TTs021216_585991_3 - If the data is not normalized or if there is presence of outliers then correlation Spearman Rank is a non-parametric test that can

# TTs021216_585991_3 - If the data is not normalized or if...

• jaimahavira2
• 4

This preview shows page 1 - 3 out of 4 pages.

If the data is not normalized, or if there is presence of outliers, then correlation Spearman Rank is a non-parametric test that can be used in place of the correlation coefficient of Pearson. The Spearman Rank Correlation (also known as Spearman rho) is the Pearson correlation coefficient in the series of data. I want to find the Spearman Rank Correlation between cigarettes per day (X as the independent variable) and cotinine (y as the dependent variable). The data for the same is given below. x (cigarettes per day) y(cotinine ) 60 179 10 283 4 75.6 15 174 10 209 1 9.51 20 350 8 1.85 7 43.4 10 25.1 10 408 20 344 Consider the scatter plot between given below. 0 10 20 30 40 50 60 70 - 200.00 400.00 600.00 f(x) = 2.48x + 139.06 R² = 0.07 scatterplot x (cigarettes per day) y(cotinine)
Clearly there is presence of outlier and hence non parametric version of Pearson correlation coefficient which is Spearman Rank correlation is recommend to test for the strength of the linear association. The initial step in calculating the Spearman Rank Correlation is to give ranks to both the variables. The data when arranged in ranks in ascending order is given below. In case of tied