1462: "Blind Trials"

This forum is for the individual discussion thread that goes with each new comic.

Moderators: Moderators General, Prelates, Magistrates

Eutychus
Posts: 448
Joined: Mon Jan 25, 2010 6:01 am UTC
Location: France

1462: "Blind Trials"

Postby Eutychus » Fri Dec 19, 2014 8:28 am UTC

Image
title="Plus, you have to control for the fact that some people are into being blindfolded."

Eight out of ten owners said their cats preferred it.
Be very careful about rectilinear assumptions. Raptors could be hiding there - ucim

User avatar
bachaddict
Handel Played it Better
Posts: 484
Joined: Wed Dec 19, 2012 7:18 am UTC
Location: Aotearoa

Re: 1462: "Blind Trials"

Postby bachaddict » Fri Dec 19, 2014 8:31 am UTC

How do you control for that? Use transparent blindfolds?
slinches wrote:Also, the OTC isn't a disease. In fact, it's the cure. As we all know, Time heals all wounds.

Thanks for the molpish wig ggh!
he/him/his

User avatar
azule
Saved
Posts: 2132
Joined: Mon Jul 26, 2010 9:45 pm UTC
Location: The land of the Golden Puppies and Rainbows

Re: 1462: "Blind Trials"

Postby azule » Fri Dec 19, 2014 8:40 am UTC

Blind them with acid while they have "sex". Because of the intense pain and disorientation, they won't realize a sugar pill is not sex. The people having sex will just have sex. Lucky fuckers.
Image

If you read this sig, post about one arbitrary thing you did today.

I celebrate up to six arbitrary things before breakfast.
Time does drag on and on and contain spoilers. Be aware of memes.

User avatar
rhomboidal
Posts: 801
Joined: Wed Jun 15, 2011 5:25 pm UTC
Contact:

Re: 1462: "Blind Trials"

Postby rhomboidal » Fri Dec 19, 2014 8:57 am UTC

Some people think that successfully copulating once every vernal equinox counts as "lots of sex." Another small problem.

User avatar
BlitzGirl
Posts: 9117
Joined: Mon Sep 20, 2010 11:48 am UTC
Location: Out of the basement for Yip 6! Schizoblitz: 115/2672 NP
Contact:

Re: 1462: "Blind Trials"

Postby BlitzGirl » Fri Dec 19, 2014 9:02 am UTC

There's got to be a fetish for sugar pills.
Knight Temporal of the One True Comic
BlitzGirl the Pink, Mopey Molpy Mome
Spoiler:
Image
Image
Image<Profile
~.Image~.FAQ->Image

User avatar
orthogon
Posts: 3099
Joined: Thu May 17, 2012 7:52 am UTC
Location: The Airy 1830 ellipsoid

Re: 1462: "Blind Trials"

Postby orthogon » Fri Dec 19, 2014 9:13 am UTC

I may be wrong, but so far hasn't she described a single-blind trial? Somehow the researchers have got to not know whether they're administering a sugar pill or a hotel room with champagne on ice and Barry White on the stereo.
xtifr wrote:... and orthogon merely sounds undecided.

User avatar
eviloatmeal
Posts: 571
Joined: Thu Dec 11, 2008 9:39 am UTC
Location: Upside down in space!
Contact:

Re: 1462: "Blind Trials"

Postby eviloatmeal » Fri Dec 19, 2014 9:32 am UTC

*** FREE SHIPPING ENABLED ***
Image
Riddles are abound tonightImage

User avatar
da Doctah
Posts: 995
Joined: Fri Feb 03, 2012 6:27 am UTC

Re: 1462: "Blind Trials"

Postby da Doctah » Fri Dec 19, 2014 10:14 am UTC

"One group will get a medication to treat hypoglycemia. The control group will instead get sugar pills."

