To overcome the shortcomings of the logit model, we developed a
heteroscedastic extreme value (HEV) model, a probit model, and a mixed-logit model.
heteroscedastic) so was transformed logarithmically (log) in order to provide a true interpretation (Atkinson and Nevill, 1998; Hopkins, 2000; Hopkins, et al., 2009).
The power decreases with increasing residual error and is almost always above 80% when |RMAX| = 100% in constant error ([[sigma].sub.i] = 25%) and
heteroscedastic error ([[sigma].sub.i] = f([C.sub.i])) scenarios (see Supplemental Material, Figure S3).
Data on hatching times for laboratory experiments were log-transformed to fulfil assumptions of ANOVA; however, variances remained
heteroscedastic. Therefore, Welch's ANOVA was used for those data, in addition to Tamhane's T2 post hoc test to further determine differences among groups.
(5.) In order to take into account that standard errors may be
heteroscedastic, we used a Huber-White sandwich estimator in these equations (StataCorp, 2001).
(6.) The ARCH (Autoregressive Conditionally
Heteroscedastic) models, introduced by Engle (1982), are specific non-linear time series models.
A regression technique accounting for
heteroscedastic and asymmetric errors.
(2006) analyse how outliers affect the identification of conditional heteroscedasticity and estimation of generalized autoregressive conditionally
heteroscedastic (GARCH) models and show that the properties of some conditional homoscedasticity tests can be distorted.
One of the first approaches to this problem was suggested by Cochran (1), who investigated maximum likelihood (ML) estimates for one-way, unbalanced,
heteroscedastic random-effects model.
Dette [8] considered the problem of testing for the equality of two curves with
heteroscedastic errors and fixed design.
Note if the variance parameters A and B are not equal to zero, Case and Shiller (1989) showed the variance of the equation error term in (4) is
heteroscedastic and proposed a feasible Generalized Least Squares (GLS) estimator to correct for this problem.
Since forecast errors are generally
heteroscedastic, we utilize the Newey-West (1987) procedure to estimate the covariance matrix of equations (1)-(3), correcting for both the f-order serial correlation and heteroscedasticity.