manufacturing organizations gain experience in producing a new product, their
production cost and production time per unit decrease. The magnitudes of the
reductions are often predictable from a mathematical model (sometimes called an
"experience curve" or a "learning curve"), and the predictions are frequently used in
planning. Recent work (Epple, Argote, and Devadas, this issue) demonstrates how
nonsimple learning-curve models can be used to investigate possible explanations of
organizational learning. This is an important use of such models, as empirical studies
(cf. Joskow and Rose 1985, and Zimmerman 1982) make clear that a number of
contingency variables may be required for an accurate explanation to be achieved.
1.2.6. Evaluation of the literature on learning from experience. A desirable feature
of the literature on learning from experience is that the studies employ multiple
methods (e.g., laboratory experiments, mathematical analysis, computer simulations,
and retrospective analysis of organizational events). Another desirable characteristic
is that the literature is replete with fresh insights. Some observations about the
difficulties encountered in organizational learning from experience, especially when
the learning is unintentional or unsystematic, are shown in Table 1.