User avatar
orthogon
Posts: 3099
Joined: Thu May 17, 2012 7:52 am UTC
Location: The Airy 1830 ellipsoid

Re: 1462: "Blind Trials"

Postby orthogon » Fri Dec 19, 2014 10:47 am UTC

I'd always wondered about double-blind trials of acupuncture. You would have to train a load of practitioners some "placebo" version of acupuncture with all the "wrong" pressure points and the wrong types of needles or whatever. Afterwards you'd have to compensate them for their wasted time and course fees. Ideally you'd make it triple-blind, with the teachers unaware that they're teaching the placebo version, but of course there's the problem of infinite regress. Eventually everyone would forget which was supposed to be the "real" version, at which point we could see if there was statistically significant evidence from the clinical trials.

In fact, the same problem exists for non-pharmaceutical interventions in conventional medicine, or "medicine" as Ben Goldacre calls it. It's possible that appendectomies don't actually cure appendicitis, it's the placebo effect of having your abdomen cut open and being given a lot of care and attention (not to mention seriously good opiates).
xtifr wrote:... and orthogon merely sounds undecided.

AdamAitch
Posts: 2
Joined: Sun Apr 28, 2013 9:16 pm UTC

Re: 1462: "Blind Trials"

Postby AdamAitch » Fri Dec 19, 2014 11:05 am UTC

With a decent supply of those pills you might be able to take over the world.

User avatar
Eternal Density
Posts: 5589
Joined: Thu Oct 02, 2008 12:37 am UTC
Contact:

Re: 1462: "Blind Trials"

Postby Eternal Density » Fri Dec 19, 2014 11:10 am UTC

I think I've been having sugar all my life.
Play the game of Time! castle.chirpingmustard.com Hotdog Vending Supplier But what is this?
In the Marvel vs. DC film-making war, we're all winners.

User avatar
eviloatmeal
Posts: 571
Joined: Thu Dec 11, 2008 9:39 am UTC
Location: Upside down in space!
Contact:

Re: 1462: "Blind Trials"

Postby eviloatmeal » Fri Dec 19, 2014 11:15 am UTC

orthogon wrote:In fact, the same problem exists for non-pharmaceutical interventions in conventional medicine, or "medicine" as Ben Goldacre calls it. It's possible that appendectomies don't actually cure appendicitis, it's the placebo effect of having your abdomen cut open and being given a lot of care and attention (not to mention seriously good opiates).

For one, I don't think that's entirely how double-blind studies work; I'm pretty sure as long as the person analysing the data is blinded (as well as the subject, of course), then that's the definition of a double-blind trial. I.e, a surgeon performing surgery isn't exactly imparting a lot of clues as to the legitimacy of the surgery to a patient who is knocked out, and I imagine you could have an entirely separate set of nurses and such doing the after-care, so that they aren't biased either. Edit: But just as a note, I think it'd become clear fairly quickly that this particular placebo didn't work, by way of patients waking up and still screaming and vomiting.

But what do I know, I'm not a science... goodly... person.

As for acupuncture, they do conduct ("single-")blind studies of acupuncture. For the control group, I think they simply have the legitimate acupuncturist poke people in places which, according to his or her training, shouldn't elicit any effect. Actually, in this study they just poked the patients with a toothpick instead of an acupuncture needle.
Last edited by eviloatmeal on Fri Dec 19, 2014 11:22 am UTC, edited 1 time in total.
*** FREE SHIPPING ENABLED ***
Image
Riddles are abound tonightImage

User avatar
Philbert
Posts: 32
Joined: Mon Jan 05, 2009 12:32 pm UTC

Re: 1462: "Blind Trials"

Postby Philbert » Fri Dec 19, 2014 11:21 am UTC

If they have some of these sugar pills left that can make you think you are having sex, I'm willing to take them off their hands.

User avatar
CocoaNutCakery
Posts: 66
Joined: Fri May 24, 2013 6:26 am UTC
Contact:

Re: 1462: "Blind Trials"

Postby CocoaNutCakery » Fri Dec 19, 2014 2:32 pm UTC

Any time I bring up a study where they test a particular proceedure or the efficacy of an exercise (with one group doing a comparable exercise) and someone essentially suggests this, just...

Rage. Rage everywhere.

The really sad part is that I actually had a doctor, an actual medical doctor, that dismissed a comparative exercise study because it wasn't double blinded.

