Hypothesis Testing for Outlier Effect on the Regression Coefficients Using F Test Statistic

Abstract

Outlying observation may lead to misleading inference plus biased estimate of the parameter and model misspecification among others. It is therefore important to highlight on it verse negative effect in the analysis of dataset. This paper, considered the effect of outlier on ordinary least squares (OLS) based on the test of hypothesis procedure, to test if there is any significant difference in testing the hypothesis of regression parameter for dataset with and without outliers using F test statistic. Four different dataset were considered to proof the set hypothesis. Our results showed that in the presence of outliers, the estimates of regression parameter sign may change which in turn may lead to wrong decision marking.