Author Topic: So, it turns out there are a lot of people that just don't like science....  (Read 257 times)

0 Members and 1 Guest are viewing this topic.

Online Ah.hell

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 13256
At least when practiced on humans anyway. 

The latest episode of "you are not so smart" interviewed a couple of scientist who have conducted research on how folks respond to controlled studies and there seems to be something like 30% of the population that just doesn't like the notion of running controlled experiments on people. 

https://youarenotsosmart.com/podcast/

Some of the examples.
Some doctor's provide treatment A to emergency room patients while other doctors use treatment B.  It is not clear whether either treatment works better than the other. When asked how the felt about the following:

1.  What doctor is around when you arrive in the emergency room is random so you will be randomly assigned to treatment A or Treatment B. 
2.  You will be randomly assigned to a Doctor who will randomly assign you to A or B. 

Almost nobody disliked Option 1 but something around 30% disliked option 2. 

Another example regarding checklists.
A.  A checklist of things to check on patient visits would be printed on the back of an attending physicians name tag.
B.  The same checklist would be printed on poster in the hospital room.
C.  Both.
D.  A/B/C would be randomly assigned to different hospitals then outcomes would be measured and the method with the best outcome would be selected for future use.

Almost nobody objected to A, B, and C but something like 30% of people objected to D.

There did not seem to be a correlation with education or even folks with science degrees. 

Please note, I'm going from memory, so some of the details will be a little off.  Also, the latest episode, 158 isn't up on the website yet. 


Offline jt512

  • Frequent Poster
  • ******
  • Posts: 2491
    • jt512
At least when practiced on humans anyway. 

The latest episode of "you are not so smart" interviewed a couple of scientist who have conducted research on how folks respond to controlled studies and there seems to be something like 30% of the population that just doesn't like the notion of running controlled experiments on people. 

https://youarenotsosmart.com/podcast/

Some of the examples.
Some doctor's provide treatment A to emergency room patients while other doctors use treatment B.  It is not clear whether either treatment works better than the other. When asked how the felt about the following:

1.  What doctor is around when you arrive in the emergency room is random so you will be randomly assigned to treatment A or Treatment B. 
2.  You will be randomly assigned to a Doctor who will randomly assign you to A or B. 

Almost nobody disliked Option 1 but something around 30% disliked option 2.


Neither #1 nor #2, as stated, is a controlled experiment.  #1 is basically a fact of life, so there is no point to objecting to it.  #2 is objectionable, because doctors should not be using randomization to allocate treatments, except in the course of a randomized controlled trial that has been deemed ethical and to which you have given informed consent.


Quote
Another example regarding checklists.
A.  A checklist of things to check on patient visits would be printed on the back of an attending physicians name tag.
B.  The same checklist would be printed on poster in the hospital room.
C.  Both.
D.  A/B/C would be randomly assigned to different hospitals then outcomes would be measured and the method with the best outcome would be selected for future use.

Almost nobody objected to A, B, and C but something like 30% of people objected to D.

There did not seem to be a correlation with education or even folks with science degrees.


If there is legitimate uncertainty over which of A, B, or C is most effective, then, personally, I would have no problem with this.  Whether such an experiment, lacking informed consent of patients, would actually be deemed ethical is another matter.
Ich verstehe nur Bahnhof.

Offline random poet

  • That's bullshit!
  • Frequent Poster
  • ******
  • Posts: 2082
  • On n'a jamais le temps, le temps nous a.
    • I have a LJ
At least when practiced on humans anyway. 

The latest episode of "you are not so smart" interviewed a couple of scientist who have conducted research on how folks respond to controlled studies and there seems to be something like 30% of the population that just doesn't like the notion of running controlled experiments on people. 

https://youarenotsosmart.com/podcast/

Some of the examples.
Some doctor's provide treatment A to emergency room patients while other doctors use treatment B.  It is not clear whether either treatment works better than the other. When asked how the felt about the following:

1.  What doctor is around when you arrive in the emergency room is random so you will be randomly assigned to treatment A or Treatment B. 
2.  You will be randomly assigned to a Doctor who will randomly assign you to A or B. 

Almost nobody disliked Option 1 but something around 30% disliked option 2.


Neither #1 nor #2, as stated, is a controlled experiment.  #1 is basically a fact of life, so there is no point to objecting to it.  #2 is objectionable, because doctors should not be using randomization to allocate treatments, except in the course of a randomized controlled trial that has been deemed ethical and to which you have given informed consent.


