Does the Format of the Message Affect What Is Heard? A Two-Part Study on the Communication of Violence Risk Assessment Data

ABSTRACT

There is currently limited understanding about how best to communicate the results of violence risk assessments to legal decision makers. To advance this literature, a two-part study was conducted assessing whether layperson perceptions of risk and dangerousness varied across common communication formats. Participants in both studies were exposed to a deidentified vignette about a defendant who was charged with a felony offense followed by expert testimony regarding that defendant’s level of risk for general violence. Study 1 (N = 103) compared perceptions across numerical, ordinal, and risk management (action-oriented) approaches using a sample of former criminal trial jurors. Results showed that, while exposure to numerical data was associated with the lowest ratings of risk, all participants overestimated risk of reoffending (ranging from about 40% to 62% compared to the expert’s estimate of 17% in 7 years and 31% in 10-years). Study 2 (N = 199) was an exploratory investigation of elaboration strategies to enhance application of numerical data in particular, using a sample consisting largely of undergraduates. Participants again overestimated risk regardless of condition, albeit to a slightly less extent than in Study 1. Based on these results, we make several recommendations to advance future research on risk communication, including using methodologies that approximate courtroom proceedings as closely as possible.

KEYWORDS:

• violence risk assessment
• risk communication
• actuarial
• structured professional judgment
• expert testimony

Acknowledgments

We thank Samuel Levulis, MS, Texas Tech University, and Steven R. Chestnut, PhD, University of Southern Mississippi, for statistical consultation. Camden E. Hoeffner was located at Texas Tech University at the time this research was conducted. Ms. Hoeffner is currently affiliated with the Department of Psychology at Sam Houston State University.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The total cost to researchers was $10 per participant (i.e., $3 for compensation and $7 for the recruitment service). QPM determines total cost based on the number of participants requested.

2. Although a different name was randomly generated for this study, the case information was otherwise identical.