Significant or not?
Post Reply
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
09-03-2016, 03:36 PM
Significant or not?
Statistics are a great tool, as long as people use them CORRECTLY. All too often people don't, even in science.

I come across this too often in journal articles, where scientists don't understand what hypothesis testing, statistically speaking. Nor do a fair number of them understand the significance if the test produces statistically significant or insignificant results.

Now, the American Statistical Association (ASA) has found it necessary to remind scientists of this.

http://www.nature.com/news/statisticians...NatureNews

Some issues with significance tests have been so misused or misapplied, that researchers (and journals) are abandoning it.
https://www.sciencebasedmedicine.org/psy...e-testing/


Statistics are DESCRIPTIONS of data. They do not predict anything. Ever. P-values (and other significance tests) can only tell you if your results are mathematically indistinguishable from your null hypothesis. They do not prove your null hypothesis, nor disprove it....dammit Angry

Being nice is something stupid people do to hedge their bets
-Rick
Find all posts by this user
Like Post Quote this message in a reply
[+] 2 users Like TheBeardedDude's post
09-03-2016, 03:43 PM
RE: Significant or not?
(09-03-2016 03:36 PM)TheBeardedDude Wrote:  Statistics are a great tool, as long as people use them CORRECTLY. All too often people don't, even in science.

I come across this too often in journal articles, where scientists don't understand what hypothesis testing, statistically speaking. Nor do a fair number of them understand the significance if the test produces statistically significant or insignificant results.

Now, the American Statistical Association (ASA) has found it necessary to remind scientists of this.

http://www.nature.com/news/statisticians...NatureNews

Some issues with significance tests have been so misused or misapplied, that researchers (and journals) are abandoning it.
https://www.sciencebasedmedicine.org/psy...e-testing/


Statistics are DESCRIPTIONS of data. They do not predict anything. Ever. P-values (and other significance tests) can only tell you if your results are mathematically indistinguishable from your null hypothesis. They do not prove your null hypothesis, nor disprove it....dammit Angry
But 87.4 per cent of all statistics are made up on the spot!
Find all posts by this user
Like Post Quote this message in a reply
[+] 2 users Like DerFish's post
09-03-2016, 03:44 PM
RE: Significant or not?
(09-03-2016 03:43 PM)DerFish Wrote:  
(09-03-2016 03:36 PM)TheBeardedDude Wrote:  Statistics are a great tool, as long as people use them CORRECTLY. All too often people don't, even in science.

I come across this too often in journal articles, where scientists don't understand what hypothesis testing, statistically speaking. Nor do a fair number of them understand the significance if the test produces statistically significant or insignificant results.

Now, the American Statistical Association (ASA) has found it necessary to remind scientists of this.

http://www.nature.com/news/statisticians...NatureNews

Some issues with significance tests have been so misused or misapplied, that researchers (and journals) are abandoning it.
https://www.sciencebasedmedicine.org/psy...e-testing/


Statistics are DESCRIPTIONS of data. They do not predict anything. Ever. P-values (and other significance tests) can only tell you if your results are mathematically indistinguishable from your null hypothesis. They do not prove your null hypothesis, nor disprove it....dammit Angry
But 87.4 per cent of all statistics are made up on the spot!

r^2 = 0.89

Bowing

Being nice is something stupid people do to hedge their bets
-Rick
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes TheBeardedDude's post
09-03-2016, 03:57 PM
RE: Significant or not?
(09-03-2016 03:44 PM)TheBeardedDude Wrote:  
(09-03-2016 03:43 PM)DerFish Wrote:  But 87.4 per cent of all statistics are made up on the spot!

r^2 = 0.89

Bowing

10 4 good buddy 15 2 15 4 and a pair is 8
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes DerFish's post
09-03-2016, 04:01 PM
RE: Significant or not?
I have reviewed a number of research papers which claim that "... even though there is no statistical difference the trend suggests that ..." AAAARGGHH!

#sigh
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes GirlyMan's post
09-03-2016, 04:04 PM
RE: Significant or not?
(09-03-2016 04:01 PM)GirlyMan Wrote:  I have reviewed a number of research papers which claim that "... even though there is no statistical difference the trend suggests that ..." AAAARGGHH!

No trend is a trend, right?

Being nice is something stupid people do to hedge their bets
-Rick
Find all posts by this user
Like Post Quote this message in a reply
09-03-2016, 04:05 PM
RE: Significant or not?
(09-03-2016 04:04 PM)TheBeardedDude Wrote:  
(09-03-2016 04:01 PM)GirlyMan Wrote:  I have reviewed a number of research papers which claim that "... even though there is no statistical difference the trend suggests that ..." AAAARGGHH!

No trend is a trend, right?

[Image: pain.gif]

#sigh
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes GirlyMan's post
Post Reply
Forum Jump: