1、Spearman秩相关系数界值表 nrs0.05rs0.01nrs0.05rs0.0160.8861.000310.3560.45970.7860.929320.3500.45280.7380.881330.3450.44690.7000.833340.3400.439100.6480.794350.3350.433110.6180.755360.3300.427120.5870.727370.3250.421130.5600.703380.3210.415140.5380.679390.3170.410150.5210.654400.3130.405160.5030.635410.3090.

2、400170.4850.615420.3050.395180.4720.600430.3010.391190.4600.584440.2980.386200.4470.570450.2940.382210.4350.556460.2910.378220.4250.544470.2880.374230.4150.532480.2850.370240.4060.521490.2820.366250.3980.511500.2970.363260.3900.501270.3820.491280.3750.483290.3680.475300.3620.467 等级相关又称秩相关（rank corre

3、lation）是一种非参数统计方法，适用于资料不是正态双变量或总体分布未知；数据一端或两端有不确定值的资料或等级资料。等级相关分析的方法有多种，在此仅介绍Spearman等级相关，它是用等级相关系数rs来说明两个变量间相关关系的密切程度与相关方向的。（12.26）式中d为每对观察值所对应的秩次之差；n为对子数。rs值界于-1与1之间，其意义同积差相关系数r。当X或Y中相同秩次较多时，宜用rs的校正值,(12.27)式中TX(或TY)=,t为X(或Y)中相同秩次的个数。当TX=TY=0时，式（12.26）与式(12.27)相同。rs是总体等级相关系数的估计值，存在着抽样误差，故计算出rs后，需作的假设检验。当n50时，可通过查rs界值