User avatar
PinkShinyRose
Posts: 835
Joined: Mon Nov 05, 2012 6:54 pm UTC
Location: the Netherlands

Re: 1462: "Blind Trials"

Postby PinkShinyRose » Fri Dec 19, 2014 2:37 pm UTC

eviloatmeal wrote:
orthogon wrote:In fact, the same problem exists for non-pharmaceutical interventions in conventional medicine, or "medicine" as Ben Goldacre calls it. It's possible that appendectomies don't actually cure appendicitis, it's the placebo effect of having your abdomen cut open and being given a lot of care and attention (not to mention seriously good opiates).

For one, I don't think that's entirely how double-blind studies work; I'm pretty sure as long as the person analysing the data is blinded (as well as the subject, of course), then that's the definition of a double-blind trial. I.e, a surgeon performing surgery isn't exactly imparting a lot of clues as to the legitimacy of the surgery to a patient who is knocked out, and I imagine you could have an entirely separate set of nurses and such doing the after-care, so that they aren't biased either. Edit: But just as a note, I think it'd become clear fairly quickly that this particular placebo didn't work, by way of patients waking up and still screaming and vomiting.

But what do I know, I'm not a science... goodly... person.

As for acupuncture, they do conduct ("single-")blind studies of acupuncture. For the control group, I think they simply have the legitimate acupuncturist poke people in places which, according to his or her training, shouldn't elicit any effect. Actually, in this study they just poked the patients with a toothpick instead of an acupuncture needle.

I think the person recording the data may be more important. But I also think double-blind is ambiguous to who is blinded aside from the patient. And of course medical ethics committees tend to be critical of trials in which your human control group is cut open and closed without doing anything else.

CocoaNutCakery wrote:Any time I bring up a study where they test a particular proceedure or the efficacy of an exercise (with one group doing a comparable exercise) and someone essentially suggests this, just...

Rage. Rage everywhere.

The really sad part is that I actually had a doctor, an actual medical doctor, that dismissed a comparative exercise study because it wasn't double blinded.

Why was that wrong? What kind of exercise? Who was not blinded for what reason? Were there other studies supporting their conclusions?

User avatar
Angua
Don't call her Delphine.
Posts: 5941
Joined: Tue Sep 16, 2008 12:42 pm UTC
Location: UK/[St. Kitts and] Nevis Occasionally, I migrate to the US for a bit

Re: 1462: "Blind Trials"

Postby Angua » Fri Dec 19, 2014 2:53 pm UTC

eviloatmeal wrote:
orthogon wrote:In fact, the same problem exists for non-pharmaceutical interventions in conventional medicine, or "medicine" as Ben Goldacre calls it. It's possible that appendectomies don't actually cure appendicitis, it's the placebo effect of having your abdomen cut open and being given a lot of care and attention (not to mention seriously good opiates).

For one, I don't think that's entirely how double-blind studies work; I'm pretty sure as long as the person analysing the data is blinded (as well as the subject, of course), then that's the definition of a double-blind trial. I.e, a surgeon performing surgery isn't exactly imparting a lot of clues as to the legitimacy of the surgery to a patient who is knocked out, and I imagine you could have an entirely separate set of nurses and such doing the after-care, so that they aren't biased either. Edit: But just as a note, I think it'd become clear fairly quickly that this particular placebo didn't work, by way of patients waking up and still screaming and vomiting.


I think for double-blinding it's the patient and the administrator who should both be blinded.

You do get single-blinded surgeries that show the surgery is as effective as control (arthroscopies for knees - I think the trial was done in Texas?). Sometimes blinding doesn't work (eg apparently that was a problem with some HIV medication trials - an effective on the medication was a change in the size of red blood cells, which the patients found out about, then went to another clinic, found out which branch they were in and shared the real drug with those getting the placebo).
Crabtree's bludgeon: “no set of mutually inconsistent observations can exist for which some human intellect cannot conceive a coherent explanation, however complicated”
GNU Terry Pratchett

User avatar
mathmannix
Posts: 1451
Joined: Fri Jul 06, 2012 2:12 pm UTC
Location: Washington, DC

Re: 1462: "Blind Trials"

Postby mathmannix » Fri Dec 19, 2014 3:54 pm UTC

Eternal Density wrote:I think I've been having sugar all my life.

