Iturriaga R, Alcayaga J, Zapata P
Laboratory of Neurobiology, Catholic University of Chile, Santiago.
Comput Methods Programs Biomed. 1989 May;29(1):37-42. doi: 10.1016/0169-2607(89)90088-6.
Many observations encountered in biological and medical research are randomly distributed in bivariate scales, and thus not susceptible to simple regression analyses. Since such data are depicted by ellipses in scatter diagrams, a computer program to calculate the confidence regions for the means or the total data of bivariate samples was written in BASIC for correlational analyses. The program, based on the principal axes algorithm, plots the calculated confidence regions as an elliptic area, using the fitted equations for its major and minor axes. The program displays the sample parameters required to perform comparisons between different groups of experimental conditions.
在生物学和医学研究中遇到的许多观测值在双变量尺度上呈随机分布，因此不易进行简单的回归分析。由于此类数据在散点图中由椭圆表示，因此用BASIC语言编写了一个计算机程序，用于计算双变量样本均值或全部数据的置信区域，以进行相关分析。该程序基于主轴算法，利用其长轴和短轴的拟合方程，将计算出的置信区域绘制成椭圆区域。该程序显示了在不同实验组条件之间进行比较所需的样本参数。
