摘要
测量的可靠性是评价测量质量的重要指标。一般认为测量的可靠性代表测量结果的准确性,这显然是一种误解,尤其是目前应用较多的几种可靠性估价方法似乎无法对可靠性作出正确的估价。为此,对于可靠性及其估价方法应重新加以审视和评析,以期重新认识并有所改善。通过分析和示例结果,可以得出如下结论:测量的可靠性不能完全表达测量结果的准确度;在可靠性定义中一致性不能由相关或相似性替代;在进行差异分析时,应注意重复测量结果的关联属性,并推荐使用非参数检验方法;可靠性分析不能替代误差分析与处理等其他方法。
The reliability of measurement is an important factor to evaluate the quality of measurement. The understanding on reliability generally focus on the accuracy of measurement. It is obviously a misunderstanding, especially for some assessment methods used widely nowadays that can＇t make correct assess to the reliability of measurement. Therefore, this paper analyzed the traditional method of assessment and its reliability in order to re-recognize it. According to the analysis and the samples, some conclusions are drawn as the following., the reliability can＇t completely express the accuracy of the measurement; the consistency to the definition of reliability can＇t be replaced by relativity or similarity; when making the analysis of variance, the relevance of repeated measuring results should be taken in, and the non-parameter test is recommended; the reliability analysis can＇t replace the error analysis and error -conduct other else methods.
出处
《体育科学》
CSSCI
北大核心
2007年第2期90-93,共4页
China Sport Science
关键词
测量
可靠性
信度
measurement
reliability
reliability
