Errata: Inference Under Heteroskedasticity and Leveraged Data, Communications in Statistics, Theory and Methods, 36, 1877–1888, 2007

This article refers to:

Inference Under Heteroskedasticity and Leveraged Data

The expression of Ψ₅ given in Sec. 3 is not correct. The correct is

The optimal value of k is correct: k = 0.7. This value has been selected, as explained in the paper, through extensive Monte Carlo experiments. Figures 1 and 2 change accordingly. Finally, the HC2, HC3, and HC4 tests are less liberal than reported in the four tables in Sec. 4. (The results for the OLS, HC0, and HC5 tests are correct.) Table 1 corrects the incorrect null rejection rates. (We omit the 1% significance level and the intermediate level of heteroskedasticity.)

Table 1 Null rejection rates of quasi-t tests based on different standard errors (HC2, HC3, HC4); designs 1, 2, 3, 4; entries are percentages.

n	λ test	10%	5%	10%	5%	n	10%	5%	10%	5%
		1.00		50.05			1.00		50.01
50	HC2	15.36	9.69	31.25	23.97	20	26.79	21.26	73.70	67.54
	HC3	9.83	5.91	18.86	14.04		7.85	6.05	20.85	16.90
	HC4	3.67	2.19	6.37	4.60		0.54	0.41	1.33	1.08
100	HC2	13.11	7.75	22.20	15.71	60	19.27	13.68	24.43	18.95
	HC3	10.43	5.97	16.95	11.60		14.78	9.99	18.73	14.19
	HC4	6.54	3.41	9.12	5.82		7.76	4.97	10.24	7.38
150	HC2	12.09	6.89	18.15	12.31	100	15.35	9.87	17.73	12.25
	HC3	10.38	5.65	14.91	9.60		12.57	7.83	14.41	9.65
	HC4	7.49	3.81	9.38	5.62		8.02	4.62	8.98	5.64
		1.00		50.30			1.00		51.30
47	HC2	11.65	6.38	14.66	9.11	19	17.31	11.72	30.84	20.85
	HC3	9.93	5.19	12.32	7.33		11.83	7.63	15.77	10.68
	HC4	9.68	5.06	11.18	6.54		6.23	3.99	5.83	4.14
94	HC2	10.83	5.69	12.44	7.01	57	13.02	7.59	20.17	13.57
	HC3	9.94	5.09	11.18	6.14		11.10	6.29	16.65	10.65
	HC4	9.82	5.03	10.49	5.66		8.61	4.75	10.63	6.02
141	HC2	10.54	5.39	11.76	6.44	95	11.85	6.59	16.04	10.06
	HC3	9.95	4.99	10.93	5.82		10.72	5.78	14.03	8.47
	HC4	9.89	4.93	10.41	5.50		9.13	4.77	10.52	5.89