查詢結果分析
來源資料
頁籤選單縮合
題 名 | 量化研究的品管--統計考驗力與效果值分析=Quality Control for Quantitative Studies: Power and Effect Size Analysis |
---|---|
作 者 | 李茂能; | 書刊名 | 國民教育研究學報 |
卷 期 | 8 2002.03[民91.03] |
頁 次 | 頁1-24 |
分類號 | 510 |
關鍵詞 | 統計考驗力分析; 效果值; 量化研究; Power analysis; Effect size; Quantitative studies; |
語 文 | 中文(Chinese) |
中文摘要 | 本文旨在剖析樣本規劃、臨床或實用顯著性與統計顯著性的密切關係,以提昇量化研究的品質。首先點出與第一類型錯誤、第二類型錯誤、統計考驗力、P值、及樣本大小有關之迷思、誤用或誤解。其次,探討統計考驗力與效果值分析的意義秈功能,同時舉一實例說明如何透過統計考驗力與效果值分析,提昇量化研究的品質。接苦,說明研究者進行樣本規劃並非難事,只要透過 Cohcn 氏查表法與統計考驗力分析軟體 C*Power,即能伙速達成。文末,試著針對論 文編審提出量化研究品管的具體方法。 |
英文摘要 | This paper addresses the close interlock between sample size determination, statistical significance, and practical/clinical significance. Several misconceptions, misuses, and misinterpretations related to Type-I error, Type-Il error, power, p-values, and sample size are highlighted. Then, those factors such as α, β, & sample size that may affect power are specified and defined. It is argued that the quality of quantitative studies can only be assured through power and effect size analysis. For a scientific conclusion of statistical significance and practical/clinical significance, power analysis should be done in the planning stage of study and minimum practical effect size should be considered with statistical significance testing. Next, computing power for any specific study may not be a difficult task via Cohen's power table and a computer program called G*Power. Finally, recommendations regarding editorial policies and reviewing practices for the integration of power analysis and effect size into the study plan or the test result are pinpointed. |
本系統中英文摘要資訊取自各篇刊載內容。