Marvin said-
My method has no dependency on a “Group C” except to have a total against which to establish a ratio. That is to say, the only dependency of the “Group C” you speak of in my method is to establish the number of JWs in New Zealand over the period of the study (1998-2007). This aggregate value for years 1998-2007 is: 126,989.
Of course your method has no dependency on Group C (NZ JWs), which is EXACTLY the point: in statistics, we try to draw conclusions ABOUT the population by SAMPLING it. As long as the sample is representative, we're OK. There are analyses which can be used to calculate the probabilty that the sample is valid (confidence intervals), but you obtain an invalid sample if you don't have reasonable level of confidence and certainty that it truly IS representative for what is claims to represent (the figure of JW deaths Worldwide).
Group C (NZ JWs: 13k people) is your (Marvin's) 'target population', and once again, you're forced to assume that Group C is a 'matched set' (in terms of it's various properties), when it's unproven that it DOES match; that issue hasn't been investigated yet (much less demonstrated), and unlike the medical researchers who participated in the study, it's impossble to create a 'matched set' by pulling charts with the same characteristics from the entire NZ population, since the patients were drawn from only a limited region. That is an assumption, and it's not only an assumption about ethnicity, but OTHER UNKNOWN FACTORS (eg variance in treatment skills, different in practice methods amongst various providers, etc) which might effect the results. You are ASSUMING that they're similar, but you cannot do that; you simply assume that the 13,000 JWs in NZ match the characteristics of the study participants.
That's the problem with working backwards, as you've done: you cannot assure that the characteristics of Group C match Group A (which was matched by creating a matched set by pulling medical records and including those who actually MATCHED, creating a Group B). Your method ASSUMES that the conditions that apply to the study participants apply to ALL JWs living in NZ, but that is an ASSUMPTION that is questionable, and should be understood by you as such, as otherwise you're only fooling yourself.
Marvin said-
I’ve not assumed any matched set against a world population. I’ve assumed that JW patients in New Zealand receive on average a better standard of care across the spectrum you cite compared with the average standard of care in the world. When it comes to comorbidities, hospital admission, treatment modalities, treatment protocols, etc. what we find among JWs in New Zealand is no different than the rest of the world overall, and I have not treated these as different.
Your assumption is not demonstrated to be valid to North America, much less the rest of the World. You ASSUME that JWs are not directed to go to area hospitals (eg Scripps in San Diego, CA) which attract skilled surgeons who specialize in performing bloodless surgery on JWs, where every area has a Healthcare Team Liason who knows to direct JWs to the hospital that is equipped to provide better specialized care. WHere any of the four hospitals known for providing experts trained and experienced in performing the latest techniques? This most definitely would be a factor for treating severe anemia, as the health care providers would have access and experience using different treatment protocols not available to a surgeon at a smaller regional hospital.
Another example?
NZ uses different protocols for severe anemia than the US, eg 'transfusion triggers', as recommended by various organizations, some of which don't apply to various countries, i.e. there is no standard treatment applicable Worldwide for severe anemia. That makes the original study problematic, since you assume the control group (the non-JWs) would be similar to non-JWs Worldwide. The standards used in NZ differ from those used in Zimbabwe, which are different from those used in India, etc.
Heck, even the skills of different providers vary, since they aren't robots: the practice of medicine allows some freedom in the treatment decisions that are made with patients, and malpractice rates vary amongst different Nations. This is why it's called "practicing medicine"; some surgeons nick more arteries during surgery, or overuse blood transfusions products in "First-World" hospitals, whereas some under-utilize blood transfusions; you're assuming a uniformity that just isn't proven.
In generating your wild hand-waving figure, you're assuming none of these factors matter or you tell yourself you've accounted for them "conservatively", but you're missing that EACH STEP TAKEN, EACH CLAIM YOU MAKE, REQUIRES supportive evidence to back it up: that's the entire REASON that so many footnotes exist in a journal article making as HUGE of a claim as yours!
You do realize that in science the one who makes the CLAIM has the burden of presenting supportive evidence (in this case of medical research, citing prior studies is the method)? Even though you may call it an "estimation" or "extrapolation", the fact is you are presenting it to the World as if it's a PROVEN FACT supported by actual investigation, when it's merely calculations you ran. It's what any first-year science student calls "dry labbing", making claims without doing any of the work. (Sorry, Marv, but digging up figures from JW publications or MEDLINE articles doesn't constitute scholarship or scientific investigation. That's not how science works).
Anything less is simply demonstrating your own stubborn refusal to follow standards of scientific methodology.
Marvin said-
As for ethnicity, in the New Zealand region there is a factor to consider regarding Maori, but this is adjusted for within the Beliaev study and it’s the findings of the Beliaev study driving my extrapolation.
So fine, the study supposedly accounted for the response of Maori JWs, but you don't KNOW that they actually respond MORE or LESS FAVORABLY to refusal of blood transfusions than European JWs or American Black JWs, or even how they respond to other conditions which cause anemia (such as sickle cell disease? Thallassemia?). You don't KNOW, but you're willing to ASSUME that it's the same, when in fact it MAY NOT.
Marvin said-
Otherwise I’ve made assumptions atop assumptions that, if anything, minimize the number of deaths due to Watchtower’s blood doctrine using the Beliaev data set.
We're getting somewhere by getting you to see that you're making assumptions. Now if youd only be able to see the uncertainty associated with your conclusion, we'd be moving forward....
Marvin said-
I recommend readers interested in this rather unique work authored by Beliaev and colleagues take time to review the original presentation in whole, and also their response to criticisms leveled by Shander and Isbister.
And you DO realize that the uniqueness of this study (as you say) doesn't work in your favor, right? A study that is "unique" or "one of a kind" is the converse of confidence-inspiring to anyone who understands science, since it means other scientists haven't validated its results.
Marvin said-
Assumptions matter for one reason, so readers can understand what’s being said. From what I’ve read of your complaints you neither understand assumptions I’ve made nor care to. That’s why I’ve not responded to more often than I have to what you’ve said.
Well, you probably don't care that I actually got a response from Dr. Beliaev, and he's glad that others have taken note of his study and are publicizing the results. If you want to get into the conversation, send me a PM and I'll send him your e-mail address (or I can cc you on the next exchange, etc). He's a generous and friendly fellow, and willing to discuss his results.
Adam