In this work, we show that Spearman’s correlation coefficient test about H0:ρs=0 found in most statistical software is theoretically incorrect and performs poorly when bivariate normality assumptions are not met or the sample size small. There common misconception tests ρs=0 robust to deviations from normality. However, under certain scenarios violation of assumption has severe effects on type ...

In this study, the concept of relative membership degree is proposed to investigate grey optimal cluster evaluation. Besides, the correlation coefficient between sample observation value and standard eigenvalue is used to reflect the similarity degree of evaluation objective and each cluster centre. Based on those mentioned above and concept of relative membership degree, an optimized grey clus...

• In the most difficult case where the number of observations n ≡ 3 (mod 4), high Defficiency of certain chemical balance weighing designs under completely symmetric covariance matrix of errors is shown. It is also proved that D-optimal design may depend on the values of the correlation coefficient.

The conventional starch determination (employing precipitation, filtration, re-dissolution, colour development) is compared with a simplified direct colorimetric procedure, which is readily automated. The reproducibility is better than 3%, with a correlation coefficient of 0,87 between the two methods.