Welcome to City-Data.com Forum!
U.S. CitiesCity-Data Forum Index
Go Back   City-Data Forum > General Forums > Health and Wellness > Daily Journals
 [Register]
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
View detailed profile (Advanced) or search
site with Google Custom Search

Search Forums  (Advanced)
Reply Start New Thread
 
Old 12-11-2017, 09:12 AM
 
Location: near bears but at least no snakes
26,633 posts, read 28,419,191 times
Reputation: 50424

Advertisements

Quote:
Originally Posted by guidoLaMoto View Post
When reading scientific studies, a couple things should be kept in mind:

"Peer review" depends on who is doing the reviewing-- a reviewer solidly entrenched in the "consensus" is having his reputation invalidated if he allows a paper to be published that goes against the consensus. How altruistic do you expect him to be?

It's not bad that drug companies pay for research and then publish the studies that help them out. The problem comes when they don't publish a study that doesn't support their product.

Beware of papers that report "Relative Risk Reduction"-- that's a way of cheating to make a single digit result look like a double digit (more impressive) result. Check the Absolute Risk Reduction....NEJM just reported a study that made the headlines about risk of breast cancer in those using BCP. They reported a "38%" increase in risk. It turns out that they found 58 Ca cases for every 100,000 woman-yrs of BCP compared to only 55/100,000 for non-users. BFD. There's a difference between statistical significance and clinical (ie- practical) significance.

Caveat lector.

If you torture the numbers long enough, you can get them to confess to anything.
Excellent. Thank you. We need to be aware, open minded, and yet skeptical of anything we read: studies, advertisements, articles, etc.

(Reminds me of a temp job I had a long time ago. It was a door to door survey and, with the questions they asked, it was guaranteed to produce the desired results. Sure enough, it "proved" just what they wanted to prove and they got the grant money they were after. I liked the results BUT it was maddening to know that the respondents (yes, they were randomly chosen) weren't even given a fair choice of answers, there was no place for "does not apply", and the entire survey would have been discarded if all questions hadn't been answered.)
Reply With Quote Quick reply to this message

 
Old 12-11-2017, 10:41 AM
 
Location: SW Florida
14,798 posts, read 11,938,114 times
Reputation: 24499
Quote:
Originally Posted by guidoLaMoto View Post
When reading scientific studies, a couple things should be kept in mind:

"Peer review" depends on who is doing the reviewing-- a reviewer solidly entrenched in the "consensus" is having his reputation invalidated if he allows a paper to be published that goes against the consensus. How altruistic do you expect him to be?

It's not bad that drug companies pay for research and then publish the studies that help them out. The problem comes when they don't publish a study that doesn't support their product.

Beware of papers that report "Relative Risk Reduction"-- that's a way of cheating to make a single digit result look like a double digit (more impressive) result. Check the Absolute Risk Reduction....NEJM just reported a study that made the headlines about risk of breast cancer in those using BCP. They reported a "38%" increase in risk. It turns out that they found 58 Ca cases for every 100,000 woman-yrs of BCP compared to only 55/100,000 for non-users. BFD. There's a difference between statistical significance and clinical (ie- practical) significance.

Caveat lector.

If you torture the numbers long enough, you can get them to confess to anything.
ROFL, I have never heard it put that way, but what a great description!

Massaging those numbers to get them to come out to fit the pre-determined conclusion. One of the federal government's favorite tricks ( which I have seen first hand as a participant in some of their studies), is their exclusion of selective data which does not fit the foregone desired conclusion. As in when a participant inputs findings into the database, there is no place, or any way to input negative results from the studies. Therefore, no negative findings, and the positive result is "proved".

I'd be surprised if this practice was limited to the guv'mint, but that is where I saw it. Amazing how many ways there are to lie, or mislead, using statistics.
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 11:04 AM
 
Location: Southern California
29,267 posts, read 16,551,673 times
Reputation: 18901
Quote:
Originally Posted by guidoLaMoto View Post
When reading scientific studies, a couple things should be kept in mind:

"Peer review" depends on who is doing the reviewing-- a reviewer solidly entrenched in the "consensus" is having his reputation invalidated if he allows a paper to be published that goes against the consensus. How altruistic do you expect him to be?

It's not bad that drug companies pay for research and then publish the studies that help them out. The problem comes when they don't publish a study that doesn't support their product.

Beware of papers that report "Relative Risk Reduction"-- that's a way of cheating to make a single digit result look like a double digit (more impressive) result. Check the Absolute Risk Reduction....NEJM just reported a study that made the headlines about risk of breast cancer in those using BCP. They reported a "38%" increase in risk. It turns out that they found 58 Ca cases for every 100,000 woman-yrs of BCP compared to only 55/100,000 for non-users. BFD. There's a difference between statistical significance and clinical (ie- practical) significance.

