Comments on: The Control Group Is Out Of Control http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/ In a mad world, all blogging is psychiatry blogging Fri, 24 Jul 2015 01:20:09 +0000 hourly 1 http://wordpress.org/?v=4.2.3 By: other mind meditation | Meditation Stuff (@meditationstuff) http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-120730 Sun, 13 Jul 2014 01:23:35 +0000 http://slatestarcodex.com/?p=1921#comment-120730 […] [12] http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/ […]

Report comment

]]> By: Phil Goetz http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-114180 Thu, 03 Jul 2014 16:14:13 +0000 http://slatestarcodex.com/?p=1921#comment-114180

The results? Schlitz’s trials found strong evidence of psychic powers, Wiseman’s trials found no evidence whatsoever.

Take a second to reflect on how this makes no sense.

It makes perfect sense. Schlitz has psychic powers. Wiseman doesn’t. They need to redo the experiment, keeping Schlitz as the starer in both groups.

Report comment

]]> By: Science smorgasbord 2 | Deadline island http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-101569 Sun, 15 Jun 2014 02:38:08 +0000 http://slatestarcodex.com/?p=1921#comment-101569 […] The Control Group Is Out Of Control […]

Report comment

]]> By: Ellie Kesselman http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-89061 Tue, 27 May 2014 20:11:44 +0000 http://slatestarcodex.com/?p=1921#comment-89061 Another online acquaintance: Gwern of Disqus comments, who has found (sometimes-amusing) fault with my comments on inane The Atlantic posts.

So. You like writing about Haskell, the Volokh Conspiracy, bitcoin and the effectiveness of terrorism. Goldman Sachs has not been extant for 300 years. I was saddened by your blithe dismissal of Cantor-Fitzgerald, post-9/11. I worked, briefly, for Yamaichi Securities, on floor 98 of Tower 2, but several years after the 1993 WTC explosion.

Please consider dropping by for a visit on any of my Wikipedia talk pages. You have 7 years’ seniority to me there. David Gerard is a decent person. He wrote your theme song, the mp3, so you must have some redemptive character traits :o) I am FeralOink, a commoner.

Report comment

]]> By: Ellie Kesselman http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-88906 Tue, 27 May 2014 14:49:18 +0000 http://slatestarcodex.com/?p=1921#comment-88906 Yes, Ilya Shpitser! I am a mere statistician and data analyst, doubter of Jonah Lehrer’s veracity, ignorantly idolatrous in my continued use of Neyman, Pearson and Fisher. I love validation.

I recognize your name. You had a lively, cordial conversation with jsteinhardt on LW, following his Fervent Defense of Frequentist Statistics. I smiled with delight as I read of your commitment there.

Report comment

]]> By: Caroline Watt http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-86422 Fri, 23 May 2014 20:34:26 +0000 http://slatestarcodex.com/?p=1921#comment-86422 Love your post, and v pleased that you approve of our KPU trial registry :-)
Just FYI, the other parapsych pre-registry that you refer to (the one Richard Wiseman and I set up for Bem replications) dates back to November 2010 and is no longer active.

Report comment

]]> By: Put Down Artist http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-85264 Thu, 22 May 2014 09:32:32 +0000 http://slatestarcodex.com/?p=1921#comment-85264 “That doesn’t tell you much until you take some other researchers who are studying a phenomenon you know doesn’t exist – but which they themselves believe in – and see how many of them get positive findings.”

This statement made my jaw drop. The illogic is stunning. You have made an assumption, assumed this assumption to be true, and are then deriding the people who are researching the question with an open mind.

I’m pretty sure this doesn’t need to be explained to you, but you don’t, in fact, know that these phenomenon don’t exist until you have studied them scientifically. No ifs, ands or buts. Possible bias by researchers has to be taken into account when considering their results, and when they themselves formulate their experiments – in any scientific research.

However, it is ridiculous to assert that because you personally don’t believe in something anyone trying to determine whether there is a way to scientifically measure and validate alleged phenomenon is automatically wrong.

This is a very worrying thought pattern I see from ‘skeptics’ all the time. It is not skepticism at all, it is a kneejerk response to things that threaten their personal world view and belief system. By this definition 99% of the people on the planet are ‘skeptics’, and the only genuine skeptics are those prepared to challenge their own world views.

I’m concerned that someone could actually take that statement seriously, so, allow me to assert that to actually prove that a group of researchers science is wrong, you actually have to go into their methodology and conclusions and find errors. To take issue with their conclusions and then assume that because you don’t like them their methodology must be flawed is the the least scientific approach imaginable.

Sheesh.

Report comment

]]> By: Anonymous http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-85147 Thu, 22 May 2014 06:08:26 +0000 http://slatestarcodex.com/?p=1921#comment-85147 It’s possible the 20m distance and however many walls wasn’t enough for sensory isolation, and one of the starers made detectable sound when moving to look at/away from the screen.

Blinding the sender to the experimental condition would avoid both accidental and malicious back channels like this. One possible design would be to have the video be sometimes delayed by 30 seconds, which would let you separate the effect of “the receiver is being watched” and “the sender thinks they’re watching”.

Report comment

]]> By: Das Versagen der Religionen - Seite 7 http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-75365 Sun, 11 May 2014 11:09:09 +0000 http://slatestarcodex.com/?p=1921#comment-75365 […] […]

Report comment

]]> By: Norm DeLisle http://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/#comment-73799 Fri, 09 May 2014 17:18:17 +0000 http://slatestarcodex.com/?p=1921#comment-73799 Very Nice! A couple of other observations. My father was a process development engineer at Dow Chemical. His design work in this area was to take a research result and test it’s commercial viability in a sizable plant process, a kind of enhanced replication. Researchers thought this was grunt work (the persistent attitude toward replication), and that all that was required was to make what they did in the laboratory bigger. But when you increase the size of a process by 5-7 orders of magnitude, a great many things become different. The lesson is that it isn’t always clear which changes in experimental conditions are important to the outcome, and the opinion of the researcher is a poor guide.
The second observation is from an article I read from the late 70-80s. It was a test of the hypothesis that niacin in large doses reduced the symptoms of schizophrenia. The design was double blind and the people who evaluated the improvement didn’t know who was receiving the niacin. However, the waiting room for the people being evaluated would hold 4-5 people at a time. Naturally they talked, and because of the niacin flush, they quickly figured out who was on placebo. It was, incidentally easy to figure out what the drug was, too. The conclusion was that one-third of the people on placebo broke the blind and went out and bought niacin. No one told the researchers because there were incentives for not telling the blind had been broken. Experimental conditions includes everything, not just what the researcher thinks is critical to successful publication.
Also, there is some evidence that placebos work even when people know they are placebos.

Report comment

]]>