Determinants of plasma retinol and beta-carotene levels - Statistics Project Example

The property of mean that included magnitude and rank of all data in a set informed the application. Non-parametric test for comparison of median of plasma beta-carotene confirms significance of the difference and establishes reliability. The test analyses the same hypothesis, narrowed down to the following hypothesis, based on median. The graphs show a similar distribution, based on skewedness, and suggest a relationship between the two variables. The suggested relationship is, further, positive because of the symmetry.

Correlation analysis results, however, undermines possible relationship as shown in the following table. The table confirms insignificance of the possible relationship between the two variables (p= 0.205> 0.005, F= 1.612). The t-test result for the regression coefficient also shows the results, based on table 6. The results identify a significant relationship between plasma beta-carotene and vitamin use, a significant relationship between plasma retinol and age and sex, but no significant relationship exist between plasma retinol and plasma