CAS MA 416 Lecture Notes - Lecture 6: Null Hypothesis
Document Summary
Y: number of minutes spent on service call. B1 is the change in number of minutes spent on service call on average when the number of copiers serviced increases 1 unit. Y= -0. 92247+15. 05x1+0. 76x2: a 95% confidence interval: (-4. 85, 6. 37). We are 95% confident that the effect of copier model on mean service time would be in the interval(-4. 85, 6. 37), which means that there is a 95% probability that copier model have no effect on mean service time. 8. 19 proc means; class type x1x2; var minsoncall numb; run; proc reg data=copiermaintenance; model minsoncall=numb type x1x2/p r clb; run; H0: b3=0(no interaction between copier type and number serviced) Ha:b3 does not equal to 0 (interaction between copier type and number serviced) We should reject the null hypothesis at 0. 10 significance level and conclude that the interaction term cannot be dropped from the model.