The Jury, John Morgan, 1861, Bucks County Museum, Aylesbury, England

In 2010, Professor Janet Randall of Northeastern University was invited by the Massachusetts Bar Association to be the linguist on their Jury Instruction Task Force. Following California's revision of their jury instructions, legal professionals in Massachusetts were concerned that those in the Bay State were incomprehensible and even dangerous to the legal process.
Professor Randall eagerly answered the call, assembling a team of undergraduates from Northeastern University to investigate the linguistic factors underlying Massachusetts's jury instructions. This team was intially named the Plain English Jury Instructions team (PEJI for short). From the beginning, PEJI was an interdisciplinary group, allowing for greater diversity of thought and creativity. After some preliminary research, the team prepared for its first study.

For our first study, 214 undergraduate students were divided into four conditions. Half heard recordings of the original Massachusetts jury instructions, while the other half heard instructions rewritten by our team into Plain English. Each of the two groups was then split into half again, one group receiving listening-only instructions and the other half receiving listening and reading instructions, where they could follow along in the texts. The subjects were then asked true/false questions about the jury instructions they had heard.

The design of our second study matched that of the first one, with one important distinction: the subjects. In our first study, we had experimented on undergraduate students. However, our subject pool did not nearly match that of a Massachusetts jury in several possibly confounding areas, especially education level. Therefore, for our second study, we studied subjects from MTurk, an online platform that allowed us to recruit a more diverse group of subjects.
In study three, we wanted to further emulate the courtroom. In studies one and two, subjects were presented with one instruction at a time and answered questions on it before moving on to the next one. However, real jurors have all of the instructions read to them in succession, and then have to follow those instructions in the case. We decided to test the effect of this by having our subjects go through all the instructions before answering questions. Study three was conducted using undergraduates.
Like study two, our fourth study replicates the previous study with a different subject pool. We again used MTurk to get subjects outside Northeastern University.
With the conclusion of study four, the lab is looking to conduct a study with a subject pool even closer to the real Massachusetts jury pool. In collaboration with Massachusetts court houses, we will be doing a new study using dismissed jurors.

In addition to this, the team has begun using computational tools to analyze jury instructions. Right now we have only looked at the correlation between word frequency and comprehension, but we hope this avenue of research will produce many insightful results.