Hypothesis tests and model parameter estimation on data sets with missing correlation information

Lukas Koch

Ideally, all analyses of normally distributed data should include the full covariance information between all data points. In practice, the full covariance matrix between all data…