Question 3
You wanted to improve the model from question two and have decided to include a new
variable, percentage of English learners (EL PCT).
SUMMARY
OUTPUT
Regression Statistics
Multiple R 0.653017122
R Square 0.426431361
Adjusted R Square 0.423680433
Standard Error 14.46448496
Observations 420
ANOVA
df SS MS F
Regression 2 64864.3011 32432.15055 155.0135989
Residual 417 87245.29254 209.221325
Total 419 152109.5936
Coefficients
Standard
Error t Stat P-value
Intercept 686.0322487 7.411312484
str -1.101295889 0.380278316
el_pct -0.649776777 0.039342549
(a) Is this second model a better or worse model than the one in question 2? Why?
(b) Are all the signs of the variables as expected? State reason.
(c) A colleague of yours believes that English learners is not significant in determining
test scores. Conduct a hypothesis test to prove whether this is true or not assuming
=0.05 i.e myu= 0.05
As question 2 is:
Question 2
The following regression output was obtained. The regression model is Test scores on
student teacher ratio. When your research assistant cut and pasted the results for you, there
were several vital pieces of information missing. See output below:
SUMMARY
OUTPUT
Regression Statistics
Multiple R 0.226362751
R Square (a)
Adjusted R Square 0.048970335
Standard Error 18.5809675
Observations 420
ANOVA
df SS MS F
Regression 1 7794.110041 7794.110041 22.57511055
Residual 418 144315.4836 345.2523531
Total 419 (b)
Coefficients
Standard
Error t Stat P-value
Intercept 698.9329523 9.467491444
str -2.279808287 0.479825567 (c)
