Figure 4
The two-panel grouped vertical bar chart is drawn on coordinate planes and compares ratings given by physicians and an A I advisor for two participant groups. In both panels, the vertical axis is labeled “Rating” and ranges from 0 to 6 in increments of 1 unit. Each panel contains grouped bars for two evaluators identified in the legend as “Physician” and “A I Advisor”. In panel (a), “Human participants”, the horizontal axis lists 3 categories from left to right as “Self”, “Individuated Other”, and “Average Other”, for a total of 6 bars. For “Self”, the Physician rating is 4.8, and the A I Advisor rating is 3.2. For “Individuated Other”, the Physician rating is 4.6, and the A I Advisor rating is 3.5. For “Average Other”, the Physician rating is 5.0, and the A I Advisor rating is 4.7. In panel (b), “G P T participants”, the horizontal axis lists 4 categories from left to right as “Self-A I”, “Self-User”, “Individuated Other”, and “Average Other”, for a total of 8 bars. For “Self-A I”, the Physician rating is 3.7 and the A I Advisor rating is 3.2. For “Self-User”, the Physician rating is 3.6 and the A I Advisor rating is 3.3. For “Individuated Other”, the Physician rating is 3.4 and the A I Advisor rating is 3.1. For “Average Other”, the Physician rating is 3.5 and the A I Advisor rating is 3.1. Note: All numerical values are approximated.Mean recommendation scores by advisor and advisee type. Panel (a) shows Longoni et al.’s (2019) participants; Panel (b) shows GPT participants (current study). Source: Authors’ own work