[HN Gopher] Randomization in Controlled Experiments
___________________________________________________________________
Randomization in Controlled Experiments
Author : pykq
Score : 19 points
Date : 2026-03-17 15:01 UTC (3 days ago)
(HTM) web link (queue.acm.org)
(TXT) w3m dump (queue.acm.org)
| hackingonempty wrote:
| > So our farmer should compare a treatment group of turkeys that
| watches TV with a control group that doesn't.
|
| For human subjects if you do that you get an "A vs. A+B study"
| which has a high probability of bias towards intervention B. Just
| sitting and talking with someone in a white lab coat will improve
| outcomes over those who do not.
|
| You must compare to some sort of placebo and the closer it is to
| the intervention you are testing the better. Testing a drug with
| side effects against an inert placebo can break blinding and show
| an effect for the drug that isn't there (Type I error.)
|
| This experimental design is very common in the world of So-Called
| Alternative Medicine (SCAM) e.g. "patients receiving conventional
| treatment vs. patients receiving conventional treatment plus a
| chiropractic technique." If you look for this you will find it
| virtually every time a SCAM practitioner presents an RCT as
| evidence for its efficacy.
| madhadron wrote:
| The ACM should probably have run this by a statistician before
| publishing it. As another commenter pointed out it opens with bad
| experimental design, then goes on about random allocation in a
| way that makes it clear that the author doesn't understand the
| subject well.
|
| You don't need sequences of independent samples for assignment.
| You need sequences that are independent of the sequence of
| subjects being assigned. If you have a very small number of
| possible sequences then you can't get independence by the pigeon
| hole principle, but both 2^(64*N) (from N random 64 bit numbers)
| or 2^64 - 1 (for N samples from a PRNG) are so large that it
| doesn't matter.
|
| The argument about a sample of seeds not producing a flat
| histogram is also weird. Of course it doesn't. That's irrelevant
| to whether uniformly sampling a sequence of N 64 bit numbers from
| the space the PRNG provides produces independent sequences from
| the subjects.
___________________________________________________________________
(page generated 2026-03-20 23:01 UTC)