Pengujian Intersep pada Uji Satu Arah Maksimum untuk Tes terkait Nonsample Prior Information

Abstract

This research studied testing the intercept for one-side hypothesis (maximum) on a simple linear regression of tests that related to non-sample prior information (NSPI). The tests, namely unrestricted test (UT), restricted test (RT), and preliminary-test test (PTT), are used. The maximum power and minimum size are considered as a good test. A simulation study is presented for showing a graphically comparison of the tests. The result shown that the NSPI did not give a significant influence. It means that the PTT was not be an alternative choice between UT and RT, but they still follow the concept of the previous research. Keywords: Linear model, power, and size.