Abstract
Learning by doing (or learning curves) is a well-known law in economics and psychology, but no consensus has been achieved on the “qualified” models for more than a century. This article explores the expression of learning by doing in a way where the expression is not involved with changing the prime factors of a learning process. If one prime factor changes dramatically during the course of a learning process, the result of the regression is actually an approach to link two different learning curves. If the two curves are distinguishable, they each obey the law of learning by doing, which will progress rapidly at the initial phase and gradually slow down to a flat end. This article presents two functions as the law of learning by doing: The general exponential model is 0.79:0.21 better than the exponential delay model, whereas the later has the ability to investigate the change of loading factors. This ability makes the models a powerful tool for entrepreneurs and managers in investment and production planning. Key words: Learning by doing, function expression, single equation models, firm behavior, empirical analysis.
Highlights
The law of the learning curves has become ubiquitous since its introduction in an 1885 study of individuals in psychology, and has been found in the manufacturing process of industrial organizations, called ―learning by doing‖ or ―organizational learning‖ in the field of economics and management (Wright, 1936)
This study aims to identify reasonable and qualified models of ―learning by doing‖ by illuminating the mechanism of the learning process
This study describes the mechanism of the plateau and the reason why previous studies could not locate a plateau
Summary
The law of the learning curves has become ubiquitous since its introduction in an 1885 study of individuals in psychology, and has been found in the manufacturing process of industrial organizations, called ―learning by doing‖ or ―organizational learning‖ in the field of economics and management (Wright, 1936). Arrow (1962) seminal work was based on the log-linear function y = pxr of Wright (1936), who found that when the output doubled, the cost of airplane declined along the power sequence infinitely to 0. This finding was and is counterintuitive, that the cost of airplane should be much higher than 0, a vast body of empirical evidence supports that the end of learning process is ambiguous.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
