Calculations from an ANOVA table

516 Views Asked by At

From running the program, we will get the corresponding ANOVA table. enter image description here

I then was able to calculate almost all except was stuck on calculated the adjusted R^2. I calculated the R^2 as RSS/tSS = 45.52940/63/8150 = 0.7144.

However when I tried Adj. R^2 formula of 1 - (RSS/(n-k-1))/(TSS/n-1)), it did not work out when I used k = 1. I say k = total # of parameters

That is, it didn't match the output from that of the table.

Also, I am needing help to calculate the standard errors and t-Value under "Parameter Estimtes" in the table.

enter image description here

1

There are 1 best solutions below

0
On

$R^2_{adj} = 1 - \frac{SSE/(25-2)}{SST/(25-1)} = 1- \frac{18.223}{63.815/24} = 0.702.$

For the standard errors recall that $$ \widehat{Var(\hat{\beta_1})} = \frac{MSE}{S_{XX}}, $$ and $$ \widehat{Var(\hat{\beta_0})} = MSE\left(\frac{1}{n} + \frac{\bar{X}_n^2}{S_{XX}}\right). $$ Where for the t.values, after calculating the s.e., just divide the parameters' estimators by their s.e., i.e, $$ t_{\hat{\beta_j}} = \frac{\hat{\beta}_j}{s.e(\hat{\beta_j)}}, \quad j=0,1. $$