Argyanti, Talitha Tertia Larissa (2019) Perbandingan Teknik Imputasi Missing Data Menggunakan Metode Mean Substitution Dan Maximum Likelihood Estimated Mean Substitution Pada Person Trait Estimation. Sarjana thesis, Universitas Brawijaya.
Abstract
Penelitian ini bertujuan untuk mengetahui perbandingan teknik imputasi mean substitution dengan maximum likelihood estimated mean substitution untuk mengatasi missing data dalam mengestimasi person-trait. Penelitian ini menggunakan data yang bersumber dari harvard dataverse. Data yang digunakan merupakan hasil dari skala Self/Peer Inventory (S/PI) dengan missing data sebanyak .035% dan skala International Personality Item Pool (IPIP) dengan missing data sebanyak 32.35% dari Oreogon Research University yang memiliki respon polytomous, 40 item dan 1000 responden. Estimasi person-trait dilakukan dengan Expected A Posteriori (EAP) berdasarkan model Graded Response Model (GRM). Hasil estimasi person-trait dikorelasikan menggunakan pearson product moment. Hasil penelitian menunjukkan bahwa estimasi person-trait dari mean substitution lebih rendah hasilnya daripada MLE mean substitution ketika missing data sebanyak 30.22%.
English Abstract
This study aims to determine the compare the imputation technique of the mean substitution and maximum likelihood estimated mean substitution to overcome missing data in estimating person trait. This research using data from harvard dataverse. The used data are the result from Self/Peer Inventory (S/PI) with missing data .035% and International Personality Item Pool (IPIP) scale with missing data 30.22% from Oregon Research University which consist of polytomous response, 40 items, and 1000 respondents. The person-trait estimation is done by Expected A Posteriori (EAP) based on the Graded Response Model (GRM) model. The person-trait estimation result will be correlated using pearson product moment. The results showed that the estimate of person-trait from the mean substitution lower than maximum likelihood estimated mean substitution when missing data 30.22%.
Other obstract
-
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FISIP/2019/887/051908305 |
| Uncontrolled Keywords: | Graded Response Model, maximum likelihood estimated mean substitution, mean substitution, missing data, Graded Response Model, maximum likelihood estimated mean substitution, mean substitution, missing data |
| Subjects: | 100 Philosophy, parapsychology and occultism, psychology > 150 Psychology > 150.2 Miscellany > 150.287 Testing and measurement |
| Divisions: | Fakultas Ilmu Sosial dan Ilmu Politik > Psikologi |
| Depositing User: | Budi Wahyono Wahyono |
| Date Deposited: | 17 Nov 2020 17:40 |
| Last Modified: | 17 Nov 2020 17:40 |
| URI: | http://repository.ub.ac.id/id/eprint/174375 |
Actions (login required)
 |
View Item |