![]() | (33.1) |
is given by![]() | (33.2) |
enter the objective function on the right-hand side of
Equation (33.1) after squaring, the effects of outliers are magnified accordingly.
of the residuals:![]() | (33.3) |
is a measure of the scale of the residuals,
is an arbitrary positive tuning constant associated with the function, and where
are individual weights that are generally set to 1, but may be set to:![]() | (33.4) |
(Andrews, Bisquare, Cauchy, Fair, Huber-Bisquare, Logistic, Median, Talworth, Welsch) are outlined below along with the default values of the tuning constants:Name | ![]() | Default ![]() |
Andrews | ![]() | 1.339 |
Bisquare | ![]() | 4.685 |
Cauchy | ![]() | 2.385 |
Fairl | ![]() | 1.4 |
Huber | ![]() | 1.345 |
Logistic | ![]() | 1.205 |
Median | ![]() | 0.01 |
Talworth | ![]() | 2.796 |
Welsch | ![]() | 2.985 |
is known, then the
-vector of coefficient estimates
may be found using standard iterative techniques for solving the
nonlinear first-order equations:![]() | (33.5) |
, where
, the derivative of the
function, and
is the value of the j-th regressor for observation
.
is not known, a sequential procedure is used that alternates between: (1) computing updated estimates of the scale
given coefficient estimates
, and (2) using iterative methods to find the
that solves
Equation (33.5) for a given
. The initial
are obtained from ordinary least squares. The initial coefficients are used to compute a scale estimate,
, and from that are formed new coefficient estimates
, followed by a new scale estimate
, and so on until convergence is reached.
, the updated scale
is estimated using one of three different methods: Mean Absolute Deviation – Zero Centered (MADZERO), Median Absolute Deviation – Median Centered (MADMED), or Huber Scaling:MADZERO | ![]() |
MADMED | ![]() |
Huber | ![]() where ![]() |
are the residuals associated with
and where the initial scale required for the Huber method is estimated by:![]() | (33.6) |
statistic as
is the M-estimate from the constant-only specification.
is calculated as:![]() | (33.7) |
statistic, and provide simulation results showing
to be a better measure of fit than the robust
outlined above. The
statistic is defined as![]() | (33.8) |
is the function of the residual value and ![]() | (33.9) |
, an adjusted value of
may be calculated from the unadjusted statistic![]() | (33.10) |
statistic is a robust version of a Wald test of the hypothesis that all of the coefficients are equal to zero. It is calculated using the standard Wald test quadratic form:![]() | (33.11) |
are the
non-intercept robust coefficient estimates and
is the corresponding estimated covariance. Under the null hypothesis that all of the coefficients are equal to zero, the
statistic is asymptotically distributed as a
.![]() | (33.12) |
), and a corresponding robust Schwarz Information Criterion (
):![]() | (33.13) |
is the derivative of
as outlined in Holland and Welsch (1977). See Ronchetti (1985) for details.
that provide the smallest estimate of the scale
such that:![]() | (33.14) |
with tuning constant
, where
is taken to be
with
the standard normal. The breakdown value
for this estimator is
.![]() | (33.15) |
using the Median Absolute Deviation, Zero Centered (MADZERO) method.
affects the objective function through
and
.
is typically chosen to achieve a desired breakdown value. EViews defaults to a
value of 1.5476 implying a breakdown value of 0.5. Other notable values for
(with associated
) are: ![]() | ![]() |
5.1824 | 0.10 |
4.0963 | 0.15 |
3.4207 | 0.20 |
2.9370 | 0.25 |
2.5608 | 0.30 |
1.9880 | 0.40 |
1.5476 | 0.50 |
from the data and compute the least squares regression to obtain a
. By default
is set equal to
, the number of regressors. (Note that with the default
, the regression will produce an exact fit for the subsample.)
refinements to the initial coefficient estimates using a variant of M-estimation which takes a single step toward the solution of
Equation (33.5) at every
update. These modified M-estimate refinements employ the Bisquare function
with tuning parameter and scale estimator![]() | (33.16) |
is the previous iteration's estimate of the scale and
is the breakdown value defined earlier.
is obtained using MADZERO
using MADZERO, and produce a final estimate of
by iterating
Equation (33.16) (with
in place of
) to convergence or until
.
times. The best (smallest)
scale estimates are refined using M-estimation as in Step 2 with
(or until convergence). The smallest scale from those refined scales is the final estimate of
, and the final coefficient estimates are the corresponding estimates of
.
for S-estimation is given by:![]() | (33.17) |
is the estimate of the scale from the final estimation, and
is an estimate of the scale from S-estimation with only a constant as a regressor.![]() | (33.18) |
statistic is identical to the one computed for M-estimation. See
“Rn-squared Statistic” for discussion.Type I (default) | ![]() |
Type II | ![]() |
Type III | ![]() |
![]() | (33.19) |
and
is the value of the j-th regressor for observation
.