Multicollinearity
Posted: Tue Nov 26, 2013 11:03 pm
Hello,
I am working on a research/study and could use some help. My study is the affects on SAT scores for 2012. I am using the dependent variable: US SAT Scores for 2012.
I am using the Independent variables: US Median Household Income; US PSAT/MATH scores; US PSAT/Critical Reading scores; and USPSAT Writing scores. (For 51 states)
(Below are the results)
It appears that multicollinearity is occurring; So I want to combine two independent variables (writing and reading) and use STATE demographics as another IND variable. What do you believe I should do??
(Following is the regression I received)
Dependent Variable: SAT
Method: Least Squares
Date: 11/27/13 Time: 00:29
Sample: 1 51
Included observations: 51
Variable Coefficien Std. Error t-Statistic Prob.
C -886.4317 266.4041 -3.327395 0.0017
MEDIAN_HOUSEHOLD_INCOME -0.001445 0.001359 -1.063508 0.2931
PSAT_MATH 30.19391 14.79523 2.040787 0.0470
PSAT_READING 0.174249 27.58466 0.006317 0.9950
PSAT_WRITING_SKILLS 22.09775 19.12802 1.155256 0.2540
R-squared 0.700140 Mean dependent var 1585.020
Adjusted R-squared 0.674065 S.D. dependent var 125.9756
S.E. of regression 71.92041 Akaike info criterion 11.48189
Sum squared resid 237937.1 Schwarz criterion 11.67129
Log likelihood -287.7882 Hannan-Quinn criter. 11.55426
F-statistic 26.85119 Durbin-Watson stat 1.864842
Prob(F-statistic) 0.000000
Thank you for your time....I really need help111
I am working on a research/study and could use some help. My study is the affects on SAT scores for 2012. I am using the dependent variable: US SAT Scores for 2012.
I am using the Independent variables: US Median Household Income; US PSAT/MATH scores; US PSAT/Critical Reading scores; and USPSAT Writing scores. (For 51 states)
(Below are the results)
It appears that multicollinearity is occurring; So I want to combine two independent variables (writing and reading) and use STATE demographics as another IND variable. What do you believe I should do??
(Following is the regression I received)
Dependent Variable: SAT
Method: Least Squares
Date: 11/27/13 Time: 00:29
Sample: 1 51
Included observations: 51
Variable Coefficien Std. Error t-Statistic Prob.
C -886.4317 266.4041 -3.327395 0.0017
MEDIAN_HOUSEHOLD_INCOME -0.001445 0.001359 -1.063508 0.2931
PSAT_MATH 30.19391 14.79523 2.040787 0.0470
PSAT_READING 0.174249 27.58466 0.006317 0.9950
PSAT_WRITING_SKILLS 22.09775 19.12802 1.155256 0.2540
R-squared 0.700140 Mean dependent var 1585.020
Adjusted R-squared 0.674065 S.D. dependent var 125.9756
S.E. of regression 71.92041 Akaike info criterion 11.48189
Sum squared resid 237937.1 Schwarz criterion 11.67129
Log likelihood -287.7882 Hannan-Quinn criter. 11.55426
F-statistic 26.85119 Durbin-Watson stat 1.864842
Prob(F-statistic) 0.000000
Thank you for your time....I really need help111