welcome redditors!to snoo-finity ... and beyond!

NBME 23 Answers

nbme23/Block 4/Question#12

Investigators conduct a prospective, ...

Strength of association, temporal relationship, dose-response gradient

Login to comment/vote.

 +2  upvote downvote
submitted by usmle11a(13),

guys watch this : https://www.youtube.com/watch?v=TnuosYuKGos

anyway (p.s i got it wrong)

A) dose-response = biological plausibility. p.s somehow equal. B) C) E) sensitivity; wrong D) my answer; consistency of other studies ( it wasnt applied to other communities)

usmle11a  guys watch this : https://www.youtube.com/watch?v=TnuosYuKGos anyway (p.s i got it wrong) A) dose-response = biological plausibility. p.s somehow equal. B) C) E) sensitivity; wrong D) my answer; consistency of other studies ( it wasnt applied to other communities) +1  
usmle11a  guys watch this : https://www.youtube.com/watch?v=TnuosYuKGos anyway (p.s i got it wrong) A) dose-response = biological plausibility. p.s somehow equal. B) C) E) sensitivity; wrong D) my answer; consistency of other studies ( it wasnt applied to other communities) +  




Can someone please further explain this question? What biostatistical analysis should I be thinking about?

vshummy  I got this wrong but best I could come up with was this was about Bradford Hill Criteria for establishing causality. And of the 9 included, F has the most that are actually included in the information given to us. I chose D but I think since we don't know about other study results, we can't include it as directly answering the question about *this* study. https://en.m.wikipedia.org/wiki/Bradford_Hill_criteria Someone double check me here: A: biologic plausibility is a weak point in the criteria, according to the wiki. Also probably not true in regards to this study. B: Sensitivity is not part of the criteria C: " " D: We don't know about consistency E: " B " +6  
mousie  Found this ... still confused about why A and D are wrong though... https://stats.stackexchange.com/questions/534/under-what-conditions-does-correlation-imply-causation +1  
2zanzibar  The three criteria for causality are: 1) empirical association (i.e. strength of association; a change in independent variable correlates or is associated with a change in dependent variable), 2) time order (i.e. temporal relationship; the independent variable must come before change in the dependent variable, or plainly stated, cause must come before effect). and 3) nonspuriousness (i.e. dose-response gradient; the relationship between 2 variables is due to a direct relationship between the two, not because of the actions of changes in a third variable... this can be evinced by a dose-dependent response). +5