Biotech Values

[DewDiligence]

I’m going to conduct a randomized controlled trial using the gaggle of geese that live at Town Hall in Wellesley, Mass:



I will bring two unmarked bags of cracked corn, one of which is non-GM. To ensure an unbiased experiment, I’ll select two people who don’t know what’s in the bags to dispense the corn from each bag simultaneously, while a third disinterested individual logs the amount of time for each ration of corn to be consumed. This will be done ten times on ten different days.

I will engage a biostatistician with no financial ties to either Monsanto or the Town Hall Geese to establish a statistical analysis plan including an appropriate non-inferiority margin for the primary endpoint of Time to Disappearance of all visible cracked corn. A secondary endpoint will be the volume of goose honking, rated on a 1 to 5 scale by the (blinded) food dispensers.

If you can make it to Wellesley, you can observe the experiment first-hand and place a wager on the outcome. I’ll even let you decide which biostatistician should design the SAP.

I presume from your posts on this board that you would be willing to bet that the GM corn will take much longer to be consumed than the non-GM corn, and hence the non-inferiority threshold will not be met for the primary endpoint. I’ll take the other side of this bet at even money for whatever amount you want to wager.