Page 40 - FINAL CFA II SLIDES JUNE 2019 DAY 2
P. 40

MODULE 8.3: ANOVA AND THE F-TEST                              READING 8: MULTIPLE REGRESSION AND ISSUES IN REGRESSION ANALYSIS






      Step 2: Determine the significance of the individual independent variables.


       Can be done using t-tests and p-values.



       •    Using p-values. Only the p-value of the coefficient for ΔI < 5% level of significance, so we conclude that only ΔI
            contributes significantly to the level of BuildCo’s annual sales.


       •    Using t-statistics: H : b = 0 versus H : b ≠ 0
                                     j
                                0
                                                   a

                                                                 Decision rule:
        The critical 2TT t-values with df = 19 are ±2.093.       Reject H if –2.093 >t > 2.093       But t GDP  = 1.33: Within range/CI!
                                                                          0


        We can’t reject Ho for GDP, and we conclude that the level of GDP does not make a statistically significant contribution to
        the variation in sales at the 5% level.


                                                       We conclude that changes in mortgage rates make a significant contribution
         Since (t = –5.758) < (t = –2.093)             to the variation in sales at the 5% level.
                 ΔI
                                  c

        The use of p-values or t-tests will always result in the same conclusions about the statistical significance of the slope
        estimate (i.e., coefficients on the independent variables). On the exam, use the p-value if it is provided!
   35   36   37   38   39   40   41   42   43   44