When are we going to change the learning curve lecture?

Most introductory textbooks in engineering economics, production management, and accounting develop the idea of a learning curve, focusing on the standard power model: y(x) = alpha x to the -beta power. However, a well-known problem with this functional form is that, as output, x, gets large, y(x) g...

Full description

Saved in:
Bibliographic Details
Published inComputers & operations research Vol. 23; no. 5; pp. 509 - 511
Main Author Hurley, W.J.
Format Journal Article
LanguageEnglish
Published Oxford Elsevier Ltd 1996
Elsevier Science
Pergamon Press Inc
Subjects
Applied sciences
Exact sciences and technology
Learning curves
Mathematical models
Miscellaneous
Operational research and scientific management
Operational research. Management science
Operations research
Studies
Production management
Economy
Learning curve
Accounting
Optimization
Online AccessGet full text

Cover

More Information
Summary:Most introductory textbooks in engineering economics, production management, and accounting develop the idea of a learning curve, focusing on the standard power model: y(x) = alpha x to the -beta power. However, a well-known problem with this functional form is that, as output, x, gets large, y(x) goes to zero. So if y measures labor hours, and the production run is large, the power model implies that the labor hours per unit will fall to zero, and for most types of production technologies subject to a learning effect, this is just not the case. Yet this technology remains the centerpiece of pedagogical efforts. It is shown that, in a forecasting context, use of the power function can lead to significant forecast error. Moreover, it is also argued that other, more realistic variations of the standard power curve, such as the De Jong model, are easy to estimate using the nonlinear optimization features available with most spreadsheets.
ISSN:0305-0548
1873-765X
0305-0548
DOI:10.1016/0305-0548(95)00031-3