研究生: |
宋品勳 |
---|---|
論文名稱: |
多維度結果依賴採樣下的長期追蹤資料對相關矩陣結構的選擇 Selecting a Working Correlation Structure for Longitudinal Data under The Multivariate Outcome-Dependent Sampling Design |
指導教授: | 呂翠珊 |
學位類別: |
碩士 Master |
系所名稱: |
數學系 Department of Mathematics |
論文出版年: | 2019 |
畢業學年度: | 107 |
語文別: | 英文 |
論文頁數: | 30 |
中文關鍵詞: | 結果依賴採樣設計 、多變量結果依賴採樣設計 、相關矩陣結構 |
英文關鍵詞: | outcome-dependent sampling design, multivariate, working correlation structure |
DOI URL: | http://doi.org/10.6345/NTNU201900704 |
論文種類: | 學術論文 |
相關次數: | 點閱:158 下載:19 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
結果依賴採樣設計在很多研究中已被證實是具有成本效益的抽樣方式。若同一位實驗對象有多於兩個觀測值,則存在無法忽略的相關性。給出不正確的相關矩陣類型結構的假設,可能影響估計值的準確性。因此,本研究的目標是考慮多個實作相關矩陣結構,依據多維度依賴採樣設計所得到的長期追蹤資料,如何找到有效的參數估計。我們的模擬試驗已證實在不同的相關矩陣下,多維度依賴採樣下的估計值皆比簡單抽樣的估計值更有效率。我們計算 AIC 和 BIC 的值來當作我們選擇合適的相關矩陣的指標。最後,我們使用此模型去分析關於牙齒修復的資料。
Outcome-dependent sampling (ODS) scheme has been shown to be a cost-effective scheme in a lot of large-cohort studies. A multivariate outcome-dependent sampling (MODS) design is a further generalization for longitudinal data collected under the ODS scheme. For multivariate responses, the correlation between the responses from the same subject cannot be neglected. Misspecified working correlation structures may lead to erroneous conclusions. In this study, we consider an ODS design for multivariate longitudinal or clustered data and model under various types of the working correlation structure. A semiparametric empirical likelihood approach is developed for the proposed design under several commonly-used working correlation structures. Simulation studies show that the proposed estimator is more efficient than the estimator from a simple random sample of the same sample size. We then used AIC and BIC to obtain the appropriate correlation matrix. We also apply our proposed approach to the dental restoration data.
A. Weaver, M., and Zhou, H.An estimated likelihood method forcontinuous outcome regression models with outcome-dependent sampling.Journal of the American Statistical Association 100(02 2005), 459–469.
Amemiya, T.Regression analysis when the dependent variable is truncatednormal.Econometrica 41, 6 (1973), 997–1016.
Caplan, D., Li, Y., Wang, W., Kang, S., Marchini, L., Cowen,H., and Yan, J.Dental restoration longevity among geriatric and specialneeds patients.JDR Clinical & Translational Research 4, 1 (2019), 41–48.PMID: 30931764.
Cosslett, S. R.Maximum Likelihood Estimator for Choice-Based Sam-ples.Econometrica 49, 5 (September 1981), 1289–1316.
Gosho, M.Criteria to select a working correlation structure for the gen-eralized estimating equations method in sas.Journal of Statistical Software57, 1 (2014), 1–10.
Hin, L.-Y., and Wang, Y.-G.Working-correlation-structure identifica-tion in generalized estimating equations.Statistics in medicine 28, 4 (2009),642–658.
Liang, K.-Y., and Zeger, S. L.Longitudinal data analysis using gener-alized linear models.Biometrika 73, 1 (1986), 13–22.
Lu, T.-S., Longnecker, M., and Zhou, H.Statistical inferencesfor data from studies conducted with an aggregated multivariate outcome-dependent sample design.Statistics in Medicine 36, 6 (3 2017), 985–997.
Pan, W.Akaike’s information criterion in generalized estimating equations.Biometrics 57, 1 (2001), 120–125.
Zeger, S. L., and Liang, K.-Y.Longitudinal data analysis for discreteand continuous outcomes.Biometrics(1986), 121–130.
Zhou, H., Chen, J., Rissanen, T., A Korrick, S., Hu, H., T Sa-lonen, J., and P Longnecker, M.Outcome-dependent sampling - anefficient sampling and inference procedure for studies with a continuous out-come.Epidemiology (Cambridge, Mass.) 18(08 2007), 461–8.
Zhou, H., Chen, J., Rissanen, T. H., Korrick, S. A., Hu, H., Salo-nen, J. T., and Longnecker, M. P.An efficient sampling and inferenceprocedure for studies with a continuous outcome.Epidemiology (Cambridge,Mass.) 18, 4 (2007), 461.
Zhou, H., Song, R., Wu, Y., and Qin, J.Statistical inference for atwo-stage outcome-dependent sampling design with a continuous outcome.Biometrics 67, 1 (2011), 194–202.
Zhou, H., Weaver, M. A., Qin, J., Longnecker, M., and Wang,M. C.A semiparametric empirical likelihood method for data from anoutcome-dependent sampling scheme with a continuous outcome.Biometrics58, 2 (2002), 413–421.