不同的多个数据样本联合分析方法中共振态显著性的差异

Difference of Resonance Significance in Different Approach of Analysis Using Multiple Datasets

粒子物理实验中数据统计量影响着物理量的测量精度和显著性的大小.为了尽可能的扩大统计量,减小测量中的统计误差,提高显著性,同时采用多个数据样本进行同一分析一直是一个有效的方法.在采用多个不同样本时,通常有两种处理方式.第一种是直接将不同的数据样本累加,然后作为一个总样本进行分析.第二种是不进行样本累加,而是在分析中采用联合拟合的方式同时提取出各个样本中信号的相关信息和显著性.本文对两种方法中共振态显著性的差异进行了研究和比较,研究结果可供使用多个数据样本开展的相关分析参考.

Statistics is important in experiment of particle physics.It will affect not only statistical uncertainty but also significance.To improve sensitive and increase significance,a common way is to combine multiple datasets to enlarge the statistics in the analysis.There are two frequently used approaches.The first one is that the datasets are added together as one whole dataset,then the analysis is performed on this unique dataset.The second approach doesn’t need to construct a whole dataset,instead,a simultaneous fit will be applied to all the datasets,and information of the signal will be extracted in this simultaneous fit.In this paper,the difference of resonance significance in these two different approaches are studied.The research and conclusion can be used as a reference for future analysis using multiple datasets.