Caveat lector.

If you torture the numbers long enough, you can get them to confess to anything.
Of course, this is true, it depends who is sponsoring the research, papers, studies etc...who is getting paid for them. I've been saying this since I arrived at C-D.
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 02:06 PM
 
Location: The Driftless Area, WI
7,112 posts, read 4,950,204 times
Reputation: 17437
Quote:
Originally Posted by Travelassie View Post
ROFL, I have never heard it put that way, but what a great description!

Massaging those numbers to get them to come out to fit the pre-determined conclusion. ....
I'd be surprised if this practice was limited to the guv'mint, but that is where I saw it.
The problem has gotten really bad in the last couple decades.

Look at the hoaxes of cholesterol, cancer from eating meat, global warming, GHGs from raising livestock, particulate matter in the air, and second hand smoke, just to name a few obvious ones.
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 02:11 PM
 
Location: on the wind
22,807 posts, read 18,101,386 times
Reputation: 73971
Quote:
Originally Posted by Travelassie View Post
ROFL, I have never heard it put that way, but what a great description!

Massaging those numbers to get them to come out to fit the pre-determined conclusion. One of the federal government's favorite tricks ( which I have seen first hand as a participant in some of their studies), is their exclusion of selective data which does not fit the foregone desired conclusion. As in when a participant inputs findings into the database, there is no place, or any way to input negative results from the studies. Therefore, no negative findings, and the positive result is "proved".

I'd be surprised if this practice was limited to the guv'mint, but that is where I saw it. Amazing how many ways there are to lie, or mislead, using statistics.
Yep, there are entire books devoted to "how to lie with statistics". Just remember that ALL sides of an argument can do the same. That's why I try to read a wide variety of sources when researching something. Eventually a common theme of results does appear and you can isolate the odd ones that stick out too far. The other problem with searching studies is the background understanding of the reader...if they don't have basic biology, chemistry, physics or other disciplines to draw on they can completely misunderstand what they are reading no matter what it tries to tell them.
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 02:16 PM
 
Location: The Driftless Area, WI
7,112 posts, read 4,950,204 times
Reputation: 17437
Quote:
Originally Posted by in_newengland View Post
Excellent. Thank you. We need to be aware, open minded, and yet skeptical of anything we read: studies, advertisements, articles, etc.

(Reminds me of a temp job I had a long time ago. It was a door to door survey and, with the questions they asked, it was guaranteed to produce the desired results. Sure enough, it "proved" just what they wanted to prove and they got the grant money they were after. I liked the results BUT it was maddening to know that the respondents (yes, they were randomly chosen) weren't even given a fair choice of answers, there was no place for "does not apply", and the entire survey would have been discarded if all questions hadn't been answered.)
Surveys are a particular problem, and most "nutritional studies," due to the ethics concerns of human experiments, are now mostly of this type. Respondents often try to please the researchers, giving them answers they think they want to hear.

A few years ago there was a TV College course on Statistics. They had a hidden camera set up to watch a fictitious agent doing a survey, stopping shoppers at a mall. It was amazing how many respondents they found who responded "yes" to the question "Should the US govt send foreign aid to the rebels on Alpha Centauri?"
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 02:29 PM
 
Location: near bears but at least no snakes
26,633 posts, read 28,419,191 times
Reputation: 50424
Quote:
Originally Posted by guidoLaMoto View Post
Surveys are a particular problem, and most "nutritional studies," due to the ethics concerns of human experiments, are now mostly of this type. Respondents often try to please the researchers, giving them answers they think they want to hear.

A few years ago there was a TV College course on Statistics. They had a hidden camera set up to watch a fictitious agent doing a survey, stopping shoppers at a mall. It was amazing how many respondents they found who responded "yes" to the question "Should the US govt send foreign aid to the rebels on Alpha Centauri?"
Lol. I can well imagine! The survey I was involved with was so long ago that we actually went door to door and sat down and interviewed the person for twenty minutes. I tried to be impartial in that I didn't show any emotion or act as though I was pleased or displeased with any answers. The problem was the survey itself--it was a survey on unemployment and I had to ask 80 year old people if they had a job. Of course they were not employed. But I had to enter that as a "no" they were "unemployed."

"How long since you have had a job?" There was no place to say that they were retired so again, it looked as though they were unemployed and couldn't find a job. The end result was that unemployment was REALLY HIGH in our area. Our area got a state grant to help people regarding jobs and it was obvious that this was the intention of the survey. This study was conducted by a university and I have never trusted statistics since then.
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 02:36 PM
bg7
 
7,694 posts, read 10,495,482 times
Reputation: 15298
Quote:
Originally Posted by newtovenice View Post
The website link works just fine.

Amazing people refuse to look at info just because they don't know the website. That thinking perpetuates ignorance.