Sweet.
I hear velociraptor tastes like chicken.

User avatar
Jackpot777
Posts: 328
Joined: Wed Sep 14, 2011 1:19 pm UTC

Re: 1462: "Blind Trials"

Postby Jackpot777 » Fri Dec 19, 2014 4:26 pm UTC

azule wrote:Blind them with acid while they have "sex". Because of the intense pain and disorientation, they won't realize a sugar pill is not sex. The people having sex will just have sex. Lucky fuckers.


Literally.

It reminded me of a scene in Annie Hall.

Alvy Singer's Therapist: How often do you sleep together?

Annie Hall's Therapist: Do you have sex often?

Alvy Singer: [lamenting] Hardly ever. Maybe three times a week.

Annie Hall: [annoyed] Constantly. I'd say three times a week.

ps.02
Posts: 378
Joined: Fri Apr 05, 2013 8:02 pm UTC

Re: 1462: "Blind Trials"

Postby ps.02 » Fri Dec 19, 2014 5:17 pm UTC

Is it normal to present to an audience a trial you haven't done yet? I don't really know because I don't science, but my impression has been that before the fact, you just write grant applications and send them to committees; the live presentations come after the research.

User avatar
Envelope Generator
Posts: 582
Joined: Sat Mar 03, 2012 8:07 am UTC
Location: pareidolia

Re: 1462: "Blind Trials"

Postby Envelope Generator » Fri Dec 19, 2014 5:30 pm UTC

da Doctah wrote:"One group will get a medication to treat hypoglycemia. The control group will instead get sugar pills."


If there's ever been a more damn-near-choked-on-my-sandwich case of beating Randall at his own joke, I want to see it.

edit: ...after I've finished the sandwich.
I'm going to step off the LEM now... here we are, Pismo Beach and all the clams we can eat

eSOANEM wrote:If Fonzie's on the order of 100 zeptokelvin, I think he has bigger problems than difracting through doors.

User avatar
Angua
Don't call her Delphine.
Posts: 5941
Joined: Tue Sep 16, 2008 12:42 pm UTC
Location: UK/[St. Kitts and] Nevis Occasionally, I migrate to the US for a bit

Re: 1462: "Blind Trials"

Postby Angua » Fri Dec 19, 2014 7:00 pm UTC

No talk about how trials can't all be the same is not complete without:

http://www.bmj.com/content/327/7429/1459
Crabtree's bludgeon: “no set of mutually inconsistent observations can exist for which some human intellect cannot conceive a coherent explanation, however complicated”
GNU Terry Pratchett

User avatar
sorceror
Posts: 53
Joined: Thu Jan 29, 2009 7:26 pm UTC
Contact:

Re: 1462: "Blind Trials"

Postby sorceror » Fri Dec 19, 2014 7:01 pm UTC

For those situations when a double-blind trial is impossible, an interesting new statistical tool that can, in the right circumstances, determine causation from correlation: https://medium.com/the-physics-arxiv-blog/cause-and-effect-the-revolutionary-new-statistical-test-that-can-tease-them-apart-ed84a988e

Obvious in retrospect, almost. Someday I hope to have an idea that's obvious in retrospect like that.

User avatar
mathmannix
Posts: 1451
Joined: Fri Jul 06, 2012 2:12 pm UTC
Location: Washington, DC

Re: 1462: "Blind Trials"

Postby mathmannix » Fri Dec 19, 2014 7:50 pm UTC

Envelope Generator wrote:
da Doctah wrote:"One group will get a medication to treat hypoglycemia. The control group will instead get sugar pills."


If there's ever been a more damn-near-choked-on-my-sandwich case of beating Randall at his own joke, I want to see it.

edit: ...after I've finished the sandwich.

Oh wow, I just got that! (I somehow read it as "glaucoma" the first time, I guess...) +1 for the Doctah.
I hear velociraptor tastes like chicken.