Quote
Another example regarding checklists.
A.  A checklist of things to check on patient visits would be printed on the back of an attending physicians name tag.
B.  The same checklist would be printed on poster in the hospital room.
C.  Both.
D.  A/B/C would be randomly assigned to different hospitals then outcomes would be measured and the method with the best outcome would be selected for future use.

Almost nobody objected to A, B, and C but something like 30% of people objected to D.

There did not seem to be a correlation with education or even folks with science degrees.


If there is legitimate uncertainty over which of A, B, or C is most effective, then, personally, I would have no problem with this.  Whether such an experiment, lacking informed consent of patients, would actually be deemed ethical is another matter.
You got it! That's why it was bothering me. There is no way an IRB is going to approve this. No informed consent.
Aujourd'hui j'ai vu un facteur joyeux.

Online Ah.hell

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 13256
I strongly recommend listening to the podcast and or reading the show notes.   Those were just examples of the sort of questions they used and just my memory of them at that.   They've framed it in numerous ways to try and figure out what folks are objecting to.  Lack of informed consent seems to be part of it but not all of it.

Regarding the first example.  Sure, situation 1 is just a fact of life but situation 2 doesn't change the circumstances in real way except it would actually improve outcomes if followed through with.


Offline jt512

  • Frequent Poster
  • ******
  • Posts: 2491
    • jt512

...situation 2 doesn't change the circumstances in real way except it would actually improve outcomes if followed through with.


Says who?
Ich verstehe nur Bahnhof.

Online Ah.hell

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 13256

...situation 2 doesn't change the circumstances in real way except it would actually improve outcomes if followed through with.
Says who?
The clear framing of the questions.

Condition one, random assortment of treatment.
Condition two, random assortment of treatment used to determine which treatment was more effective.

Offline jt512

  • Frequent Poster
  • ******
  • Posts: 2491
    • jt512

...situation 2 doesn't change the circumstances in real way except it would actually improve outcomes if followed through with.
Says who?
The clear framing of the questions.

Condition one, random assortment of treatment.
Condition two, random assortment of treatment used to determine which treatment was more effective.


You are asking whether a doctor, acting on his own, should be able to randomize patients to treatments to determine which is more effective.  In the real world, the answer is absolutely not, for at least the following reasons.


1. An individual doctor should not be permitted to decide whether there is legitimate scientific uncertainty between the risks and benefits of the two treatments to justify a trial.


2. An individual doctor cannot be trusted to randomize patients in a statistically valid manner.


3. An individual doctor does not have the expertise or the resources to compute the required sample size for the trial, monitor the trial for safety, determine when the trial should be terminated, and analyze the results.


4. How would the doctor promulgate the trial results to the profession?  What journal would publish the results of a "trial" that had no IRB approval, no informed consent, no statistician, no safety committee, and no oversight?  Even if the results managed to get published, who would trust them?


5. Patients have the right to know that they are participating in a trial, and that some established medical body has deemed the trial medically important, that the trial will not put them at undue risk, that the trial has responsible oversight, and that the results of the trial will improve future medical care.  None of this would be possible by a doctor acting on his own.


I see no evidence that people objecting to the scenario you outlined are objecting to science per se, but rather are objecting to being subjected to sloppy, unethical science.
Ich verstehe nur Bahnhof.

Online Ah.hell

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 13256
Listen to the podcast, I'm clearly not doing a good job of explaining the research in question.  I should have known not to provide an example.  The point was and is, as framed in various ways, sometimes but not always involving doctors, lots of folks seem to object to running experiments that involve people for reasons that are not entirely clear.  Some of this effect may well be objections to experimenting with out informed consent but that is not actually clear.

That is not the only way the questions were framed, it was merely a memorable example. 

Offline jt512

  • Frequent Poster
  • ******
  • Posts: 2491
    • jt512
Listen to the podcast.

Sorry.  It's not my job to make up for your inadequate summary of the podcast by listening to it.  It's your job to adequately summarize it.

Ich verstehe nur Bahnhof.

Online Ah.hell

  • Poster of Extraordinary Magnitude
  • **********
  • Posts: 13256
Listen to the podcast.

Sorry.  It's not my job to make up for your inadequate summary of the podcast by listening to it.  It's your job to adequately summarize it.
Quote
At least when practiced on humans anyway. 

The latest episode of "you are not so smart" interviewed a couple of scientist who have conducted research on how folks respond to controlled studies and there seems to be something like 30% of the population that just doesn't like the notion of running controlled experiments on people. 
« Last Edit: July 15, 2019, 05:45:56 PM by Ah.hell »

 

personate-rain