The Effect of Feedback on Identity Shift in Computer-Mediated Communication

Abstract

The hyperpersonal model of computer-mediated communication (CMC) suggests manners by which online communication transforms relational communication and self-perception. Criticism of the model includes concerns over the linkage among its four theoretical components. Recent research on identity shift in CMC suggests that senders' online selective self-presentation provides sufficient dynamics to modify individuals' personality following an online identity performance. The present research extends these findings by examining effects on identity shift due to the influence of feedback to an individual following a self-presentation that deliberately emphasizes a specific personality characteristic. Results support hypothesized interaction effects and illuminate the personality-modification effects of feedback on someone in a public blog setting, as well as feedback in a private expressive condition. Major implications for the status of the hyperpersonal model are presented, with additional questions regarding the computers-as-social-actors effect.

Notes

1. These induction procedures mirror Gonzales and Hancock's (2008) methods (which appropriated aspects of Tice's, 1992, offline identity shift research), with one important difference. Both Tice (1992) and Gonzales and Hancock (2008) induced publicness using two simultaneous strategies: by making salient the presence of a graduate researcher within earshot of the subject, and by having subjects identify themselves by name, major, hometown, age, and dormitory name when they began their recordings. In the private condition, subjects were provided identification numbers and instructed that they would not be identified, and that no one would witness them recording their responses (see also Kelly & Rodriguez, 2006). Tice (1992) explicitly asserts that public behavior is specifically that which can be linked to a person's identity, and private behavior is that which cannot. However, Gonzales and Hancock's (2008) argued that a blog engenders the activation of public commitment to an identity whereas a private electronic document does not and, therefore, selectively self-presenting in blogs stimulated identity shift because of the public nature of the medium. In that case, having the public/blog participants identify themselves, while keeping word-processing subjects anonymous, provides a potential confound: It is not discernable whether the blog versus private word-processor media conditions caused the differences in perceived publicness and identity shift, or whether differences arose due to the self-identification procedure that was embedded in only the public condition. The present research did not employ the self-introduction elements used in previous studies, relying on only the blog versus private document to differential perceptions of public versus private media.

2. Furr (2004) defines r _contrast as reflecting “the unique association between the contrast and that part of the outcome variable that is unrelated to other known sources of variance”; it partials out “variability in the outcome that is associated with any possible contrasts other than the give contrast” (p. 11).

3. Additional contrast tests included the sets of weights depicted in the first three rows of Table 1, representing the following effects on self-reported extraversion: First, contrast weights reflected the main effect of introverted/extraverted self-presentation alone, t(204) = 3.12, p = .001 (one-tailed), r _contrast = .214. A second test reflected only the Feedback/No Feedback × Introversion/Extraversion factors, without consideration of public/private, t(204) = 3.23, p < .001 (one-tailed), r _contrast = .221. A third test examined only the Public/Private × Introversion/Extraversion factors, without consideration of feedback, t(204) = 1.61, p = .054 (one-tailed), r _contrast = .112. The most disparate of these effect sizes do not differ significantly from each other, z = 1.01 (see Howell, 2007).