ABSTRACT
Steven Chaffee stated that ‘the most likely “effect” of communication…is further communication' (p. 72). The present study asks to what extent this offering, referred to as ‘The Chaffee Principle,’ is realized in extant communication research. A manual content analysis of two top-tier communication journals (N = 2,040) offers an assessment of empirical attention to a variety of communicative acts in mass, interpersonal, and mediated-interpersonal communication research, as well as across subfields, contexts, methods, and time. The analysis finds that communication-to-communication effect sizes (r = .28) largely surpass those found in more dominant areas, and yet receive relatively scant attention in the communication literature. A call is made for a more systematic study of The Chaffee Principle to advance communication research.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 The distinction between direct and indirect CCEs dealt with whether the posed relationship predict any intermediary (either causal or noncausal) variables between two CAs.
2 The specific question from the survey was: While definitions of the broader concept of communication are important, for this questionnaire our focus is on how one defines a "communicative act." More specifically, if a student were to ask you "What is a communicative act?", what definition would you offer?
3 The specific question from the survey was: Is this definition derived from any particular source(s)? If so, we would greatly appreciate it if you could provide the last name of the primary author and the title of the article in the field below.
4 The different types of communicative acts were derived from iterative coding by the research team of non-study content used to identify the variety of communicative acts in empirical research.
5 According to Gardner (2001), cell value is significantly different if the adjusted standardized residual in the cell exceeds the Z score required for the Bonferroni-adjusted alpha level calculated by dividing the desired alpha level (i.e. .05) by the number of cells. For example, to estimate the relationship between X and Y, the 2*5 table in this study produced an adjusted alpha of .05/10 = .05, and the Z score at this level was 2.57.
6 When accounting for the directionality of effect sizes, the pooled effect was (r = .21, 95% CI [.24, .31], p = .001).
7 Six studies were excluded from this analysis since they reported on CCEs that were both causal and noncausal.
8 Six studies were excluded from this analysis since they reported on CCEs that were both direct and indirect.