“It is especially telling that no matter what landmark we applied to the methodology, we have yet to get a negative result. There is, in fact, a real chance that you simply cannot get a negative result from this method.”
There could be a few reasons that contribute. My guess, and I haven’t completely proven this yet, is that the proximity metric (1/distance) drives the result. My coauthor ran the methodology with randomized outcomes and still got positive results. That, along with a few other tests, suggests the proximity metric is weighting and biasing the outcomes.
Thanks. I’ll be interested to see how the Harvard authors respond. Their papers, as you’ve explained in detail, seem extraordinarily weak, even without this apparent bias in their methodology that you’ve observed.
“It is especially telling that no matter what landmark we applied to the methodology, we have yet to get a negative result. There is, in fact, a real chance that you simply cannot get a negative result from this method.”
Do you have a guess for why this might be true?
There could be a few reasons that contribute. My guess, and I haven’t completely proven this yet, is that the proximity metric (1/distance) drives the result. My coauthor ran the methodology with randomized outcomes and still got positive results. That, along with a few other tests, suggests the proximity metric is weighting and biasing the outcomes.
Thanks. I’ll be interested to see how the Harvard authors respond. Their papers, as you’ve explained in detail, seem extraordinarily weak, even without this apparent bias in their methodology that you’ve observed.