Walden and Browne [11] indicated that CP 690550 convergence to the correct decision does not occur after 100 decisions in the WB model, which also applies to the G-WB model. Therefore, we have compared the results obtained after simulating 1000 replicates by 100 decision makers using the WB and G-WB models, and the results are shown in Fig. 4.
Fig. 4. Convergence with the WB and G-WB models.Figure optionsDownload full-size imageDownload as PowerPoint slide
Fig. 4 shows that the probability of making a correct decision is higher with the WB model than the G-WB model before the 15th15th decision maker (including the 15-th15-th decision maker), but the results are different after the 15-th15-th decision maker. Furthermore, the probability of making an incorrect decision is much lower with the G-WB model than the WB model with 100 decision makers. Therefore, our results demonstrate discontinuous variation the G-WB model is more precise than the WB model. However, more replicates are necessary to analyze the convergence further, so we have performed simulations of 10,000 decisions and the results are shown in Fig. 5.
Fig. 4. Convergence with the WB and G-WB models.Figure optionsDownload full-size imageDownload as PowerPoint slide
Fig. 4 shows that the probability of making a correct decision is higher with the WB model than the G-WB model before the 15th15th decision maker (including the 15-th15-th decision maker), but the results are different after the 15-th15-th decision maker. Furthermore, the probability of making an incorrect decision is much lower with the G-WB model than the WB model with 100 decision makers. Therefore, our results demonstrate discontinuous variation the G-WB model is more precise than the WB model. However, more replicates are necessary to analyze the convergence further, so we have performed simulations of 10,000 decisions and the results are shown in Fig. 5.