The Differential Effect of Numeracy and Anecdotes on the Perceived Fallibility of Forensic Science

Abstract

Contrary to popular belief, forensic science – including forensic DNA testing – is not infallible. The rate at which errors occur exerts an inordinate impact on the probative value of a DNA match. Previous research indicates that jurors are insensitive to this effect. The current study tests two possible explanations for the observed insensitivity: (1) juror innumeracy or (2) quantified error rates are not sufficiently vivid. Jury-eligible adults (n = 568) read a synopsis of a rape trial in which the quantified error rate was manipulated (either 1-in-10 or 1-in-100), as was the vividness of an error (the laboratory technician was anecdotally portrayed as: sloppy, biased, both or none). Overall, both manipulations affected participants’ verdicts. However, numerate participants were affected by the quantified error rate but not anecdotal information, whereas innumerate participants were affected by anecdotal information but not the error rate. The results indicate that the well-known effect on the use of abstract vs. concrete information is moderated by numeracy. On a practical level, the results suggest that, depending on their numerical proficiency, jurors rely on different types of information when evaluating the possibility that forensic scientific evidence is fallible.

Key words:

• forensic science
• judgment and decision-making
• numeracy

Note

Notes

1. This approach (splitting the sample based on numeracy scores) was utilized for the ease of interpretability of the results. The three-way interaction of error rate, anecdotal information and numeracy (as a log-transformed continuous variable) is statistically significant (Wald = 7.432, df = 3, p < .05). Note that the strong negative skew (–1.92, Kurtosis = 4.91) of the numeracy scores necessitated the logarithmic transformation (Mosteller & Tukey, 1977).