A New Robust Diagnostic Plot for Classifying Good and Bad High Leverage Points in a Multiple Linear Regression Model

Table 6

The regression estimates, standard errors (in the brackets), and PCE for the second modified Artificial data set.

Variables	Full data	Regular data
		⁢Remove only the suspect BLP; detected by OLS-MD plot; cases		⁢Remove only the suspect BLP; detected by LMS-RMD plot; cases , , , , , , and		⁢Remove only the suspect BLP; detected by MGt-DRGP plot; cases , , and
		Estimate	PCE	Estimate	PCE	Estimate	PCE
	1.026; (1.412)	0.889; (1.292)	13.35; 8.50	0.217; (0.319)	78.85; 77.41	−0.377; (0.196)	136.74; 86.12
	−0.865; (1.457)	−0.934; (1.332)	7.98; 8.58	−0.190; (0.326)	78.03; 77.63	0.421; (0.203)	148.67; 86.07
Constant	1.185; (0.549)	1.243; (0.502)	4.89; 8.56	0.545; (0.080)	54.01; 85.43	0.431; (0.078)	63.63; 85.79
Observation	20	19		13		17
	0.067	0.030	55.22	0.149	122.39	0.264	294.03
Residual SE	1.817 (df = 17)	1.661 (df = 16)	8.59	0.204 (df = 10)	88.77	0.236 (df = 14)	87.01
-statistics	0.614	0.246	59.93	0.874	42.35	2.513	309.28

Note: ; .