MEASUREMENT MODELS FOR CONTINUOUS AND DISCRETE VARIABLES (FROM STRUCTURAL EQUATION MODELS IN THE SOCIAL SCIENCES, 1973, BY ARTHUR S GOLDBERGER AND OTIS D DUNCAN - SEE NCJ-52004)

• NCJ Number: 52007
• Author(s): N W HENRY
• Date Published: 1973
• Length: 17 pages
• Sponsoring Agency: Russell Sage Foundation
• Corporate Author: Seminar Press, Inc
• Publication Format: Document
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America
• Note: REVISION OF A PAPER PRESENTED AT A CONFERENCE COSPONSORED BY THE SOCIAL SCIENCE RESEARCH COUNCIL AND THE SOCIAL SYSTEMS RESEARCH INSTITUTE OF THE UNIVERSITY OF WISCONSIN, MADISON, NOV 12-16, 1970

Annotation

THE PURPOSE OF THIS ARTICLE IS TO SHOW HOW ASSUMPTIONS OF PATH-ANALYTIC MODELS CLOSELY PARALLEL DICHOTOMOUS MODELS, AND BASIC CHARACTERISTICS OF CONTINUOUS-VARIABLE MODELS ARE REVIEWED.

Abstract

THE DISCUSSION ON REPEATED MEASUREMENT OF ONE CONTINUOUS VARIABLE FOCUSES ON THE IDENTIFICATION OF RESTRICTIONS AND VARIANCES. TWO WAYS OF DEVELOPING MEASUREMENT MODELS FOR DICHOTOMOUS ATTRIBUTES ARE EXAMINED: (1) TREAT VARIABLES AS NUMBER VALUED (BINARY OR 'DUMMY') AND (2) DEFINE CONDITIONAL PROBABILITIES OF RESPONSE UNCERTAINTY AND CHANGE UNDER VARIOUS CONDITIONS, LEADING TO A STYLE OF ANALYSIS THAT IS COMPARABLE TO MARKOV CHAIN ANALYSIS. WAYS OF RESTRICTING GENERAL MEASUREMENT MODELS INVOLVING DICHOTOMOUS ATTRIBUTES ARE CONSIDERED, ALONG WITH BINARY VARIABLE ANALYSIS AND POLYTOMOUS QUALITATIVE VARIABLES. APPLICATIONS OF MODELS RELATING LATENT AND MANIFEST VARIABLES ARE DISCUSSED. SUPPORTING EQUATIONS ARE GIVEN. (DEP)