Issue No. 03 - May/June (1992 vol. 9)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/52.136161
<p>Part of adopting an industrial process is to go through a learning curve that measures the rate at which the average unit cost of production decreases as the cumulative amount produced increases. It is argued that organizations buy integrated CASE tools only to leave them on the shelf because they misinterpret the learning curve and its effect on productivity. It is shown that learning-curve models can quantitatively document the productivity effect of integrated CASE tools by factoring out the learning costs so that managers can use model results to estimate future projects with greater accuracy. Without this depth of understanding, managers are likely to make less-than-optimal decisions about integrated CASE and may abandon the technology too soon. The influence of learning curves on CASE tools and the adaptation of learning-curve models to integrate CASE are discussed. The three biggest tasks in the implementation of learning-curves in integrated CASE settings, locating a suitable data site, collecting the data, and validating the results, are also discussed.</p>
data collection; data validation; software engineering; learning curve; integrated CASE tools; behavioural sciences; human factors; software engineering; software tools
C. F. Kemerer, "How the Learning Curve Affects CASE Tool Adoption," in IEEE Software, vol. 9, no. , pp. 23-28, 1992.