User avatar
Dmytry
Posts: 68
Joined: Mon Jun 01, 2009 12:39 pm UTC
Contact:

Re: 1462: "Blind Trials"

Postby Dmytry » Fri Dec 19, 2014 8:27 pm UTC

The way I think of it, if the claimed effect is small, close in magnitude to the observed magnitude of placebo effect, it shouldn't really matter so much to anyone except those researching biochemistry if it is pure placebo effect or not. If the cure is very close to the placebo effect, it is also almost entirely ineffective.

It's like we have +-5% electronic kitchen scales (internally four force sensors under the feet), which can be brought down to +-1% if we carefully place the weight on an exact same spot, with the scales on same hard surface tested for levelness. The procedure for bringing it down to 1%, while perhaps very interesting and even fascinating, has no importance to most cooks.

orthogon wrote:In fact, the same problem exists for non-pharmaceutical interventions in conventional medicine, or "medicine" as Ben Goldacre calls it. It's possible that appendectomies don't actually cure appendicitis, it's the placebo effect of having your abdomen cut open and being given a lot of care and attention (not to mention seriously good opiates).

Well, the effect size is really huge compared to anything ever measured for placebos in that condition, so we know it works. The non-blind trials have an error margin, the double-blind trials have an error margin, when the effect is substantially larger than the error margin we know it works.

Thinking in absolutes is what's the problem. Any effect, however minuscule, is sufficient to deem something an "effective treatment"; lack of effect is sufficient to deem it "snake oil". As a consequence, we have this proliferation of very, very weak cures (to conditions we can't currently cure), and we spend time squabbling over the practically unimportant question of whenever said cures "really work" or perhaps the trial got un-blinded a little bit by side effects. You have those "anti-depressants" that lower the depression score by a point or two compared to the placebo (with both them and the placebo lowering the score by well over a couple points), and they're deemed an "effective treatment", and there's great deal of argument as to whenever they actually do anything, or is it an un-blinding artefact, or is that the weight gain getting measured as decrease in depression. But the truth is, it does not matter - the effect is incredibly weak in either case (you could compare that to antibiotics which work like 'take this for a week, 90% patients are cured, don't take it, 90% of patients are dead').
Last edited by Dmytry on Fri Dec 19, 2014 8:47 pm UTC, edited 5 times in total.

User avatar
cellocgw
Posts: 2067
Joined: Sat Jun 21, 2008 7:40 pm UTC

Re: 1462: "Blind Trials"

Postby cellocgw » Fri Dec 19, 2014 8:33 pm UTC

orthogon wrote:I may be wrong, but so far hasn't she described a single-blind trial? Somehow the researchers have got to not know whether they're administering a sugar pill or a hotel room with champagne on ice and Barry White on the stereo.


I'd rather have sugar pills and no Barry White over any kind of sex with Barry White.

Wasn't it GBShaw who said something like "The expense is damnable, the pleasure momentary,and the position ridiculous" ? (not talking about sugar pills here)
resume
Former OTTer
Vote cellocgw for President 2020. #ScienceintheWhiteHouse http://cellocgw.wordpress.com
"The Planck length is 3.81779e-33 picas." -- keithl
" Earth weighs almost exactly π milliJupiters" -- what-if #146, note 7

User avatar
Flumble
Yes Man
Posts: 2263
Joined: Sun Aug 05, 2012 9:35 pm UTC

Re: 1462: "Blind Trials"

Postby Flumble » Fri Dec 19, 2014 8:42 pm UTC

sorceror wrote:For those situations when a double-blind trial is impossible, an interesting new statistical tool that can, in the right circumstances, determine causation from correlation: https://medium.com/the-physics-arxiv-blog/cause-and-effect-the-revolutionary-new-statistical-test-that-can-tease-them-apart-ed84a988e

Obvious in retrospect, almost. Someday I hope to have an idea that's obvious in retrospect like that.

I may have gone through the article too fast, but I don't see a claim for telling correlation and causation apart —only the claim that, given a pair of variables with a causal relationship, they can calculate which is the cause and which is the effect... with a high degree of certainty.

