I find ppl have difficulty understanding #evidencebasedmedicine (EBM) & how Science works.
EBM doesn't try to "prove" anything but is focused on REFUTATION. Let's say u have a new medicine which you believe works better than previous medicine/no medicine. Thats your theory...1/n
EBM doesn't try to "prove" anything but is focused on REFUTATION. Let's say u have a new medicine which you believe works better than previous medicine/no medicine. Thats your theory...1/n
#EBM sets out to refute this theory. Hence all trials start with a null hypothesis: New Treatment A is no better than placebo/treatment B.
You then use statistics to find out whether the difference between the two was likely to happen by chance. P < 0.05 only means that...2/n
You then use statistics to find out whether the difference between the two was likely to happen by chance. P < 0.05 only means that...2/n
there is less than 5% chance that the difference (if any) between the 2 treatments was likely to be due to chance, & is a real difference.
In this case, you reject the a priori Null hypothesis ( that there is no difference between 2 treatments) You accept...3/n
In this case, you reject the a priori Null hypothesis ( that there is no difference between 2 treatments) You accept...3/n
your original hypothesis that Treatment A is better than Treatment B/placebo, because you are unable to REFUTE this hypothesis. Until another study comes along.
Purists: sorry if this sounds too simplistic. Meant for Joe Public.
If you are interested, do read Karl Popper..END.
Purists: sorry if this sounds too simplistic. Meant for Joe Public.
If you are interested, do read Karl Popper..END.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
