Van der Kloot W
Department of Physiology and Biophysics, SUNY, Stony Brook 11794, USA.
J Neurosci Methods. 1996 Apr;65(2):151-5. doi: 10.1016/0165-0270(95)00159-x.
This paper describes some statistical methods for working with data on quantal sizes. Since quantal sizes often do not fit to normal probability distribution functions, statistics based on the normal distribution are inappropriate. Resampling methods can be used to determine confidence limits and to test whether two sets of data differ by chance. Some hypotheses about the nature of quanta are based on apparent peaks and valleys in histograms. Confidence limits can be placed on the bins in the histogram by using the Kolomorogov-Smirnov statistic or by resampling. The confidence limits should assist in the evaluation of the significance of the valleys.
本文描述了一些处理量子大小数据的统计方法。由于量子大小通常不符合正态概率分布函数,基于正态分布的统计方法并不适用。重采样方法可用于确定置信区间,并检验两组数据是否偶然不同。一些关于量子性质的假设基于直方图中明显的峰值和谷值。通过使用柯尔莫哥洛夫-斯米尔诺夫统计量或重采样,可以为直方图中的区间设置置信区间。置信区间应有助于评估谷值的显著性。