Amended on 22 May 2017 

It is the most important issue that the observed values to handle with are quantitative and reliable such as derived from HPLC-photometric method.   Non-quantitative methods, such as ELISA method, PCR method, direct photometric method, and MS method, etc., give usually non-conclusive statistical results.    

Since the distribution patterns in biology and biochemistry have shown no Gaussian distribution (Gauß-Verteilung), a non-parametric method of Mann-Whitney’s U test has to be performed.    It has been concluded that P≦0.05 (one-tailed test; n1 = 3, n2 = 3; all the three values in one group was considered to have the same rank) between the two groups was significant.   Statistical significance has been calculated by using the Mann-Whitney’s U test, and has been found to be significant (P = 0.05, one-tailed test: n1=3, n2=3) (Please see file; Rat DEN Np-Fuco).   I have still been considering that P = 0.05 (one-tailed test) is really significant, since such a significant phenomenon can not be usually obtainable; i.e., another mathematical consideration and calculation using Bernoulli trials should also be performed.

If two median values of two groups are compared, then calculation of P value using all values of the two groups is performed by using Mann-Whitney's U test with a commercial software such as Statflex Lite, Version 2 (J. I. P. Co., Tokyo, Japan) (please see file; Dr. Terentyeva Urine BIN). 

Such expression has been used in the file "JCB fucoidan active transport" as "Azide inhibited the transport activity by 62.9% (significance; p < 0.05; Mann–Whitney’s U test; two-tailed test).  Median values (n = 6) were indicated. Ranges of control were 22–99 for 1 h, 59–91 for2 h, 77–87 for 3 h, and 90–96 for 6 h, respectively. Ranges of azide-containing sample were 10–47 for 1 h, 24–49 for 2 h, 39–60 for 3 h, and 30–67 for 6 h, respectively."  

If the effect of drug or foodstuff before and after treatment is to be obtained; i.e.. the values of before and after treatment are obtainable from one same person.   Then, Wilcoxon's test is used (please see file; CA agg Dr. De Felice, in this paper representstive example of not significant is presented expressed as follows "No statistically significant difference was observed between the esterase activities of aggregated bCA and bCA monomer (Wilcoxon’s test, P > 0.05).") (this calculation is also possible by using commercial software such as Statflex Lite, Version 2).