Some
type of data are more affected by erratic forces than others. Many day-to-day irregularities are not
predictable at all. However, if the
days are accumulated into weeks, and weeks into months, it is possible to
detect trends. Generally, the shorter the
interval of time classification, the higher the tendency there is to observe
erratic fluctuations in the series.
Only
occasionally are irregular fluctuations important enough to be given special
study. The erratic nature of the
fluctuation would have to cause a significant change in the economic series to
justify extensive study. Exceptions
could include the effect of a strike on production, a convention, etc.