论文部分内容阅读
在普物实验结果的表示结式中,误差项一般只考虑偶然误差,而偶然误差在大量重复测量中将遵守高斯分布。高斯分布可用分布曲线表示,可以从曲线的高低、宽狭大概可以看出数据离散性的好坏,高而狭就较好,低而宽则较差,画图法虽然直观但不精确。一般采用特征量来进行定量的描述:1.用分布曲线的高度h来描述,h称为精密度常数,h大则离散性好。2.用分布曲线的宽度来描述,即用置信度来表示;方法是先确定一个概率值P_r,然后求出与之相应的区间β,使所得的偶然误差中的误差绝对值小于β的概率为P_r。因P_r的取值不同又分为:①标准误差σ:其P_r=0.683。②平均误差
In the generalized experimental results, the error term generally only considers accidental errors, and the occasional error will follow the Gaussian distribution in a large number of repeated measurements. Gaussian distribution curve can be used to express, from the level of the curve can be seen about the width of the discrete data is good or bad, high and narrow is better, low and wide are poor, although the drawing method is intuitive but not accurate. The general use of quantitative characteristics to describe: 1. Distribution curve to describe the height h, h is called the precision constant, h, then the dispersion is good. 2. The width of the distribution curve to describe, that is, to use confidence to express; method is to determine a probability value P_r, and then find the corresponding interval β, so that the resulting error in the absolute value of the error is less than the probability of β P_r. Due to the different values of P_r is divided into: ① standard error σ: P_r = 0.683. ② average error