Judgments of probable guilt

Although the arguments did not affect the number of subjects making judgments consistent with fallacious reasoning, they did affect subjects’ judgments of probable guilt. As Table 29.3 shows, subjects thought the detective’s subjective probability of guilt should be higher after reading the Prosecution argument ( M = .31) than after reading the Defense argument ( M = .24; F (1,71) = 7.89, p < .01). Order of presentation also infl uenced these judgments. Subjects in the Pros-Def condition thought the detective’s estimate of guilt should be higher than did subjects in the Def-Pros condition, F(1,71) = 8.44, p < .005. The order effect is probably due to a simple anchoring phenomenon. Subjects who received the Prosecution argument fi rst made higher initial judgments than subjects who received the Defense argument fi rst. The initial judgments then served as an anchor point for subjects’ second judgments. It is interesting to note, however, that a 2 × 2 analysis of variance examining the effects of type of argument (Prosecution vs. Defense) and order of presentation (Pros-Def vs. Def-Pros), revealed a signifi cant argument by order interaction, indicating there was less variation by type of argument in the Def-Pros condition than in the

Pros-Def condition [ F (1,71) = 4.43, p < .05]. One way of looking at this fi nding is that the Defense argument, when received fi rst, “anchored” subjects’ judgments more fi rmly than did the Prosecution argument. This interpretation is, of course, consistent with the previously noted fi nding that the Defense argument was more persuasive than the Prosecution argument.