Simultaneous Determination Of Adjusted Ranks Of Sample Observations And Their Sums And Products

Abstract: This paper proposes a systematic method for the simultaneous determination of adjusted ranks of sample observations and their sums and products adjusted for possible presence of tied observations in the sampled populations for use in further analyses. When computations involving paired data sets, as in the computation of the Spearman Rank Correlation Coefficient, this procedure intrinsically obtained the sums of ranks, products of ranks and sums of squares of ranks, automatically adjusting these sums for more accurate results. The proposed method is illustrated with some data and used to estimate ties adjusted Spearman Rank Correlation Coefficient and the bias that would have arisen if there were no adjustments for ties in the sampled populations.

REFERENCES

[1] Gibbons J.D. (1973). Nonparametric Statistical Inference. McGrawhill Book Company, New York.

[2] Hollander, M and Wolfe D.A. (1999). Nonparametric Statistical Methods (2nd Edition). Wiley � Inter Science ,New York.

[3] Oyeka C.A., Osuji, GA and Nwankwo, C.C. (2013) �Simplified Method of Rank Determination in Rank Order Statistics� Mathematical Theory and Modeling Vol 3, No 14.

[4] Oyeka Ikewelugo Cyprian Anene, Nwankwo Chike H, Awopeju K. Abidemi (2014). The G- Method for Rank Determination in Rank Order Statistics for �C Samples. American Journals of Theoretical and Applied Statistics. Vol 3, No1, pp 18-24.

[5] Siegel Sidney (1956): �Nonparametric Statistics For the Behavioural Sciences�. International Students Edition. McGraw � Hill, KOGAKUSHA Ltd.