In April 1987, the British Medical Journal announced with great fanfare that most doctors know very little about statistics. This startling and embarrassing conclusion came about after Wulff and colleagues sent a questionnaire to 250 Danish doctors to assess their knowledge of elementary statistical expressions (standard deviation, correlation coefficient, etc) and found “thought-provoking evidence of ignorance.”
From nine multiple choice questions, the doctors who replied produced a median correct response of 1.4. The authors concluded: “The statistical knowledge of most doctors is so limited that they cannot be expected to draw the right conclusions from those statistical analyses which are found in medical journals.” As for the 102 doctors who did not reply at all, they might have known even less! Of course, the extrapolation from 148 replies to the statement about “most doctors” is itself an example of studying a sample and then drawing global conclusions!
Be that as it may, the British Medical Journal took an appropriate stance when it pleaded that “journals need to ensure that results can be explained in terms that the statistically amblyopic can still comprehend. Expansive statements claiming ‘fact’ from probability should be discouraged and, despite the policy of some editors, all medical journals should allow space in their letter columns for disagreement and informed counter opinions.”
Matthew DR, McPherson K. Doctors’ ignorance of statistics. Br Med J, 7 1987, 294, 856.
🖨️ Print post
I’d like to do a survey asking how many doctors would readily conduct a “study” if it were paid for by large food company conglomerates, large chemical companies, or even political entities, then report it as health-groundbreaking news. Britain does this all the time–uses fuzzy math, flawed data (who doesn’t remember the global warming scandal?), and corporate/political backing, so I don’t trust anything coming out of Britain….and now they have Rupert Murdoch to further taint their reputation!