| 胡宝文.2009.动物学定量与定性资料分析中的常见错误.动物学杂志,44(6):41-47. |
| 动物学定量与定性资料分析中的常见错误 |
| Common Errors in Qualitative and Quantitative Data Analysis in Zoological Research |
| 投稿时间：2009-03-29 修订日期：2009-08-03 |
| DOI： |
| 中文关键词: t-检验 方差分析 卡方检验 列联表 定量资料 定性资料 |
| 英文关键词:t-test Analysis of variance Chi-squareχ2 Contingency table Qualitative data Quantitative data |
| 基金项目:国家科技支撑项目“中国重要生物物种资源监测和保育关键技术与应用示范资助”(No.2008BAC39B04) |
|
| 摘要点击次数: 2121 |
| 全文下载次数: 2653 |
| 中文摘要: |
| 合理选择正确的统计方法是提高科研工作质量和可信度的关键。然而有些科研工作所采用的统计方法往往不合理,甚至是错误的。例如,常常忽视定量资料中t-检验和方差分析的前提条件——正态性、方差齐性。对单因素多水平的定量资料直接使用组间两两比较的t-检验;对定性资料中列联表的分析一律通用Pearson χ2检验,包括对Pearson χ2检验结果的理解出现偏差,从而做出错误的推论。本文对国内一些动物学研究论文中所采用的具体统计方法提出辨析。 |
| 英文摘要: |
| Choosing right statistical method is a key point to improve quality and credibility of scientific research.However,some methods used by biologists are often unreasonable or even incorrect.For instance,the normality and homogeneity of variance are often neglected in the t-test or Single-Factor Analysis of Variance.Two sample t tests are often employed to address the difference between multiple levels in one factor.The Pearson’s χ2 has been frequently adopted in the contingency table,but inappropriate analyses and inferences are very common.Five examples from the wildlife literatures are used to illustrate the above questions. |
| 附件 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
