Using influence function matrix as outlier detecting tool based on pooled serial correlation coefficients

In this paper, we incorporated auto-correlation function (ACF), partial auto-correlation function (PACF) and inverse auto-correlation function (IACF) into the influence function as a graphical tool for detecting outliers. Depending on the number of positive and negative values of the influence function based on critical values obtained for different lags an observation is identify as outlier. Both the simulated data and Botswana meat sales data confirms the efficacy of using the pooled correlation coefficients in influence function matrix as outlier detection device.