B is better because B points out that the survey might have under-estimated the prevalence of P in the general population. If 5% of the general population have P, then P is unlikely a cause for the fatal accidents. ? Some arguments will tell you that two events are correlated (happened together) and then conclude that one event causes the other. The problem with this kind of causal argument is that it relies on many questionable assumptions. If the conclusion states that A causes B (If reading my posts, a reader’s understanding of CR is increased), look for evidence that says - B causes A (A reader with good understanding of CR tends to read my posts). - C causes B or A or both (CDer whose English is superb are more likely to read my posts and have better understanding of CR). -A and B are not actually correlated (The survey overestimated the increased understanding of CR for most readers after reading my posts).
-- by 会员 sdcar2010 (2011/12/17 21:41:38)
谢谢! |