h Each Sum of squares has a corresponding degrees of freedom df associated with

# H each sum of squares has a corresponding degrees of

This preview shows page 25 - 30 out of 66 pages.

h) Each Sum of squares has a corresponding degrees of freedom, df associated with it. i) Total df is n - 1, one less than the number of observations. j) Regression df is the number of independent variables in the model. For SLR, it is 1. k) Error df is the difference between total df and Regression df. l) Mean Squares (MS) are the sum of squares divided by the corresponding df. m) F statistic or F ratio is the ratio of SS divided by the corresponding df. n) In the context of regression, the p value reported in the table gives us an overall test for the significance of our model. o) As p value (2 e-16) is less than 0.05, we reject the null hypothesis that there is no relationship between predictor and response and conclude that there is a significant relationship between the predictor and response variables. VIII. Tests of significance using ANOVA Regression 5-Jan-20 Great Learning 26 6. Simple Linear Regression (SLR) Model building IX. Hypothesis test for regression coefficient (t test) and p values H 0 : There is no relationship between X and Y or β 1 = 0 H A : There is a relationship between X and Y or β 1 ≠ 0 Test statistic follows a t distribution with n-2 df where n is the number of observations. We reject the null hypothesis at 5% level of significance, when the p value corresponding the t statistic < 0.05. Regression 5-Jan-20 Great Learning 27 IX. Hypothesis test for regression coefficient (t test) and p values We observe that the overall mean square of prediction error is 4.53 Model parameters estimated: β0 = 2.2526; β1 = 1.9689 Regression equation is given as Revenue = 2.2526 + 1.9689 * Promotion Expenses Interpretation of the equation: For every one unit increase in Promotion expenses, revenue will increase at the rate of 1.9689 on the average. R square is 0.996 tells us the model explains 99.6% of the variation in the dependent variable, Revenue. 6. Simple Linear Regression (SLR) Model building Regression 5-Jan-20 Great Learning 28 6. Simple Linear Regression (SLR) Model building IX. Hypothesis test for regression coefficient (t test) and p values Standard error of the residual is 2.04 on 48 degrees of freedom Standard error of the X is 0.0191 and t value is 103.16. For this two-tailed test with alpha = 0.05 & degrees of freedom = 48 (50 -2), t critical value is 2.01 Our t calculated value of 103.16 > t critical value of 2.01. Alternatively, p value of 2e-16 < 0.05 So we have evidence to reject the null hypothesis and conclude that there exists a relationship between X and Y and β 1 ≠ 0 Regression 5-Jan-20 Great Learning 29  #### You've reached the end of your free preview.

Want to read all 66 pages?

• • • 