Section Threeforecasting the behavior of industries or specific companies.^39
Thus, although the value of macro-level fore-casting may have
declined, more micro-oriented applications, such as human
resource forecast ing, currently have greater credibility
and value.
Other Considerations
The sheer number of forecasting techniques also can pose
problems for those who must make decisions on which
techniques to use. Several factors should be considered in
making these decisions, including time horizon, level of
technical or mathematical sophistication required of the
forecaster, cost, whether appropriate data are available, the
stability of the data on which forecasts will be based, and level
of detailed data required. Other factors include accuracy,
whether the technique is effective at identifying turning points,
whether the forecast output includes a measure of central
tendency, and an indication of the breadth of potential
outcomes. One precaution regarding forecasting horizons is
that the desire to extend the forecast far out into the future
should be tempered with a recognition that with long-range
time horizons, the process will become more complex and time
consuming. The costs may also increase disproportionately to
the forecast’s value. A general recommendation for improving
forecasting, which appears to have a great deal of validity, is to