论文部分内容阅读
目的 :探讨两水平方差成份模型在评价调查者间变异的应用价值。方法 :针对对问题的应答为连续性正态分布的数据 ,分别拟合了两水平方差成份模型进行对比分析。结果 :问卷中各个问题存在着一定的调查者间变异 ,但逐年改进。结论 :对于问卷项目的应答为连续性正态分布的数据 ,在拟合重复调查的情形下 ,可以采用两水平方差成份模型评价问卷项目应答的信度。
Objective: To explore the application value of the two-level variance component model in evaluating inter-individual variation. Methods: For the data with continuous normal distribution, the two-level variance component models were fitted and compared. Results: There are certain inter-investigator variations in each question in the questionnaire, but it improves year by year. Conclusion: The response to the questionnaire project is a continuous normal distribution data. In the case of repeated surveys, the two-level variance component model can be used to evaluate the reliability of the questionnaire project response.