A Comparison of Autometrics and Penalization Techniques under Various Error Distributions: Evidence from Monte Carlo Simulation

Table 3

Simulation results for Frechet distributed errors.

	Methods	0.25		0.90
		Potency	Gauge	Potency	Gauge
{100, 20}	Autometrics	0.995	0.01	0.719	0.032
	SCAD	0.999	0.155	0.965	0.499
	MCP	0.999	0.12	0.964	0.498
{100, 25}	Autometrics	0.996	0.011	0.731	0.031
	SCAD	0.999	0.115	0.937	0.438
	MCP	0.999	0.079	0.94	0.441
{100, 30}	Autometrics	0.995	0.011	0.734	0.050
	SCAD	1	0.089	0.918	0.374
	MCP	1	0.054	0.916	0.379
{300, 20}	Autometrics	0.999	0.01	0.983	0.009
	SCAD	1	0.091	0.999	0.179
	MCP	1	0.078	0.998	0.18
{300, 25}	Autometrics	1	0.010	0.987	0.010
	SCAD	1	0.063	0.997	0.17
	MCP	1	0.05	0.997	0.168
{300, 30}	Autometrics	1	0.011	0.985	0.012
	SCAD	1	0.045	0.995	0.153
	MCP	1	0.032	0.995	0.141