User avatar
Dmytry
Posts: 68
Joined: Mon Jun 01, 2009 12:39 pm UTC
Contact:

Re: 1462: "Blind Trials"

Postby Dmytry » Fri Dec 19, 2014 8:56 pm UTC

Flumble wrote:
sorceror wrote:For those situations when a double-blind trial is impossible, an interesting new statistical tool that can, in the right circumstances, determine causation from correlation: https://medium.com/the-physics-arxiv-blog/cause-and-effect-the-revolutionary-new-statistical-test-that-can-tease-them-apart-ed84a988e

Obvious in retrospect, almost. Someday I hope to have an idea that's obvious in retrospect like that.

I may have gone through the article too fast, but I don't see a claim for telling correlation and causation apart —only the claim that, given a pair of variables with a causal relationship, they can calculate which is the cause and which is the effect... with a high degree of certainty.

Yeah, and for the placebo, the treatment is still the cause of the placebo effect.

In any case, I think placebo-controlled trials are massively misused. Let's say we have this ancient Chinese treatment for appendicitis, it involves rubbing in a special oil at special points. What matters is that it is extremely ineffective compared to appendectomy. It doesn't matter in the slightest if it beats placebo a little bit or not. But the way the world works, they find it beats the placebo a little bit and then you have giant headlines - this alternative appendicitis treatment really works! No it does not "work" for any non-negligible value of "work", not for most patients anyway.

rmsgrey
Posts: 3655
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1462: "Blind Trials"

Postby rmsgrey » Fri Dec 19, 2014 8:59 pm UTC

Dmytry wrote:The way I think of it, if the claimed effect is small, close in magnitude to the observed magnitude of placebo effect, it shouldn't really matter so much to anyone except those researching biochemistry if it is pure placebo effect or not. If the cure is very close to the placebo effect, it is also almost entirely ineffective.


I wouldn't say a cure comparable to placebo is "almost entirely ineffective" - the whole reason for double-blind trials in the first place is that placebo is so effective that it completely screws up your results if you don't correct for it. Something that's only "as effective as placebo" is still reasonably effective - just not necessarily in the way(s) claimed for it.

User avatar
jdgalt
Posts: 1
Joined: Fri Dec 19, 2014 9:19 pm UTC

Re: 1462: "Blind Trials"

Postby jdgalt » Fri Dec 19, 2014 9:38 pm UTC

It might actually work, if the non-control group is "having sex" by taking the pills they used in Demolition Man. Or Barbarella.

User avatar
12obin
Posts: 91
Joined: Tue Mar 26, 2013 1:15 am UTC

Re: 1462: "Blind Trials"

Postby 12obin » Fri Dec 19, 2014 10:26 pm UTC

I'd really like to think that I could tell the difference between being poked with a toothpick or a needle.
Robin, she.

User avatar
CocoaNutCakery
Posts: 66
Joined: Fri May 24, 2013 6:26 am UTC
Contact:

Re: 1462: "Blind Trials"

Postby CocoaNutCakery » Sat Dec 20, 2014 12:33 am UTC

PinkShinyRose wrote:
CocoaNutCakery wrote:Any time I bring up a study where they test a particular proceedure or the efficacy of an exercise (with one group doing a comparable exercise) and someone essentially suggests this, just...

Rage. Rage everywhere.

The really sad part is that I actually had a doctor, an actual medical doctor, that dismissed a comparative exercise study because it wasn't double blinded.

Why was that wrong? What kind of exercise? Who was not blinded for what reason? Were there other studies supporting their conclusions?


Originally, I was just going to post one of the many facepalm gifs that are out there, but you know, maybe it would help to educate you a little bit. Here is what a single-blind exercise study would look like.

Group 1 will go walking for 30 minutes each day, but the people in the group won't know that they're walking.
Group 2 will do comparable resistance band training each day, but the people in the group won't know that they're doing resistance band training.

Can you see what's wrong with that?

Angua wrote:No talk about how trials can't all be the same is not complete without:

http://www.bmj.com/content/327/7429/1459


Oh, I know SO many people that should sign up for that trial. For science.

rmsgrey
Posts: 3655
Joined: Wed Nov 16, 2011 6:35 pm UTC

Re: 1462: "Blind Trials"

Postby rmsgrey » Sat Dec 20, 2014 12:50 am UTC

jdgalt wrote:It might actually work, if the non-control group is "having sex" by taking the pills they used in Demolition Man. Or Barbarella.