Marcia Angell, former EIC of the New England Journal of Medicine, has spoken at length and written at least one book about the handshake wink wink relationship between the drug companies and the *respected* medical journals.

If you wish to learn about the subject perhaps Dr. Angell's work will be more to the liking of those who don't want to clink on the link the OP posted. Every doctor who publishes is paid in some way by the pharma company who creates the drug the doctor is supporting. This has been proved again and again.

https://www.amazon.com/Truth-About-D.../dp/0375760946

The Truth About the Drug Companies | by Marcia Angell | The New York Review of Books


https://www.youtube.com/watch?v=uDbQNBla6aU

https://www.wanttoknow.info/truthaboutdrugcompanies

These are the 1st four links that popped up. So yeah.


Except that a huge number of doctor publications in peer-reviewed journals don't regard drugs at all. They are medical observations, different types of non-drug interventions, and applied scientific research. So what's their iniquitous purpose? The evil advancement of knowledge? I mean according to the usual mindset here I didn't think MDs got out of bed unless Pfizer was paying for their egg roll and Merck was fueling their Ferrari.


Plus get this - science is falsifiable! Yea - its true - another doc. in the world can do the same experiment and see if it works - or even do a different experiment proving it wrong! You can improve knowledge all the time! (c.f. the charlatans duping the "alternatives" where every single thing is hearsay and specious theorizing - no accountability! Its a great business to be in).
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 02:47 PM
bg7
 
7,694 posts, read 10,495,482 times
Reputation: 15298
Quote:
Originally Posted by guidoLaMoto View Post
The problem has gotten really bad in the last couple decades.

Look at the hoaxes of cholesterol, cancer from eating meat, global warming, GHGs from raising livestock, particulate matter in the air, and second hand smoke, just to name a few obvious ones.


For centuries of deforestation, car pollution, and factory pollution not to have effected change in the global climate would be (absent divine intervention or a magical space vacuum removing all pollution up to the Kuiper belt), absolutely impossible. The extent to which climate change is anthropomorphic is one debate for sure, but, short of lying to yourself, of course human activities have affected climate.


Second hand smoke studies - validated in pets of heavy smokers.


Cancer form eating meat? I'm a meat eater but pretending epidemiological data doesn't show a higher incidence of certain cancers in processed or non-red meat eaters.....That's just putting hands over your ears.


Sure you can argue causation or correlation all you want but most will draw the line
Reply With Quote Quick reply to this message
 
Old 12-11-2017, 02:55 PM
bg7
 
7,694 posts, read 10,495,482 times
Reputation: 15298
Quote:
Originally Posted by guidoLaMoto View Post
When reading scientific studies, a couple things should be kept in mind:

"Peer review" depends on who is doing the reviewing-- a reviewer solidly entrenched in the "consensus" is having his reputation invalidated if he allows a paper to be published that goes against the consensus. How altruistic do you expect him to be?

It's not bad that drug companies pay for research and then publish the studies that help them out. The problem comes when they don't publish a study that doesn't support their product.

Beware of papers that report "Relative Risk Reduction"-- that's a way of cheating to make a single digit result look like a double digit (more impressive) result. Check the Absolute Risk Reduction....NEJM just reported a study that made the headlines about risk of breast cancer in those using BCP. They reported a "38%" increase in risk. It turns out that they found 58 Ca cases for every 100,000 woman-yrs of BCP compared to only 55/100,000 for non-users. BFD. There's a difference between statistical significance and clinical (ie- practical) significance.

Caveat lector.

If you torture the numbers long enough, you can get them to confess to anything.


That's was statistically significant?
At what level - the p was 0.05?


But that's why we have stats - so we can be reasonably sure its not just by chance. It surely won't mean much to the physician seeing a dozen cancer patients everyday. It surely would make a difference as knowledge to a woman who has tested genetically predisposed to certain cancer types and is trying to reduce all reasonably foreseeable chances.


Its one thing to see a publish-or-die situation meaning researchers making a paper out of a small difference. But then again its quite another to have 55 vs 58 deaths per 100,000 plane miles, statistically significant, on the airline you are choosing. One kills 580 people per million miles, the other 550 - a whole crash less. Means the aircraft maintenance program of the former clearly needs attending to. I mean - which airline would you fly on knowing that?
Reply With Quote Quick reply to this message
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.

Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.


Reply
Please update this thread with any new information or opinions. This open thread is still read by thousands of people, so we encourage all additional points of view.

Quick Reply
Message:


Over $104,000 in prizes was already given out to active posters on our forum and additional giveaways are planned!

Go Back   City-Data Forum > General Forums > Health and Wellness > Daily Journals

All times are GMT -6.

© 2005-2024, Advameg, Inc. · Please obey Forum Rules · Terms of Use and Privacy Policy · Bug Bounty

City-Data.com - Contact Us - Archive 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37 - Top