Each measurement was fitted to an error model and assigned a significance measurement (type I error).