so you think Samsung shouldn't've recalled 2.5M phones just because a few dozen exploded? http://www.komando.com/happening-now/373069/samsungs-fix-for-exploding-smartphones-dumps-responsibility-on-end-users …
-
-
I suggest an RCT. Buy 4000 Samsung S7. Control: Buy 4000 iPhones. Result: No significant difference in explosions.
1 reply 0 retweets 1 like -
Replying to @VeryGummy @mihotep
One of those studies was of 800,000 people, more than enough statistical power to determine an increase.
1 reply 0 retweets 0 likes -
An RCT with 800k subjects?
1 reply 0 retweets 0 likes -
Replying to @VeryGummy @mihotep
Nope the RCT had up to 10k the observational studies with similar results were up to 800k
1 reply 0 retweets 0 likes -
Even if you surveyed the entire population you couldn't filter out a 0.6‰ point increase in risk.
1 reply 0 retweets 0 likes -
Replying to @VeryGummy @mihotep
That would require approx. 250k individuals to demonstrate a stat significant difference.
1 reply 0 retweets 0 likes -
Only if it's all random noise. Even the weakest confounders or measurement problems would bury such a tiny effect.
1 reply 0 retweets 0 likes -
Replying to @VeryGummy @mihotep
You'd ideally want several types of evidence, as well as systematic reviews combining a number of studies...
1 reply 0 retweets 0 likes -
worth noting that if the effect is .06% then of those 659 suicides you'd expect ~3 caused by Chantix.
2 replies 0 retweets 0 likes
You can't have it both ways; if the effect size is tiny then almost no one is effected.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.