- CFA Exams
- CFA Level I Exam
- Topic 1. Quantitative Methods
- Learning Module 10. Simple Linear Regression
- Subject 3. Analysis of Variance
CFA Practice Question
An analyst is currently reviewing the following regression analysis:
At a significance level of 5%, which of the following statements would best reflect the significance of the regression coefficients together as a whole?
A. The calculated F-statistic is greater than the critical F-statistic of 7.25, thus making the model significant.
B. The calculated F-statistic of 13.71 is inside the bound of the acceptance range, thus making the model insignificant.
C. The calculated F-statistic of 13.71 is outside the bound of the acceptance range, thus making the model significant.
Explanation: To test the fit of the simple linear regression, we can calculate an F-distributed test statistic and test the hypothesis H0:b1 = 0 versus Ha:b1 != 0, with 1 and n-2 degrees of freedom.
H0: b1 = b2 = b3 = 0.
Hb1: At least one of the coefficients does not equal to 0.
1. Set up the test (F-test):
H0: b1 = b2 = b3 = 0.
Hb1: At least one of the coefficients does not equal to 0.
2. Compute the 'Calculated F': Fcalc = MSSR/MSSE = 7391/539 = 13.71.
3. Compute the 'Critical F'. Fcalc at the 5% significance level with a d of f(MSSR) = 3 (numerator) and d of f(MSSE) = 46 (denominator) equals: (approximately) 2.81.
4. Decision: since the calculated F statistics is greater than the critical value, the null hypothesis is rejected. Thus, at least one of the coefficients does not equal 0, giving the regression equation significance.
User Contributed Comments 0
You need to log in first to add your comment.