When Online Meets Offline: The Effect of Modality Switching on Relational Communication

Abstract

Collaborative partnerships developed via text-based computer-mediated communication (CMC) commonly shift interactions to alternative formats. Extant research indicates that shifting from one modality to another, or “modality switching,” can have profound positive and negative effects on relational outcomes. Drawing on social presence theory (Short, Williams, & Christie, 1976) and social information processing theory (SIPT; Walther, 1992, 1996), the present study examined the influence of meeting FtF after varying lengths of time interacting via CMC on relational communication. Consistent with predictions, remaining online yielded greater intimacy and social attraction than the other conditions in which FtF contact occurred. With respect to the CMC conditions, modality switching modestly enhanced relational outcomes in the “early” switching partnerships but more strongly dampened those of “late” switching ones.

Keywords:

• Computer-Mediated Communication
• Relational Communication
• Social Information Processing Theory
• Modality Switching

A previous version of this paper was presented at the 89th Annual Meeting of the National Communication Association held November 19–23, 2003 in Miami, FL.

A previous version of this paper was presented at the 89th Annual Meeting of the National Communication Association held November 19–23, 2003 in Miami, FL.

Acknowledgements

The authors would like to express their gratitude to Frank Boster, Alan Sillars, and the anonymous reviewers whose comments greatly improved this manuscript.

Notes

A previous version of this paper was presented at the 89th Annual Meeting of the National Communication Association held November 19–23, 2003 in Miami, FL.

1. The lottery system described was employed initially to assign participants to the six conditions. However, data collection in the three CMC and FtF-only conditions was completed prior to that in the remaining reverse modality switching conditions (FtF-ES, FtF-LS). As a result, approximately one-half of those participants were assigned in an alternating manner until the cell sizes reported were achieved.

2. Separate preliminary analyses examined the presence of sex differences and dyad composition, respectively, within the overall design. The results revealed significant main effects for participant sex on four outcomes: social orientation, t(862) = 3.99, p<.001, η²=.02; information seeking, t(862) = 7.37, p<.001, η²=.05; predicted outcome value, t(862) = 4.55, p<.001, η²=.02; and uncertainty, t(862) = 4.84, p<.001, η²=.02. Females reported greater social orientation and information seeking, more positive outcome values, and less uncertainty than did males. Significant main effects for dyad composition (cross-sex vs. same-sex) emerged on three outcomes: intimacy, t(430) = 3.22, p<.001, η²=.02; social attraction, t(430) = 3.38, p<.001, η²=.03; and information seeking, t(430) = 4.27, p<.001, η²=.05. Cross-sex dyads reported greater intimacy and information seeking, whereas same-sex dyads reported greater social attraction than their counterparts. The analyses, however, failed to reveal any significant two- or three-way interactions with the other independent variables for either participant sex or dyad composition. Consequently, they were eliminated from further analyses.

3. The attrition rate for each condition was as follows: For the face-to-face only condition, n at Time 1 was 85 dyads, 79 dyads at Time 2, and 75 dyads at Time 3; for the face-to-face early switch condition, n at Time 1 was 82 dyads, 75 dyads at Time 2, and 70 dyads at Time 3; for the face-to-face late switch condition, n at Time 1 was 84 dyads, 77 dyads at Time 2, and 72 dyads at Time 3; for the CMC -only condition, n at Time 1 was 85 dyads, 77 dyads at Time 2, and 74 dyads at Time 3; for the CMC-early switch condition, n at Time 1 was 84 dyads, 78 dyads at Time 2, and 74 dyads at Time 3; for the CMC-late switch condition, n at Time 1 was 83 dyads, 75 dyads at Time 2, and 67 dyads at Time 3. A comparison between the dyads who completed each stage of the study and those that did not failed to reveal any systematic differences on any indicators employed in the present study. Beyond modality switching, approximately one-half of the dyads were informed at the start of the study the manner in which they would be completing all three tasks, whereas the remaining one-half were informed of their mode at the start of each stage. No significant effects emerged (p>.27). As a result, the two conditions were collapsed across modality switching conditions and eliminated from further analyses.

4. Items adapted from the Afifi and Metts’ (1998) uncertainty measure were, “As a result of this interaction, I feel I know him/her” not as well as I thought/much better than I thought; “Regarding my ability to predict his/her future behavior, my partner's communication behavior made me feel” a lot less confident/a lot more confident; “Regarding my ability to predict the type of relationship we should develop, this interaction made me feel” a lot less confident/a lot more confident; “As a result of this interaction, my ability to predict his/her future behavior in the relationship” strongly increased/strongly decreased.

5. Information seeking items were, “I asked my partner questions to get him/her to open up;” “I encouraged him/her to disclose personal information while completing the task;” “I followed up his/her comments with questions;” “I asked him/her more information about what we discussed;” “I did not try to find out more information than what I normally would;” and “I did not actively seek information from my partner while completing the task.”

6. Contrasts utilized to assess each pattern were as follows, unless otherwise noted in the text: For the linear pattern proposed for the CMC-only and “early” modality switch conditions, −1, 0, 1; For the pattern proposed for the “late” modality switch condition, −1, 2, −1 as it was expected to indicated a dampening or mitigation of gains from CMC interaction. In addition, FtF-only and reverse modality switching conditions were probed using the same contrasts employed in examining their corresponding CMC conditions. The contrasts utilized an MS denominator for each dependent variable derived through initial analyses that addressed the interdependence inherent in dyads by treating them as nested within the condition×time interaction. The net effect of this approach is the reduction of Type I errors.

7. Despite the overall patterns reported for each condition for the six outcomes, it is important to note that modality switching, in and of itself, was not perceived as an impediment to task completion. Supplemental data analysis on task-related outcomes to be reported elsewhere indicate tasks completed via face-to-face interaction—irrespective of the timing—were rated significantly more positively (e.g., less effortful, time consuming, pleasant, etc.) than those completed via CMC (p<.20). Among the three CMC conditions examined (CMC-only, ES, and LS conditions), CMC-only was perceived as most effortful and time-consuming. This is in spite of the fact that, anecdotally, participants in the CMC modality switching conditions reported that they were looking forward to meeting their partners in person.

Additional information

Notes on contributors

Artemio Ramirez Jr

Artemio Ramirez, Jr. (PhD, University of Arizona) is an Assistant Professor in the School of Communication, the Ohio State University, Columbus, Ohio

Shuangyue Zhang

Shuangyue Zhang (PhD, the Ohio State University) is an Assistant Professor of Communication at Sam Houston State University in Brownsville, Texas