Peter H. Schönemann
Professor Emeritus • Department of Psychological Sciences • Purdue University

Abstract 33


Peter H. Schonemann

Factorial definitions of "intelligence": Dubious legacy of dogma in data analysis

In Borg, I. (ed.) Multidimensional data representations: When and why. Ann Arbor: Mathesis Press, 1981,  p.325-374.


  1. Early definitions of "intelligence": Too many answers in search of a question
  2. Spearman: A new question, with a new answer
  3. Thomson: Another answer to the same question
  4. Wilson: Questions some of the answers
  5. Thurstone: Simple answers for most questions
  6. Joreskog: Answers only half the question
  7. Jensen: Crystallized answers to fluid questions


This is an invited talk I delivered at a conference, Multidimensional Data Representations; When and Why, at the Rheinisch-Westfalische Technische Hochschule, Aachen West Germany, in June 1980.  Clyde Coombs and Louis Guttman were among the featured speakers. The paper  summarizes my views at the time of the history of the classical factor model and its problems. Though  technical in places, it was intended to be comprehensible to interested graduate students. It covers Thomson's Sampling Theory, Simple Structure, LISREL, and factor indeterminacy (but not Spearman's hypothesis, of which I was unaware at the time).