Demolition Man used helmets, not pills.

User avatar
Dmytry
Posts: 68
Joined: Mon Jun 01, 2009 12:39 pm UTC
Contact:

Re: 1462: "Blind Trials"

Postby Dmytry » Sat Dec 20, 2014 1:11 am UTC

rmsgrey wrote:
Dmytry wrote:The way I think of it, if the claimed effect is small, close in magnitude to the observed magnitude of placebo effect, it shouldn't really matter so much to anyone except those researching biochemistry if it is pure placebo effect or not. If the cure is very close to the placebo effect, it is also almost entirely ineffective.


I wouldn't say a cure comparable to placebo is "almost entirely ineffective" - the whole reason for double-blind trials in the first place is that placebo is so effective that it completely screws up your results if you don't correct for it. Something that's only "as effective as placebo" is still reasonably effective - just not necessarily in the way(s) claimed for it.

Well, placebo effect is small to non-existent for any objectively measurable quantities, so it's probably a combination of anti-complaining effect on the patient and anti-listening effect on the doctor.

edit: it's interesting to contrast the subjectively measured pain, with subjectively measured depression. For pain, extremely effective painkillers exist in nature and were identified correctly even by primitive tribes, because the effect is very strong. They definitely work when the patient needs to sit still when you're cutting their leg off, which is probably as objective as it gets for something as subjective as pain. For depression, "antidepressants" seem not to lower, or even raise the least biased measure of depression: suicide rate.

edit2: and interestingly, antidepressants are almost indistinguishable from placebo (a case can be very well made that they are a placebo and the minor effects are due to un-blinding by side effects or perhaps some small effect that weight gain has on reporting - something as simple as the patient being concerned with the weight gain and talking about it with the doctor, leaving less time for discussion of the depression, can decrease the reported depression symptoms). To wrap it up, my point is, yes, it is true that when you're treating something with a placebo, you do get decreases in recorded symptoms. But it has not been conclusively demonstrated that this is indicative of any medical improvement or is even harmless. As far as we know it is entirely possible that the placebos affect only reporting and recording of the symptoms, but have no effect on well being of the patient, or even cause harm to the patient; at least for the depression, that is weakly supported by available data on suicides.
Last edited by Dmytry on Sat Dec 20, 2014 3:09 am UTC, edited 4 times in total.

User avatar
PinkShinyRose
Posts: 835
Joined: Mon Nov 05, 2012 6:54 pm UTC
Location: the Netherlands

Re: 1462: "Blind Trials"

Postby PinkShinyRose » Sat Dec 20, 2014 1:43 am UTC

CocoaNutCakery wrote:
PinkShinyRose wrote:
CocoaNutCakery wrote:Any time I bring up a study where they test a particular proceedure or the efficacy of an exercise (with one group doing a comparable exercise) and someone essentially suggests this, just...

Rage. Rage everywhere.

The really sad part is that I actually had a doctor, an actual medical doctor, that dismissed a comparative exercise study because it wasn't double blinded.

Why was that wrong? What kind of exercise? Who was not blinded for what reason? Were there other studies supporting their conclusions?


Originally, I was just going to post one of the many facepalm gifs that are out there, but you know, maybe it would help to educate you a little bit. Here is what a single-blind exercise study would look like.

Group 1 will go walking for 30 minutes each day, but the people in the group won't know that they're walking.
Group 2 will do comparable resistance band training each day, but the people in the group won't know that they're doing resistance band training.

Can you see what's wrong with that?

That makes sense, I thought it might be exercise with resistance band versus not-so-much-but-still-some-resistance band. Or some exercise for a more complicated issue where moving differently would suffice for making it ineffective (because it doesn't train the relevant muscle).

To be fair: at least here medical education doesn't involve developing academic skills very thoroughly.

