Get those Senior members out of the Corportae uniform and into the USAF style

[Eclipse]

Google is fixing the symptom search issue, just like it has fixed everything else...

http://www.usatoday.com/story/tech/news/2016/06/20/google-health-symptom-medical-conditions/86061086/

[Pace]

Funny thing about evidence-based medicine, the statisticians recently released a unified position statement that the healthcare industry has been using P-value incorrect / inappropriately to justify using some data over other data for years. This basically means that a significant majority of "evidence based practice" is potentially wrong. [heads exploding] [/drops mic and walks away...]

[THRAWN]

Meh. It works for me. If some egghead in a labcoat disagrees he can go tune up his sliderule...

[LSThiker]

Funny thing about evidence-based medicine, the statisticians recently released a unified position statement that the healthcare industry has been using P-value incorrect / inappropriately to justify using some data over other data for years. This basically means that a significant majority of "evidence based practice" is potentially wrong. [heads exploding] [/drops mic and walks away...]

This is not new.  It has been known for quite some time.  Journals will occasionally hire statisticians to go back through the old articles and examine them.  I believe (as I cannot find the article right now) that Journal of Virology had something like 50-60% of published articles used the wrong statistical test.  That is, they should have used a ANOVA instead of a t-test or used an ANOVA but did not include a post-hoc such as Tukey or used one-way instead of two-way:

http://www.nature.com/news/scientific-method-statistical-errors-1.14700
http://www.sciencedirect.com/science/article/pii/S0887617705001071
http://imedea.uib-csic.es/bc/ecopob/docs/pdfsgrupo/articulos/2008/14.Acta%20Oecol%20alex(2008).pdf


Reported:
https://www.sciencenews.org/article/odds-are-its-wrong
https://www.sciencenews.org/article/research-cant-be-right-statistics-done-wrong
http://nautil.us/issue/4/the-unlikely/sciences-significant-stats-problem

The thing about science though is that it is self-correcting (albeit it may take years or decades).  That is why science is more than a body of knowledge, but rather a way of skeptically examining the world.  This is why I always make the difference between a scientist and a person that has study science.  Anyone can get a degree in the sciences, but that does not necessarily make them a scientist. 

Unfortunately in today's world, science has taken a hit due to the lack of money.  The funding situation in this country is poor with the NIH having an ~15% fund rate (the lowest it has been).  Essentially, you are more likely to get cancer than get a grant to cure cancer. 
 

[Pace]

I am aware that clinicians poorly use their limited knowledge of statistics to develop their research. I was, however, quite impressed that the statisticians' national association released a statement that effectively said "you're doing it wrong, your results don't mean what you think they mean, knock it off."

[LSThiker]

I am aware that clinicians poorly use their limited knowledge of statistics to develop their research. I was, however, quite impressed that the statisticians' national association released a statement that effectively said "you're doing it wrong, your results don't mean what you think they mean, knock it off."

It is not just clinicians. It includes scientists, medical scientists, and clinicians.

Like I said this not new and these statements have been published before.

The reason why they are not getting caught is due to funding. Scientists have less time as they need to constantly write new grants and papers as that is the only real thing that matters in the world of academic research. While community service is required by most schools, it does not get the PI any credit. The wonderful you need to do it but I wont give you credit for doing it. So the peer review process has taken a hit as scientists take less time to review publications. In the past, a lot of these mistakes were caught. Today not so much. The perfect example is the Wakefield paper on autism which had enough stat errors that a grad student could easily catch it. Unfortunately it was not caught and we are living up to this mess

I had arguments with my advisor about SD vs SEM. SEM bars always look smaller than SD by nature so he always wanted us to publish with SEM bars. But SEM bars are not always appreciate.  As a student I just gave up arguing with him.  Now, I try not to publish with SEM unless it is necessary.

[Майор Хаткевич]

https://www.youtube.com/watch?v=0Rnq1NpHdmw

[LSThiker]

https://www.youtube.com/watch?v=0Rnq1NpHdmw

Watched it.  Mostly spot on. 

I always hate the "break through" or "science controversy" science articles from the media.  They make me laugh.

[supertigerCH]

autism?  statisticians?

huh?


what's this thread about again?  i'm lost...

[RNOfficer]

Google is fixing the symptom search issue, just like it has fixed everything else...

http://www.usatoday.com/story/tech/news/2016/06/20/google-health-symptom-medical-conditions/86061086/

Note that the new Google "symptom checker" runs only on iPhone and Androids and NOT on desktops or laptops. The alternative I recommend is

http://www.mayoclinic.org/symptom-checker/select-symptom/itt-20009075

[RNOfficer]

A medical explanation for why it's so hard to loose weight

http://www.kevinmd.com/blog/2016/07/obesity-changes-way-endocrine-system-functions.html

Obesity changes the way our endocrine system functions and how hormones communicate back to the brain. One hormone important in regulating energy and caloric intake is named leptin. Leptin is a hormone secreted by fat cells and acts on the brain to regulate eating and energy balance. Both fasting or exercise decrease leptin levels, while obesity increase leptin. With prolonged increased leptin production (from fat cells) a state of leptin resistance develops.

In short, leptin can no longer communicate to the brain, "Hey, I'm full, stop eating." The brain becomes accustomed to this higher, constant leptin level seen with obesity, and when an obese person does lose weight, leptin levels decrease and the brain interprets these as abnormally low (compared to the "new normal"). The brain thinks the obese person on a diet is starving, and so a new demand from the brain for more calories is signaled ("I'm hungry"). This is working against efforts to lose weight. Frustrating!

Of course there are other reasons such as emotional eating, which is common.