根据LZ所述,应该指的是X和Y的二维正态分布。二维总体的正态性检验主要从以下几个方面进行检验:
1.二维正态分布的边际分布服从一维的正态分布,所以要先检验其边际分布
1)利用散点图(小样本)和直方图(大样本),若出现分布不是呈现中间大两边小,或一边特别长的情况,则不服从正态分布。
2)利用Q-Q图可进一步检验。Q-Q图是指sample quantile VS population quantile,结果应是点基本分布在一条直线上。还可计算相关系数。
Considering fitting a simple linear regression model by following model form: Y=α+βX+ε
Then X and ε (error term) are assumed to have a bivariate normal distribution with zero correlation and variances of Var(X) and Var(ε), respectively.
JingJu
to visualize bivariate distribution:
an example to test the bivariate normality but the normality for each variable (that is what i am always thinking).
Jingju