[quote author=“cpl”]I’m really quite pleased to have the data from 39 of 43 studies brought to my attention. I was wrong, rationalists really are smarter than religionists and I can now rely upon data instead of “common sense” or the opinion of vested interests for that information to support that assertion.
Oh dear. Assuming this is not irony, where to begin.
I am not concerned that Dawkings misattributes a “meta-analysis” between two lightweight mags such as Mensa and Sceptic in a verbal interview. For that matter, it may have appeared in both.
What does trouble me is that he appears not to know what a meta-analysis is, since he says “in 39 out of 43 [studies]—that’s all but four—there is a correlation”. (So what? How good were those studies? I could design 400 studies to find a positive correlation between intelligence and religiosity and they would all be worthless.)
I am further troubled that he does not seem to be aware that meta-analysis is notoriously unreliable when it comes to research reports - let alone that what he mentioned probably wasn’t even a meta-analysis, since the website refers to the Sceptic piece as a summary. And Dawkins certainly describes it that way.
The New England Journal of Medicine rejected meta-analysis in favor of randomized clinical trials: that is to say, when the results of a meta-analysis conflict with the results of a properly done clinical trial, one is advised to accept the results of the trial.
Here is part of what the editors had to say in support of their position:
“[M]ajor problems with the implementation of meta-analyses have been common. There have been a wide variety of these, including failure of the investigator performing the meta-analysis to understand the basic issues, carelessness in abstracting and summarizing appropriate papers, failure to consider important covariates, bias on the part of the meta analyst, and, perhaps most often, overstatements of the strength and precision of the results. It is not uncommon to find that two or more meta-analyses done at about the same time by investigators with the same access to the literature reach incompatible or even contradictory conclusions. Such disagreement argues powerfully against any notion that meta-analysis offers an assured way to distill the ‘truth’ from a collection of research reports.”
This article was called “The promise and problems of meta-analysis” and it was published in 1997. You have to have a subscription to NEJM to access their web content now, or you can pay-per-view, so if anyone is interested go to the site and search.
Bear in mind that they were criticizing meta-analysis when it comes to the kind of data that can be obtained from clinical trials. Imagine how much worse it is in other areas of research.
In the end, meta-analysis works well for research that uses common (as in mutual… sigh) mathematical parameters but very badly for everything else. As someone here on the forum said recently, GIGO.
Dawkins made a fairly heavyweight claim. I’d have liked to see some equally sturdy evidence for it - or at least not merely something so flimsy. All he did was expose his position to ridicule with that statement. Then again, it was just an interview. Perhaps he has published all kinds of gold-standard data.
The last line “I feel smarter already!” is a quip, a brief moment of humor offered up at my own expense, a reference to what was clearly my ignorance which has now been cured by your direction to the evidence.
Again, assuming this is not irony, I apologize if you think that your ignorance is now cured by the page I directed you to. Such was not my intent. I find that my ignorance on the subject of intelligence vs. religiosity is no more enlightened by perusing that avalanche of information. That is to say, I am not more able to draw a conclusion one way or the other.
Being rational and somewhat historically inclined, I value real data over even the most expert of opinion. (One of my favorite books is “The Experts Speak,” a compendium of “expert” analysis which time has proven false.)
Indeed. It sounds like an excellent example of how hard it is to acquire “real data” - especially on topics as uncertain as “intelligence” and “religiosity”.
I’m also naturally inclined to see great humor in human folly, especially my own. I don’t see either of those traits as a problem. Apparently you do, for which I am truly sorry.
“‘In that case, tinkerty-tonk.’ And I meant it to sting.”
This forum does seem to be solemn, perhaps because we are all so very intelligent. It’s not usually that hard to recognize jokes in writing, even when someone’s being laconic AND ironic. (Then again, unintended humor can be the funniest of all.) So I suppose I need to tell you explicitly not to worry. I love a joke on myself. It’s a shame so few people are able to bring one off.