Further, anything like large scale efficacy should show not just a negative correlation, but a massive one. Massive.
I'll let you cherry-pick this time...show me one single world geography where its constituents show massive negative efficacy. Just one. There are hundreds. Just show one.
Further, anything like large scale efficacy should show not just a negative correlation, but a massive one. Massive.
I'll let you cherry-pick this time...show me one single world geography where its constituents show massive negative efficacy. Just one. There are hundreds. Just show one.
I'll go a step further: You state that you have systemically different results. Post them on a substack or blog for all to see by the end of the day...since you've already done the work as you claim. You don't have to do any particular great work explaining anything---you can save that part for another day if you want to define your functions. But you've staked your professional reputation on the claim of having done the work and found different results. Just a screenshot would suffice.
"Further, anything like large scale efficacy should show not just a negative correlation, but a massive one. Massive." This is not true. In statistics we actually have a name for why this is not true, Simpson's Paradox. There are plenty of situations in applied work, where something might show negative correlation in a population but positive correlation in every sub-group along one dimension, like sex, or the other way around, positive correlation in the population, but negative correlation in the sub-groups. This is especially true in treatment/control trials if the treatment is self-selected by patients. When people are choosing the treatment or control group, there may be confounders that hide the actual relationship between the treatment and disease.
Further, anything like large scale efficacy should show not just a negative correlation, but a massive one. Massive.
I'll let you cherry-pick this time...show me one single world geography where its constituents show massive negative efficacy. Just one. There are hundreds. Just show one.
I'll go a step further: You state that you have systemically different results. Post them on a substack or blog for all to see by the end of the day...since you've already done the work as you claim. You don't have to do any particular great work explaining anything---you can save that part for another day if you want to define your functions. But you've staked your professional reputation on the claim of having done the work and found different results. Just a screenshot would suffice.
https://loren-56681.medium.com/for-mathew-crawford-ccc0c556ab09
As you requested.
"Further, anything like large scale efficacy should show not just a negative correlation, but a massive one. Massive." This is not true. In statistics we actually have a name for why this is not true, Simpson's Paradox. There are plenty of situations in applied work, where something might show negative correlation in a population but positive correlation in every sub-group along one dimension, like sex, or the other way around, positive correlation in the population, but negative correlation in the sub-groups. This is especially true in treatment/control trials if the treatment is self-selected by patients. When people are choosing the treatment or control group, there may be confounders that hide the actual relationship between the treatment and disease.