Things we'd like to simulate

[jjc54]

• Median appropriateness by RAND correlation to Kappa index
• Expert criteria measure (which will be industry-dependent, how to scale numerically?)
• Panel size ~ to statistical power (RAND recc. n = 9 or 15; why?)
• First -> second round discussion (in prelim code!)
• Impact of a third round with additional discussion (i.e., number of rounds & discussions)
• Impact of changing what median threshold is considered a "consensus"