J. Scott Armstrong

Professor J. Scott Armstrong has been a member of the Wharton Marketing Faculty since 1968. In 1989, a University of Maryland study ranked Professor Armstrong among the top 15 marketing professors in the U.S. In 1996, he was selected as one of the first six Honorary Fellows by the International Institute of Forecasters. He is a co-developer of new methods including rule-based forecasting, causal forces for extrapolation, simulated interaction, and structured analogies. In addition to forecasting, Professor Armstrong has published papers on survey research, educational methods, applied statistics, social responsibility, strategic planning, and scientific peer review.

  1. Home
  2. Book Authors
  3. J. Scott Armstrong
Eclectic Research Construct Validation

Chapter 1: Eclectic Research and Construct Validation

Chapter 1 of Models of Buyer Behavior

Eclectic Research and Construct Validation outlines two competing research strategies, eclectic research and intensive research, utilized by social scientists. Studies on income-sales and price-sales relationships offer real-world examples of these techniques, as well as demonstrate the superiority of eclectic research. The text argues that eclectic research provides more in-depth information at a reasonable cost than intensive research, and would thus be more helpful and cost efficient to social scientists.

Predictive Validation of Multivariate Models: Regression vs. Trees

Chapter 3: Predictive Validation of Multivariate Models: Regression vs. Trees

Chapter 3 of Multivariate Methods for Market and Survey Research

The chapter provides a comparative analysis of regression models and tree analysis. The text describes the superiority of tree analysis due to its capabilities with exploratory research, fewer assumptions about data, and with large sample sizes. The criterion for superiority is based on the model’s ability to make predictions on new data.