User avatar
CocoaNutCakery
Posts: 66
Joined: Fri May 24, 2013 6:26 am UTC
Contact:

Re: 1462: "Blind Trials"

Postby CocoaNutCakery » Sat Dec 20, 2014 2:13 am UTC

PinkShinyRose wrote:
CocoaNutCakery wrote:Originally, I was just going to post one of the many facepalm gifs that are out there, but you know, maybe it would help to educate you a little bit. Here is what a single-blind exercise study would look like.

Group 1 will go walking for 30 minutes each day, but the people in the group won't know that they're walking.
Group 2 will do comparable resistance band training each day, but the people in the group won't know that they're doing resistance band training.

Can you see what's wrong with that?

That makes sense, I thought it might be exercise with resistance band versus not-so-much-but-still-some-resistance band. Or some exercise for a more complicated issue where moving differently would suffice for making it ineffective (because it doesn't train the relevant muscle).

To be fair: at least here medical education doesn't involve developing academic skills very thoroughly.


Image

And the award for completely missing the point goes to...

I've bolded the issue. Can you see it now?

User avatar
jc
Posts: 356
Joined: Fri May 04, 2007 5:48 pm UTC
Location: Waltham, Massachusetts, USA, Earth, Solar System, Milky Way Galaxy
Contact:

Re: 1462: "Blind Trials"

Postby jc » Sat Dec 20, 2014 2:39 am UTC

Angua wrote:No talk about how trials can't all be the same is not complete without:
http://www.bmj.com/content/327/7429/1459


That's fun to read in an actual medical journal. A somewhat less sarcastic reply that I've run across occasionally is a more general comment on the idea that only properly controlled experimental tests can be considered "scientific". If this were true, we'd have to conclude that the entire field of astronomy is a pseudo-science, since astronomers obviously perform very few experimental tests on their subject matter. But in fact, astronomy is considered by all scientists to be one of the "hardest" of the "hard sciences". It's not actually all that hard to understand how purely observational data can be the basis of good scientific theories. The only problem is with people thinking that there can only be one method of performing "science".

OTOH, I do really like the idea of trying to organize a properly-controlled double-blind study of parachutes vs. placebos, with the critics of the use of uncontrolled observational data as the participants in the study.

User avatar
da Doctah
Posts: 995
Joined: Fri Feb 03, 2012 6:27 am UTC

Re: 1462: "Blind Trials"

Postby da Doctah » Sat Dec 20, 2014 3:56 am UTC

12obin wrote:I'd really like to think that I could tell the difference between being poked with a toothpick or a needle.

Wait, are we still talking about the sex thing?

User avatar
azule
Saved
Posts: 2132
Joined: Mon Jul 26, 2010 9:45 pm UTC
Location: The land of the Golden Puppies and Rainbows

Re: 1462: "Blind Trials"

Postby azule » Sat Dec 20, 2014 4:33 am UTC

Micropenis.





That is all.
Image

If you read this sig, post about one arbitrary thing you did today.

I celebrate up to six arbitrary things before breakfast.
Time does drag on and on and contain spoilers. Be aware of memes.

RogueCynic
Posts: 409
Joined: Sun Nov 22, 2009 10:23 pm UTC

Re: 1462: "Blind Trials"

Postby RogueCynic » Sat Dec 20, 2014 5:07 am UTC

Wasn't a study done some years ago that convinced some rats they were having sex when they weren't? And how does masturbation fit in with this study?
I am Lord Titanius Englesmith, Fancyman of Cornwood.
See 1 Kings 7:23 for pi.
If you put a prune in a juicer, what would you get?

User avatar
Neil_Boekend
Posts: 3220
Joined: Fri Mar 01, 2013 6:35 am UTC
Location: Yes.

Re: 1462: "Blind Trials"

Postby Neil_Boekend » Sat Dec 20, 2014 9:58 am UTC

Jc, one glorious day we will figure out how to do double blind astronomy research. The next day the military will steal the supernova machine.
Mikeski wrote:A "What If" update is never late. Nor is it early. It is posted precisely when it should be.

patzer's signature wrote:
flicky1991 wrote:I'm being quoted too much!

he/him/his


Return to “Individual XKCD Comic Threads”

Who is online

Users browsing this forum: Google Feedfetcher, hetas, mscha and 96 guests