spelling out a variety of factors that affect the ability to see and to remember that should be factored in. In addition, the instructions stressed that "a witness's level of confidence, standing alone, may not indicate the reliability of the identification." 

This, of course, was applauded by those who have studied the problem of eyewitness identification and false convictions, with Innocence Project Director Barry Scheck predicting that "these instructions will revolutionize the way that juries scrutinize identification evidence." The problem is, according to one of the first investigations to test whether it works, it doesn't. Or, at least, not quite. A team of University of Arizona researchers (Papailiou, Yokum, & Robertson, 2015) exposed mock jurors to a video-recorded reconstruction of a murder trial involving what the instructions would consider either strong or weak eyewitness testimony, and then participants heard either the standard instructions in common use throughout the states or the new and improved New Jersey instructions. Mock jurors were less likely to convict on the basis of eyewitness testimony after receiving the instruction, but were not more sensitive to quality. "The New Jersey instruction," the researchers concluded, "did not improve jurors' ability to discern quality; rather, jurors receiving those instructions indiscriminately discounted 'weak' and 'strong' testimony in equal measure." So the instructions seemed to be effective in highlighting the issue, but not entirely effective in correcting the problem. In this post, I will take a look at the research, noting some implications for all instructions, not just the eyewitness instructions. 

The Research: Greater Salience Without Better Sensitivity

In the naive view of criminal jurors, there is no stronger evidence than the finger from the stand pointing toward the defendant as the witness declares, "That is the man, right there." Given what we know about the frailty of human memory, the power of suggestion, and the inevitability of reconstruction, that can be misleading.  And the study at least showed that mock jurors receiving the improved New Jersey instruction got that, and were less likely to convict based on eyewitness testimony. Looking at the people convicted by eyewitness testimony who were later exonerated by DNA testing, criminal defense attorney Alan Zegas, interviewed by NPR's Nell Greenfieldboyce, noted, "there should be skepticism of eyewitness testimony." The problem is that the court's proper intention in the instructions is that there should be skepticism, not outright rejection. Strong eyewitness testimony should be used, and weak eyewitness testimony should not be. 

But on that score, the results of the study were less favorable for the instruction. "On the one hand, the use of the New Jersey instruction substantially reduced juror reliance on weak identification evidence, as compared to the Florida-based instruction," the University of Arizona team notes, "On the other hand, the New Jersey instruction also equally reduced juror reliance on strong identification evidence." In this case, the researchers speculate that their mock jurors may have taken from the instructions that they should discount testimony that falls beneath a certain threshold, but did not gain enough understanding of what that threshold should be.

The Implications: Don't Take Instructions for Granted

These results don't necessarily doom the effectiveness of more explicit or educational instructions on the eyewitness issue or other issues. One possible confounding factor that the authors acknowledge is the length of the instructions: fifteen minutes for the New Jersey instructions versus five minutes for the conventional instructions. In a one-week trial, that time wouldn't make a significant difference, but in the context of a shortened 35-minute trial, it is a very large proportional difference. But the fact that the New Jersey instructions amounted to a whopping 43 percent of the total information mock jurors received, and still did not equip the mock jurors to draw distinctions between weak and strong eyewitness testimony should not be comforting. 

Here are some implications that apply to all instructions.  

Don't Assume Comprehension or Utilization

As I've noted on these pages before, instructions to the average ears are dense, technical, and legalistic. Even "plain English" requires a fair amount of work from jurors at an analytical level. Empirical investigations speak to the practical need to take comprehension with a few grains of salt, since most analyses have found only a weak correlation between instructions and effects on juror behavior. In this study, most mock jurors self-reported that they understood the instructions, but the results did not support the conclusions that they were differentiating between strong and weak testimony, which was of course the main point of the instruction.  

Test Your Instructions

So the solution is to not take effectiveness for granted but to test it, just as this research team did. Instead of assuming, "If the jurors heard it, then they will understand it and apply it" scrutinize that assumption. One way to check is to use your anticipated instructions, or at least the key instructions, in a mock trial. If you want a more detailed examination, or perhaps a comparison between two or more instruction options, consider running focus groups that look primarily at the instructions. That can be an ideal way to answer the questions, "Will they understand them?" and more importantly, "How will they use them?"

Address Both the Informative and the Agenda-Setting Function of Instructions

Instructions are obviously designed to instruct and to inform, but that isn't all they do. They also convey a message of salience, or what is important. One way to think about the difference is to consider the mass media: They don't tell us what to think (at least we like to think so), but they are very effective at telling us what to think about. Even if they don't determine our attitudes, they do set an agenda. Interpreting this study results, it seems that New Jersey's expanded discussion serves to increase salience but not comprehension, and jurors get the message on the problems of eyewitness testimony, but are less clear on what to do about it. The authors explain, "Jurors might interpret the judge's decision to read the enhanced instruction -- one which belabors the inaccuracies of eyewitness testimony, without much counterbalanced discussion of its conditions of accuracy -- as an implicit signal that said testimony should, in fact, not be trusted. Why else, a juror might wonder, would the judge bother to make such a fuss about it." 

One overarching problem is that our current methods of instructing juries are not educationally grounded: People simply don't learn well in response to a list of abstract rules that are read to them. If we truly prioritized the effective understanding and use of law by juries, then the instruction phase would likely look much different. It would be concrete, applied, visual, and interactive. For the time being, though, it is unlikely that jurors will be instructed by an effective teacher using examples and visual aids, and following up with a question and answer period. So for better or worse (probably for worse), that means that our goal is to discover ways to wring as much meaning as we can from traditional jury instructions.