The Finnish Journal of Business Economics 1/94, 36-49.
In this paper it is our purpose to revisit the least absolute
deviation estimation in regression analysis, consider some of its
theoretical properties, and consider its implementation from a
computational mathematical programming point of view. We also
consider goodness of fit statistics as well as approximate
distributions of the associate test statistics for the parameters.
Furthermore, we suggest a new goodness of fit statistic, called the
LAD-coefficient of determination, that is adapted to the metrics
used in LAD-estimation. Finally, some examples are provided to
illustrate the behavior of the procedures in data that include
outliers.
Seppo Pynnönen, PhD
Unversity of Vaasa, P.O.Box 700, FI-65101 Vaasa, Finland
e-mail; Internet: sjp(a)uwasa.fi
Timo Salmi
Professor of accounting and business finance
University of Vaasa, P.O.Box 700, FI-65101 Vaasa, Finland
e-mail; Internet: ts(a)uwasa.fi