Eyewitnesses Credible and Discredible1

Undergraduate subjects read descriptions of two armed robbery cases that varied in the degree to which physical and circumstantial evidence alone would evoke a judgment of guilt. The effect on such judgments of the addition of an eyewitness identifying the defendant, and the effect of discrediting such eyewitness testimony were studied, along with the effects of altering the evidentiary standard (reasonable doubt or preponderance of the evidence), in both between- and within-subjects designs. The different cases, the different conditions of testimony, and the different standards of proof had large effects on judgment. Testimony discrediting an eyewitness was fully effective in lowering the rate of guilt judgments, regardless of the strength of the physical and circumstantial evidence, the standard of proof, or whether or not subjects made serial judgments after reading successive increments to the summaries.