Suppose we have two Binomial proportion GLMs Model 1 and Model 2, where one of t

Question

Suppose we have two Binomial proportion GLMs Model 1 and Model 2, where one of the models is nested within the other. Suppose we have On the basis of this information, which model is preferred? Explain clearly why. Provide as many reasons as you can based only on this information. If you use a hypothesis test as part of your answer, you should clearly state Your null and alternative hypotheses in terms of the two models. your test statistic and its approximate distribution if one of the models is true (specify which one) your decision rule if testing at significance level a 001. and your decision. You may also choose to include another reason/reasons as part of your answer.

Explanation / Answer

here residual deviance means variance

we use F-test to test equality of two variances

and F=s12/s22 =8.9/25.2=,0.3531

(first part) the null H0 and alternative hypothesis Ha are given as

(second part)

and F=s12/s22 =8.9/25.2=,0.3531

(third part) decsion rule

F<F(1/2,N11,N21)=0.141 ( using ms-excel command =FINV(1-0.01/2,8,11)) or

F> (F/2,N11,N21)=5.68 ( using ms-excel command =FINV(0.01/2,8,11) )

(fourth part)

since the calculated F=0.3532lies betwee 0.141 and 5.68, so we fail to reject null hypothesis and conclude that variance of both the model are statistically not different. either model can be used.

F<F(1/2,N11,N21)=0.141 ( using ms-excel command =FINV(1-0.01/2,8,11)) or

F> (F/2,N11,N21)=5.68 ( using ms-excel command =FINV(0.01/2,8,11) )

(fourth part)

since the calculated F=0.3532lies betwee 0.141 and 5.68, so we fail to reject null hypothesis and conclude that variance of both the model are statistically not different. either model can be used.

  

Related Questions

Navigate

• Browse (All)
• Browse S
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.