人工增雨统计检验结果准确度分析

采用自然复随机化方法对人工增雨作业效果检验中常用的4种统计检验方案进行评估结果准确度分析，以“假效果”来反应评估结果准确度，指出影响评估结果准确度的各个因子并对不同统计方案的评估结果准确度进行比较。结果表明：评估结果准确度与作业样本数和历史样本数都有关系，准确度随样本数增多而提高，但随着样本数的增多，准确度的提高将趋于平缓；在相同条件下，区域历史回归试验较其它3种方案评估结果准确度更高，在规定显著度水平下要达到基本相同的评估结果准确度，区域历史回归试验所需样本数最少，其次为双比分析，但序列试验所需样本数最多，对比试验“假效果”区间分布成偏态，不宜用来做效果检验。还介绍了一种较少使用的统计分析方案——回归比率分析法，通过“假效果”分析表明，回归比率法比区域历史回归方案评估结果准确度更高。

Accuracy Analysis of Statistical Evaluation Result in Precipitation Enhancement Experiment

By using the natural complex randomization method,the accuracy analysis is made of four kinds of statistical test schemes generally used in precipitation enhancement effectiveness evaluation,in which the "false effectiveness" is used to reflect the accuracy of evaluation to find out the different factors affecting the accuracy and to compare the accuracy of the different statistical schemes.It is found that the accuracy of the evaluation is relative to the quantityies of operation samples and historical sam...