Four essays on IT-mediated phenomena: IT knowledge management, mobile telecommunication, and e-commerce

This dissertation studies three IT-mediated phenomena: (1) learning behavior of IT knowledge workers, (2) users' consumption behavior of wireless communication services, and (3) the evolution of product-level uncertainty in online shopping and its role in determining consumer's online shopping behavior. It consists of four essays all of which employ a combination of economic modeling and econometric analysis.;In the first essay entitled "The Learning Curve of IT Knowledge Workers in A Computing Call Center," I study the organizational learning curves in IT technical support services. Using data from a computing call center at a university, I estimate group/organizational learning curves using a combination of traditional learning models and duration models. The work also introduces (1) a new knowledge classification (surface-level application knowledge vs. in-depth technical knowledge) and (2) the classification of IT problem types. In addition to demonstrating that learning rate in IT technical support is 7.1% (compared to 20% observed in manufacturing), the study explores (a) the evolution of the variance of resolution times as a performance measure, (b) the moderating effect of both knowledge types and IT problem types on the learning progress, and (c) the knowledge transfer across IT problem types.;The second essay entitled "Empirical Analysis of Mobile Voice and SMS service: A Structural Model" empirically examines the pricing effects on wireless telecommunication service demand. I, first, develop a structural model to understand mobile users' behavior in individual consumption of voice and Short Message Service (SMS) services under one-way and 'step' nonlinear pricing. Second, I estimate the structural parameters specified in the model with consumption data on subscribers to a cellular service provider in an Asian country. I find (a) that SMS and voice services are small substitutes, (b) an asymmetric pattern in cross-price elasticities, (3) the own-price elasticity of voice services is small compared to that of fixed phone service, (4) younger users' demand is far more inelastic than that of older users. Finally, I conduct policy experiments that capture the effects of change in the strategic pricing scheme.;In the third essay titled "On Product-level Uncertainty and Online Purchase Behavior: An Empirical Analysis," I (1) develop theories to describe how an individual consumer's product-level uncertainty is a function of their online shopping experience and the information communication methods and (2) build an analytical model to derive research hypotheses. Using individual-level transaction data from a large South Korean online retailer, the study shows that (a) as online shopping experience increases, consumers are likely to purchase more intangible products. However, this is contingent on product price, (b) consumers tend on average to purchase cheaper product lines as they accumulate online shopping experience and (c) online vendors can induce the purchase of intangible products by taking advantage of retailer reputation in the physical world and through the use of digitized video commercials.;The fourth essay entitled "Trajectory-based Consumer Segmentation and Product Positioning in the Online Markets" aims to understand the distinctive longitudinal online shopping pattern and further suggest how to use this better understanding. I, first, identify clusters of individuals following similar progressions of shopping behavior on three dimensions: (I) product intangibility, (2) product price, and (3) the number of transactions in a unit of time. I find interestingly that estimated trajectories do not overlap at any time in all three behaviors. Second, using a Bayesian approach and knowledge acquired from trajectory analysis, I develop the framework of an optimal product positioning strategy that predicts the relative probability that an individual customer is interested in a particular good considering multiple product attributes.