Resources allocated to software maintenance constitute a major portion
of the total lifecycle cost of a system and can affect the ability of
a firm to react to dynamic environments. This paper illustrates how t
he human information processing model can be used to develop a strong
theoretical foundation for describing factors affecting software maint
enance. This foundation is then used to synthesize recent empirical fi
ndings and to direct attention towards empirical questions which need
investigation. In addition, the human information processing model is
shown to provide explanatory as well as descriptive power, providing a
basis for managerial action and recommendations aimed at decreasing m
aintenance effort which can decrease maintenance cost and increase the
responsiveness of maintenance to changing organizational needs.