YIChen

研究方向

劑量反應研究中，為探討某一種物質之效應，經常比較數個漸增劑量水準(處理組)與零劑量對照組。實例常見於動物實驗或臨床試驗中，評估某種藥物、食品添加物或是殺蟲藥效應之研究。漸增之劑量經常優於或至少與零劑量具有相同之效應，此即簡單樹形處理效應。若漸增劑量水準產生較佳效應，則為順序處理效應。但是，由於研究之物質在高劑量水準時，經常引發反效果，劑量反應呈現先升後降的形態，處理效應則呈傘形。

針對一元配置和集區設計下的完整資料，或是一元設計之下的重複測量資料，我們建立無母數方法進行數個處理組與一對照組之比較。此外，針對右方設限資料，不但建立各種型態處理效應的檢定，也提供多對一和相鄰處理效應的多重比較。

針對劑量反應中最低有效劑量，亦即優於零劑量對照組的最低劑量水準，我們已經在一元配置和集區設計之下提出無母數方法加以鑑別。目前則分別朝多組資料，存活資料及多維度資料的分析加以研究。

除上述的統計研究之外，我們也嘗試與其他研究領域的學者合作，推廣統計之應用。目前我們根據全球8個散佈不同磁緯的電離層觀測站資料，利用兩段迴歸模式配適電離層最大電子濃度與太陽活動性測值，研究太陽活動性對電離層最大電子濃度飽和效應的影響。此外，也研究電離層最大電子濃度變化和台灣強震的時空關係。

Nonparametric Procedures for Patterned Treatment Effects

To investigate the effect of a substance in a dose-response study, several increasing dose levels of the substance (treatments) are usually compared with a zero-dose control. An example of this arises in animal experiments or clinical trials to evaluate the effect of a drug, a food additive or a pesticide. Note that the non-zero dose levels are usually anticipated to be better or at least equal to the zero-dose control, yielding a simple-tree pattern of treatment effects. It also occurs very often that increasing dose levels produce better treatment effects, hence, ordered treatment effects are expected. However, due to adverse effects possibly occurring at high doses, the dose-response shows an up-down ordering which is referred as an umbrella pattern of treatment effects.

Nonparametric procedures are proposed for comparing several treatment with a control under the one-way layout,block design and one-way repeated measures design,respectively ([3]).Testing procedures are suggested for patterned treatment efforts with right-censored survival data.([1],[6],[7]).Multiple comparisons are further developed for the many-to-one problem and the adjacent treatments evaluation with right-censored survival data ([4],[5]).

Identification of the minimum effective dose (MED) in a dose-response study is discussed for the one-way layout and block design ([2],[8]).Current investigations focus on the MED identification for several groups data,survival data and multivariate data,respectively.

We investigated the relationship between the sun spot number and the plasma density (f。F₂) in the ionosphere observed by eight ionospheric stations scattered in different latitudes ([9]). The space-time relationship between the f。F₂ variation and Taiwan's large earthquakes is also explored ([10],[11]).

Selected Recent Publications:

1. Chen, Y. I. (1998). Simple-tree weighted logrank tests for right-censored data. Ann. Inst. Statist. Math. 50, 311-324.

2. Chen, Y. I. (1999). Nonparametric identification of the minimum effective dose.54,1236-1240.

3. Chen, Y. I. (1999). Rank-based tests for dose finding in non-monotonic dose response settings.54,1258-1262.

4. Chen, Y. I. (2000). Multiple comparisons in carcinogenesis study with right-censored survival data. Statist. Med.19,353-367.

5. Chen, Y. I. and Chi, Y. (2000). Multiple comparisons between successive treatments for randomly right-censored survival data. J. Statist. Plan. Infer.83,137-151

6. Chen, Y. I. and Wolfe, D. A. (2000). Umbrella tests for right-censored survival data. Statistica Sinica.10,595-612.

7. Lee, W. C. and Chen, Y. I. (2001). Weighted Kaplan-Meier tests for umbrella alternatives. To appear in Ann. Inst. Statist. Math.

8. Chen, Y. I. and Jan, S. L. (2001). Nonparametric identification of the minimum effective dose for randomized block designs. To appear in Comm.Statist.-Simul. & Comput.

9. Chen, Y. I. , Lin, J. Y. and Chen, S. C. (2000) Statistical investigation of the saturation effort of sunsopts on the ionospheric f。F₂.Physics and Chemistry of the Earth 25,359-362.

10. Liu, J. Y., Chen, Y. I., Pulinet, S. A., Tsai, Y. B. and Chuo, Y. J. (2000). Seismo-ionospheric signatures prior to M ≧6. Taiwan earthquakes. Geophy.Res.Lett.27,3113-3116.

11. Liu, J. Y., Chen, Y. I., Chuo, Y. J. and Tsai, H. F.(2001). Variations of ionospheric total electron content during the Chi-Chi earthquakes. Geophy.Res.Lett.28,1383-1386.