Jump to content

Err America files Chapter 11


KD in CA

Recommended Posts

  • Replies 598
  • Created
  • Last Reply

Top Posters In This Topic

Conversely, you could take the error completely out of your simulation.  Test your population, then test them again.  "On average", people who score very well/poorly will score less well/poorly even in the absence of error, because chance (i.e. the probability distribution of a normal distribution) dictates it.

That's not the way I set up my simulation. Each given member's true I.Q. stayed constant for the two tests. Therefore, someone who scored a 155 on the first error-free I.Q. test would score a 155 on the second, the third, and the fiftieth.

Link to comment
Share on other sites

I am right.  I can mathematically prove I'm right.  I've explained why I'm right.  Other people have explained why I'm right.  You're just too friggin' stupid to understand.

834668[/snapback]

You don't understand the phenomenon my simulation is intended to exhibit. You don't understand how my simulation was set up. You didn't understand the Wikipedia article about regression toward the mean, nor the HyperStat article to which I linked. You didn't understand that my simulation is set up in the same way, and intended to prove the same point, as the simulation to which HyperStat linked. Nearly every word you've written about the relationship between measurement error and regression toward the mean has been based on a faulty and incorrect understanding of the heart of the issue. You still don't understand it, despite my Herculean efforts to explain it to you.

Link to comment
Share on other sites

You don't understand the phenomenon my simulation is intended to exhibit. You don't understand how my simulation was set up. You didn't understand the Wikipedia article about regression toward the mean, nor the HyperStat article to which I linked. You didn't understand that my simulation is set up in the same way, and intended to prove the same point, as the simulation to which HyperStat linked. Nearly every word you've written about the relationship between measurement error and regression toward the mean has been based on a faulty and incorrect understanding of the heart of the issue. You still don't understand it, despite my Herculean efforts to explain it to you.

834686[/snapback]

 

Yes, I do. You're just wrong.

Link to comment
Share on other sites

You don't understand the phenomenon my simulation is intended to exhibit. You don't understand how my simulation was set up. You didn't understand the Wikipedia article about regression toward the mean, nor the HyperStat article to which I linked. You didn't understand that my simulation is set up in the same way, and intended to prove the same point, as the simulation to which HyperStat linked. Nearly every word you've written about the relationship between measurement error and regression toward the mean has been based on a faulty and incorrect understanding of the heart of the issue. You still don't understand it, despite my Herculean efforts to explain it to you.

834686[/snapback]

The Sabres have been tied after regulation a lot this year. Does this mean they are better at regressing toward the mean than most other teams?

Link to comment
Share on other sites

That's why they're so good.  They're less wrong than most teams.

836486[/snapback]

Maybe that would look good on a shirt. "We're less wrong than most teams. (Except when we try to pick a new friggin' logo.)" It would definitely look better than the new logo.

 

I'm not sure what crayons should be used to color the shirt though.

Link to comment
Share on other sites

Maybe that would look good on a shirt.  "We're less wrong than most teams.  (Except when we try to pick a new friggin' logo.)"  It would definitely look better than the new logo.

 

I'm not sure what crayons should be used to color the shirt though.

837358[/snapback]

 

Should we call JP-era back to help with the coloring?

Link to comment
Share on other sites

  • 2 weeks later...
×
×
  • Create New...