Rex Eupseiphos,
Thank for sharing your thoughts, it is an interesting conversation to have for the benefit of the other board members, as, in my opinion, there is no more important topic as this will decide the approval of Leronlimab or otherwise.
Some comments:
Yup, but that is problem with investing
one has to make decisions without all the data in hand. I wish we had the same population as the Montefiore cohort, meaning we had our placebo group, and more patients, ideally 75 or more. But this will be having the data from the trial, wouldn't it?
So, one can say (statistically speaking): I don't have enough information therefore I will not invest.
Or: with the data I have, publicly available, I run my statistics (imperfect as I don't have all the data available, of course) and reach an actionable conclusion that, perhaps, can help me to make the BEST decision and hopefully make money.
Or in lay terms: I have 88% reported with patients in N.Y. this number can be higher and lower (as you rightly point out). What happens then ?? Well, if the number is higher we are in better shape. If the number is lower one can ask: how much lower before what I already know means that Leronlimab did not help the patients in Montefiori??
And invest or not accordingly.
As I mentioned before in another post where this was discussed more extensively, even with this percentage going down we are still OK with the results of the small sample in Montefiori (if one decides to factor-in these results).
Now, of course, it might be that the patients in Montefiore where outliers, or the 2634 reported form the Richardson not representative, but, hell, we have to make do with what we have. Don't we??
The drawback from sharing one's (decision-making) results is that there is implicitly a risk-tolerance transfer to others. My simulations are good enough for me and they might not be for other investors. As always, this is not investment advise. I am just sharing my results with the hope it helps somebody.
The example is a two sample test. One is the Montefiore cohort (let's call it Patterson cohort) and the other is the N.Y. cohort (Richardson cohort). Now, the two samples have to be each independently obtained from a different given population. I cannot guarantee that none of the Patterson cohort patients was included in the Richardson cohort (violating the independence rule) but, given the large difference in numbers, this would not affect the results in a noticeable manner.
Thank for sharing your thoughts, it is an interesting conversation to have for the benefit of the other board members, as, in my opinion, there is no more important topic as this will decide the approval of Leronlimab or otherwise.
Some comments:
Quote:But that's only half the problem. You need a good sample to compare it to. "88%" is guaranteed to be wrong but we don't know how wrong or in what direction.
Yup, but that is problem with investing
one has to make decisions without all the data in hand. I wish we had the same population as the Montefiore cohort, meaning we had our placebo group, and more patients, ideally 75 or more. But this will be having the data from the trial, wouldn't it? So, one can say (statistically speaking): I don't have enough information therefore I will not invest.
Or: with the data I have, publicly available, I run my statistics (imperfect as I don't have all the data available, of course) and reach an actionable conclusion that, perhaps, can help me to make the BEST decision and hopefully make money.
Or in lay terms: I have 88% reported with patients in N.Y. this number can be higher and lower (as you rightly point out). What happens then ?? Well, if the number is higher we are in better shape. If the number is lower one can ask: how much lower before what I already know means that Leronlimab did not help the patients in Montefiori??
And invest or not accordingly.
As I mentioned before in another post where this was discussed more extensively, even with this percentage going down we are still OK with the results of the small sample in Montefiori (if one decides to factor-in these results).
Now, of course, it might be that the patients in Montefiore where outliers, or the 2634 reported form the Richardson not representative, but, hell, we have to make do with what we have. Don't we??
The drawback from sharing one's (decision-making) results is that there is implicitly a risk-tolerance transfer to others. My simulations are good enough for me and they might not be for other investors. As always, this is not investment advise. I am just sharing my results with the hope it helps somebody.
Quote:that there's a world of difference between a two-sample test and a one-sample test, and they will certainly use two-sample tests.
The example is a two sample test. One is the Montefiore cohort (let's call it Patterson cohort) and the other is the N.Y. cohort (Richardson cohort). Now, the two samples have to be each independently obtained from a different given population. I cannot guarantee that none of the Patterson cohort patients was included in the Richardson cohort (violating the independence rule) but, given the large difference in numbers, this would not affect the results in a noticeable manner.
7
