My Math Forum Regression coefficient

 January 18th, 2017, 11:03 AM #1 Member   Joined: Jan 2017 From: California Posts: 45 Thanks: 4 Regression coefficient Hi everyone, First I ran a simple regression model. Second I ran a multiple regression model by introducing a new independent variable. Both $R^2$ and $Adjusted$ $R^2$ improved which was a good thing. The coefficient for the simple regression also improved. How can we explain why this coefficient improved? Thanks in advance
 January 18th, 2017, 03:06 PM #2 Member   Joined: Jan 2017 From: California Posts: 45 Thanks: 4 for more precision I meant coefficient of the independent variable
 January 18th, 2017, 03:07 PM #3 Senior Member     Joined: Sep 2015 From: CA Posts: 1,207 Thanks: 614 some sort of systematic bias in your experiment? like temperature dependent noise or something like that
January 18th, 2017, 03:38 PM   #4
Member

Joined: Jan 2017
From: California

Posts: 45
Thanks: 4

Quote:
 Originally Posted by romsek some sort of systematic bias in your experiment? like temperature dependent noise or something like that
Nah nothing like that

January 18th, 2017, 03:49 PM   #5
Senior Member

Joined: Sep 2015
From: CA

Posts: 1,207
Thanks: 614

Quote:
 Originally Posted by dthiaw Nah nothing like that

was it $y = m x$ or $y = m x + b$ ?

January 18th, 2017, 03:55 PM   #6
Member

Joined: Jan 2017
From: California

Posts: 45
Thanks: 4

Quote:
 Originally Posted by romsek your simple regression was it $y = m x$ or $y = m x + b$ ?
$y=mx + b$

January 18th, 2017, 04:18 PM   #7
Senior Member

Joined: Sep 2015
From: CA

Posts: 1,207
Thanks: 614

Quote:
 Originally Posted by dthiaw $y=mx + b$
got me then..

I can see why using 2 degrees of freedom would improve the overall fit but I can't see why that would improve your linear regression.

January 28th, 2017, 08:02 PM   #8
Senior Member

Joined: Dec 2012
From: Hong Kong

Posts: 789
Thanks: 285

Math Focus: Stochastic processes, statistical inference, data mining, computational linguistics
Quote:
 Originally Posted by dthiaw Hi everyone, First I ran a simple regression model. Second I ran a multiple regression model by introducing a new independent variable. Both $R^2$ and $Adjusted$ $R^2$ improved which was a good thing. The coefficient for the simple regression also improved. How can we explain why this coefficient improved? Thanks in advance

Can you show us the data and/or the printouts?

Or were you actually looking at Pearson's correlation coefficient the second time, and the coefficient of determination the first??

Or maybe you're looking at the regression coefficient between the new independent variable and the dependent variable?

Last edited by 123qwerty; January 28th, 2017 at 08:06 PM.

 Thread Tools Display Modes Linear Mode

 Similar Threads Thread Thread Starter Forum Replies Last Post chesspoker Algebra 2 April 2nd, 2015 12:59 PM Davuit Advanced Statistics 0 May 9th, 2013 09:24 AM ixmardukxi Algebra 1 March 8th, 2013 04:32 AM hollymxox Algebra 0 March 21st, 2012 04:46 PM Francis410 Physics 6 February 13th, 2011 09:04 PM

 Contact - Home - Forums - Cryptocurrency Forum - Top