Search All of the Math Forum:

Views expressed in these public forums are not endorsed by NCTM or The Math Forum.

Notice: We are no longer accepting new posts, but the forums will continue to be readable.

Topic: Regression equations and rounding.
Replies: 1   Last Post: Nov 6, 1996 10:57 AM

 Messages: [ Previous | Next ]
 DougKuhlmann-PhillipsAcademy-Math Posts: 119 Registered: 12/6/04
Regression equations and rounding.
Posted: Nov 5, 1996 11:33 AM

On the TI-83, if your machine is set to 2 decimal rounding (under the MODE
menu) and you calculate, say, a linear regression line, the coefficients are
calculated to the full fourteen digits but only displayed to 2 decimal
accuracy. However, if you then store the regression equation into Y1 by
choosing it from the menu, it only stores two decimal places. For example,
if the machine calculates the equation y=ax+b and determines that
a=.1683645274534 and b=245.23785957587 when you then store the regression
equation into Y1 you get Y1=.17x+245.24. If you then change the MODE
setting to Float but do not recalculate the regression equation and try to
store the equation into Y2 you get Y2=.17+245.24. I.E. the machine somehow
rounds the equation permanently, even though the coefficients a and b are not.

Strange.

Doug

--
Doug Kuhlmann (508)-749-4242