The following parts involve ANOVA concepts in general. e) To calculate mean squa
ID: 3223779 • Letter: T
Question
The following parts involve ANOVA concepts in general. e) To calculate mean squares, the corresponding sum of squares is divided by which corresponding value? f) In an ANOVA table, the calculated test statistic comes from an F distribution with two parameters: numerator degrees of freedom (ndf), and denominator degrees of freedom (ddf]. If SSR has k df and SST has n - 1 df, what are the ndf and ddf, in terms of k and n? g) Consider using ANOVA to compare a "Full" model to a "Reduced" model. In performing the hypothesis test, which model does the null hypothesis support: the "Full" or "Reduced" one? Briefly explain why. h) If a test from part (g) produces in a small p-value, which model does the test suggest is preferred: the "Full" or "Reduced" one?Explanation / Answer
e) To calculate the mean sum of squares , sum of squares is divided by respective degrees of freedom.
f) SST has n-1 degrees of freedom and SSR has k degrees of freedom so F statistic will have ndf and ddf as, ndf = n-1 & ddf = k because SST degreees of freedom are considered as numerator and SSR degrees of freedom are considered as degrees of freedom.
g) Full model explains the total variation of the data and Reduced model does partial explanation of data. The null hypothesis is set as Reject reduced model in favour of full model and alternative hypothesis will be Reject full model in favour of reduced model. So null hypothesis supports Full model.
f) If in test we get small p-value then we will reject the null hypothesis i.e reject reduced model in favour of full model so full model will be preferred.
Related Questions
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.