Behavior Research Methods, Instruments, & Computers 2000, 32 (1), 33-46 The study of emotionalprocesses in communication: I. Measuring emotionalization in everyday face-to-face communicative interaction MARIE-LOUISE KAsERMANN, ANDREAS ALTORFER, KLAUS FOPPA, STEFANJOSSEN, and HEINRICH ZIMMERMANN University of Bern, Bern, Switzerland The drawbacks of traditional research into emotional processes have led us to develop a set of methodologies for investigating them in everyday face-to-face communication. The conceptual basis of these procedures is a model of the eliciting conditions of emotional processes as well as a concep- tualization of the emotional processes themselves. On the basis of the assumption of conversation as a rule-governed process, one can describe its default temporal, formal, and functional features, for which we use the MASEDITand SEQ programs, and the minimal model of communicative exchange, respectively. Violations of these default rules can be identified as unexpected/temporally unpredictable events eliciting emotionalization. The nature of emotionalization is determined by the psychological principle of "standard and deviation." Its investigation under natural conditions requires the following: Anoninvasive method of data acquisition (including procedures for rejecting faulty or missing values), measurement (high-resolution recording of physiological, psychomotor, and vocal variables), and the (nonstatistical) construction of an inventory or "relevant effects" (contrastive and template analysis). Finally, we depict three routes of investigating time courses of activation changes as dependent and in- dependent variables and as a target of modification and reflection. The Problem of Investigating Emotional Processes in Social and Communicative Interaction Emotional processes, including unspecific negative dis- tress and positive eustress, form a pervasive and impor- tant aspect of everyday life. Most of these processes may be elicited by features of social and communicative con- texts (e.g., Scherer & Tannenbaum, 1986; Scherer, Wall- bott, Matsumoto, & Kudoh, 1988), and their ongoing course and outcome are critical for healthy individuals' local functioning and global well-being. This can be in- ferred, for instance, from psychoneuroimmunological evidence (e.g., Glaser et aI., 1987; Kiecolt-Glaser & Glaser, 1991). But it also holds true for psychosomatic phenomena (Pennebaker, Hughes, & O'Heeron, 1987) and some psychopathological phenomena. For example, schizophrenicindividuals' rate of relapse into acute phases of illness IF:aly depends on characteristics of the familial style of conversation (Doane, West, Goldstein, Rodnick, & Jones; 1981; Vaughn & Leff, 1976). There, relatives' emotionally overinvolved or hostile remarks function as stressors, Unfavorably combined and occurring at a crit- ical'frequency, they form the communicative antecedents Parts ofthis research were funded by Grants 1114-043635.95/1 and 1113-051050.97/1 from the Swiss National Science Foundation. The authors thank Nicole Gilgen for her assistance with transcription, MAS EDIT, and SEQ analyses. Correspondence should be addressed to M.-L. Kasermann, University of Bern, Psychiatric Institutions, Depart- ment of Psychiatric Neurophysiology, Bolligenstrasse Ill, CH-3000 Bern 60, Switzerland (e-mail: mkaesermann@puk.unibe.ch). 33 of distress in the schizophrenic index person. Further- more, the way in which the schizophrenic and/or his/her partner cope with such a stress reaction may be instru- mental to aggravating or to ameliorating a given condi- tion (Kasermann, Altorfer, & Jossen, 1998). The same, although perhaps in a less obvious fashion, holds true for communicative relationships between healthy interactants (e.g., Kasermann, 1995; Kasermann & Altorfer, 1991; Levenson & Gottman, 1985). lt is uncontroversial then that features of the actual so- cial and communicative exchange affect the optimal (emotional) functioning of an individual; therefore, they should be scrutinized closely. However, there is one very intriguing problem with most research into human emo- tion (Kaserrnann, 1995): Analyses of everyday emo- tional processes in real time are rare in psychology (for exceptions, see Epstein & Fenz, 1965, and Schedlowski et aI., 1993). Furthermore, as a rule, the direct measures of, for example, autonomic activity, that they use do not represent the course of emotional processes, since it is tied to specific features of social or communicative inter- action. Rather, they report average-based indices for var- ious types of situations (e.g., Gottman & Levenson, 1992; Levenson, Carstensen, & Gottmann, 1994; Levenson & Gottman, 1983, 1985). Other approaches mostly use in- direct measures, allowing only for inferences about fea- tures of the relevant situations. Evidence in schizophren- ics cited above, for instance, stems from correlation studies concerned with the co-occurrence between indirectly iden- tified stressors (assessed with the Camberwell Family In- Copyright 2000 Psychonomic Society, Inc.