| This research study focuses on the firm decision maker as the learner while the learning outcome is the updated knowledge base playing an important role in the firm's decision making. The learning process, where the learner obtains additional knowledge, has two phases. In the information acquisition phase, the decision maker acquires internal information by managing the internal data from past experience (for example, learning-by-doing), and/or acquires the external information by communicating with others via social learning activities (e.g., conversation, cooperation, and collaboration). On the other hand, the knowledge updating phase involves the use of a learning mechanism translating the collected information into additional useful knowledge feeding into the existing knowledge base.; This research addresses (i) how the learner (the firm decision maker) learns and seeks to formalize the learning process, (ii) how the decision maker chooses among different knowledge management schemes, and (iii) how social learning behavior reflects on production heterogeneity. This thesis research develops a conceptual model focusing on the definition of knowledge, the different ways of executing learning process, and the way the updated knowledge base influences future decision making. The theoretical model is investigated by using a mathematical model where the decision maker maximizes the firm's profit over time under production and knowledge management constraints. The optimization conditions point out the marginal costs and benefits of learning and guides the decision maker to allocate the physical input and the effort for knowledge management.; Learning strategies, such as always learn, wait to learn, learn-in-bursts , and quit learning, are observed in the deterministic dynamic programming model as the output price changes. Considering the learning decisions under uncertainty, two stochastic dynamic programming models are constructed where the market and technological uncertainties are represented by the stochastic properties of output price and knowledge base accumulations.; The empirical model is used to reveal the connection between social learning and production heterogeneity. A latent class stochastic frontier model (LCSFM) is introduced to estimate the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) India data. (Abstract shortened by UMI.)... |
