Hostname: page-component-7b7567c8c-kwt9d Total loading time: 0 Render date: 2026-11-12T16:59:02.221Z Has data issue: false hasContentIssue false
  • English
  • Français

A Coefficient of Imbalance for Content Analysis

Published online by Cambridge University Press:  01 January 2025

Irving L. Janis and
Raymond H. Fadner
Irving L. Janis
Affiliation:
Special War Policies Unit, Department of Justice
Raymond H. Fadner
Affiliation:
Special War Policies Unit, Department of Justice
Get accessRights & Permissions [Opens in a new window]

Abstract

This article presents a Coefficient of Imbalance applicable to any type of communication that may be classified into favorable content, unfavorable content, neutral content, and non-relevant content. The combined influence of the average presentation of relevant content and the average presentation of total content is reduced to two components, the coefficients of favorable imbalance and of unfavorable imbalance. A precise definition of imbalance is developed and measured against ten criteria.

Information

Type
Original Paper
Information
Psychometrika , Volume 8 , Issue 2 , June 1943, pp. 105 - 119
Copyright
Copyright © 1943 The Psychometric Society

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Article purchase

Temporarily unavailable

Footnotes

*

The authors wish to express their appreciation to Mr. Harold Elsten, Dr. Natan C. Leites, and Mr. Ithiel Pool for valuable suggestions and criticisms.

References

* For a summary of the outstanding attitude-scales, such as those of L. L. Thurstone and of R. Likert, see Gardner and Lois Barclay Murphy and Theodore M. Newcomb, Experimental social psychology, New York: Harpers, 1937, Ch. 13.

** Some of the pioneering studies are the following: Gordon W. Allport and Janet M. Faden, The psychology of newspapers: five tentative laws, Public Opinion Quarterly, 1940, 4, 687-703; Thomas S. Green, Jr., Mr. Cameron and the Ford Hour, Public Opinion Quarterly, 1939, 3. 669-75; Harold D. Lasswell, The World attention survey, Public Opinion Quarterly, 1941, 5, 456-462; David Nelson Rowe, Japanese propaganda in North China, 1937-1938, Public Opinion Quarterly, 1939, 3, 564-80; Douglas Waples and Bernard Berelson, Public Communications and Public Opinion, Chicago: Univ. Chicago Press, 1941; Quincy Wright and C. J. Nelson, American attitudes toward Japan and China, 1937-1938, Public Opinion Quarterly, 1939, 3, 46-62. For a discussion of the methodological problems of content analysis, see N. C. Leites and I. de Sola Pool, On content analysis, Document No. 26, Experimental Division for the Study of War Time Communications, Library of Congress.

† Harold D. Lasswell, A Provisional Classification of Symbol Data, Psychiatry, 1938, 1, 197-204 ;—and Associates, The politically significant content of the press: coding procedures, Journalism Quarterly, 1942, 19:1, 12-23 ;—Communications research and politics, in Print, Radio, and Film in a Democracy (edited by Douglas Waples), Chicago: Univ. Chicago Press, 1942, pp. l18-32;—Amalyzing the content of mass communication: a brief introduction, Document No. 11, EXperimental Division for the Study of War Time Communications, Library of Congress.

‡ Waples and Berelson, ibid. Appendix A; and unpublished manuscripts by N. C. Leites and I. de Sola Pool.

* Cf. p. 117.

* Since “t” is required to be held constant, an increase of neutral content implies an equal decrease in non-relevant content so that the value of “t” remains unchanged. Cf. p. 117 for a fuller discussion of “t”.

** Division provides an index to the attention devoted to the topic or symbol, .

* It is also true that increasing the total content by the factor “a” decreases the Coefficient by the factor “a” because “t” appears as a single factor in the denominator.

** When there is neutral content (n ≠ ,0), the Coefficient varies directly as the ratio of the favorable to the unfavorable-plus-neutral content . This is a reasonable relationship since u + n constitute the “non-favorable” relevant content, (r – f), corresponding to the “non-favorable” or unfavorable content in the simpler ease stated by the criterion.