UJI HIPOTESIS DAN INTERVAL KEPERCAYAAN RATIO FUNGSI LINEAR PARAMETER RANCANGAN NESTED TIGA LEVEL

Ibnu Malik, (1327031010) (2015) UJI HIPOTESIS DAN INTERVAL KEPERCAYAAN RATIO FUNGSI LINEAR PARAMETER RANCANGAN NESTED TIGA LEVEL. Other thesis, UNIVERITAS LAMPUNG.

[img]
Preview
Text
ABSTRACT.pdf

Download (198kB) | Preview
[img]
Preview
Text
ABSTRAK.pdf

Download (149kB) | Preview
[img]
Preview
Text
COVER DALAM.pdf

Download (50kB) | Preview
[img]
Preview
Text
COVER LUAR.pdf

Download (50kB) | Preview
[img]
Preview
Text
DAFTAR GAMBAR.pdf

Download (148kB) | Preview
[img]
Preview
Text
DAFTAR TABEL.pdf

Download (146kB) | Preview
[img]
Preview
Text
HALAMAN PENGESAHAN.pdf

Download (120kB) | Preview
[img]
Preview
Text
HALAMAN PERSETUJUAN.pdf

Download (113kB) | Preview
[img]
Preview
Text
KATA PENGANTAR.pdf

Download (37kB) | Preview
[img]
Preview
Text
MOTO.pdf

Download (278kB) | Preview
[img]
Preview
Text
PERNYATAAN.pdf

Download (54kB) | Preview
[img]
Preview
Text
PERSEMBAHAN.pdf

Download (36kB) | Preview
[img]
Preview
Text
RIWAYAT HIDUP.pdf

Download (85kB) | Preview
[img]
Preview
Text
DAFTAR ISI.pdf

Download (9kB) | Preview
[img]
Preview
Text
BAB I.pdf

Download (158kB) | Preview
[img]
Preview
Text
BAB II.pdf

Download (387kB) | Preview
[img]
Preview
Text
BAB III.pdf

Download (89kB) | Preview
[img] Text
BAB IV.pdf
Restricted to Registered users only

Download (637kB)
[img]
Preview
Text
BAB V.pdf

Download (40kB) | Preview
[img]
Preview
Text
DAFTAR PUSTAKA.pdf

Download (86kB) | Preview

Abstract

Rancangan nested tiga level merupakan suatu rancangan percobaan yang mempunyai lebih dari satu ukuran unit percobaan dan terdiri dari tiga faktor. Pada model linear umum rancangan nested tiga level dapat ditulis Y=Xθ+ε, dan model tersebut mempunyai kolom tak berperingkat penuh. Kondisi ini pada beberapa model menimbulkan masalah dalam pendugaan, yaitu pendugaan parameter tidak unik. Model Reduction Methods (MRM) dapat digunakan untuk mentransformasi model tak berpeingkat penuh dengan kendala menjadi model penuh dan tak terkendala. Tujuan dari penelitian ini adalah (1) mentransformasi model tak berpeingkat penuh dengan kendala menjadi model penuh dan tak terkendala, (2) menduga parameter, (3) menguji hipotesis, (4) menentukan interval kepercayaan dari ratio fungsi linear parameter, dan (5) melakukan simulasi menggunakan program R. Berdasarkan hasil tersebut dapat disimpulkan bahwa (1) MRM cocok digunakan untuk entransformasi model tak berpeingkat penuh dengan kendala menjadi model penuh dan tak terkendala (2) hasil simulasi menunjukan karakteristik penduga yang tak bias dan mempunyai varians minimum, (3) pada pengujian hipotesis, power yang dihasilkan baik dan (4) Teorema Fieller dapat digunakan untuk menentukan interval kepercayaan ratio dari fungsi linear parameter. . Kata Kunci : Model reduction methods, pendugaan parameter, pengujian hipotesis, ratio fungsi linear parameter ABTRAK BAHASA INGGRIS Three level of nested design is an experimental design which has more than one size of experimental unit, and consist of at three factors. In general linear model three level of nested design can be written as Y=Xθ+ε, and the model is non full colomn rank. This type of condition in some model create a problem of estimation, namely the etimation parameter is not unique. Model Reduction Methods (MRM) is used to the transform non full rank model with contraint into full rank model and uncontraint. The aim of this research is (1) to transform non full rank model with contraint into full rank model and uncontraint, (2) to estimation parameter, (3) to testing hypotheses, (4) to determine confidence interval ratio of linear function of parameters, and (5) to do simulation using R programe. Based on the results, it can be concluded that (1) MRM is suitable being used to tranform non full rank model into full rank model, (2) the result of the simulation shows characteristic parameter is unbiased and has minimum variance, (3) in the testing hypotheses, the power is good and (4) Filler’s theorem can be used to determine confidence interval ratio of linear function of parameters. Keyword : Model reduction methods, parameter estimation, testing hypotheses, ratio of linear function of parameters

Item Type: Thesis (Other)
Subjects: Q Science (General) > QA Mathematics
Divisions: Fakultas MIPA > Prodi Magister Ilmu Matematika
Depositing User: 1855314 . Digilib
Date Deposited: 06 Jan 2016 08:20
Last Modified: 06 Jan 2016 08:20
URI: http://digilib.unila.ac.id/id/eprint/16606

Actions (login required)

View Item View Item