| Author(s) | Moawad El-Fallah Abd El-Salam |
| Affiliation | Department of Psychology, College of Social Sciences, Al-Imam Muhammad Ibn Saud Islamic University, Riyadh, Saudi Arabia |
| Title | A Robust Measure for the Correlation Coefficient |
| Source | Journal of King Saud University. Administrative Sciences. Volume 19, No 1. (2007/1427) |
| Abstract | In this paper, we investigate the robustness of some well known correlation coefficients, namely, Pearson’s, Spearman’s and Kendall’s. The empirical evidence shows that these correlation coefficients are sufficiently non-robust against outliers. That is, they do not have high breakdown points. As an alternative, a robust estimator for the correlation coefficient is proposed. This estimator is based on the least median of squares. It is shown that this correlation coefficient has a higher breakdown point than the well known correlation coefficients. Keywords: Correlation coefficient, Outliers, Robustness, Least median of squares, High breakdown point. |
