HSD: Democratic Epistemics in Lab-Based Processes

Fairleigh Dickinson University, Teaneck NJ

Investigators

Abstract

In pure science, the results of any one lab are subject to challenge from any other lab. In forensics and other areas, however, it may be unlikely for the results of one lab to be challenged by any other lab. In this research project, the investigators will study alternative network structures for lab-based social processes. They will conduct experiments with human subjects, constructing alternative network structures in the lab and comparing their performance. In initial studies, they will compare two basic structures. In the first, a Sender, who possibly represents a crime lab, will examine evidence and report the result to a Receiver, who represents a judge or jury. In the second, several labs will examine the same evidence and report their findings. The investigators expect the second system to have a lower error rate both in the messages sent to the Receiver and in the decisions made by Receivers based on those messages. The investigators will employ the techniques of economic systems design, which is a branch of experimental economics, but which applies the criterion of veracity rather than efficiency. Instead of asking "Does this system produce a more efficient allocation of resources?", the investigators ask "Does this system produce a lower rate of errors?" The project will provide insights of value to fields like medical testing, drug screening, and forensic science. Better forensic science should mean more justice. Better medical testing and drug screening should mean improved health and safety. The problem now is the monopoly element in these fields. In forensics, for example, only one lab looks at any piece of evidence. No lab is likely to look at evidence that went first to some other lab. Monopoly can create bad results, and it can lead to errors. Josiah Sutton, for example, was convicted of rape by DNA evidence. This evidence was later shown to be wrong, but Sutton spent four years in jail for a crime he did not commit. In medical testing and drug screening, too, mistakes create serious problems. The investigators expect this project to help the creation of backup systems. In forensics, for example, sometimes more than one lab often should examine the evidence. The project's experiments should help determine which backup systems are best.