Abstract
The growing ubiquity of mobile telephony able to send e-mail raises new questions, and renews old issues, about the effect of the medium on a message. This article reports experimental results testing effects of user- and medium-generated cues on perceptions of message senders. Grounded in warranting theory, we assess the interaction of low- and high-warrant cues on perceptions of the sender's professionalism, hypothesizing senders of grammatically accurate messages are perceived as more credible. However, we also hypothesize an interaction between grammatical accuracy and the system-generated high-warrant cues from the mobile device used to send the message. Responses from 111 students assessing the credibility of an e-mail sender indicate that, although a message's user-generated content (grammatically accurate vs. erred) influences receiver's perceptions, negative attributions are attenuated by cues reflecting the transmission medium (i.e., a message's mobile signature block). Findings offer theoretical implications for warranting theory. Additionally, results suggest practitioners need to craft a message and indicate the transmission medium strategically to mitigate any impacts on attributions of professionalism to message receivers.
Acknowledgements
The authors would like to thank two anonymous reviewers for their careful and critical reviews, as well as Dave Schrock and his assistance in data collection and Larry Beauchamp his mathematical insight. An earlier version of this work was presented in New Orleans at the 2011 meeting of the National Communication Association.
Notes
1. Convenience samples drawn from higher education are sometimes suspect for their ecological validity (Dipboye & Flanagan, Citation1979; McNemar, Citation1946). Several factors suggest that results can be generalized beyond the present study. First, we tailored the stimuli material to a topic that was highly salient to students to prime their engagement: employability following graduation. Second, although undergraduates with limited work experience may not perceive professionalism the same as those with lengthy work histories, previous work has indicated student samples are reliable measures of the dependent variables operationalized in this research, specifically credibility (Flanagin & Metzger, Citation2000; Rains, Citation2007; Stephens, Houser, & Cowan, Citation2009). Finally, several studies (Gordon, Slade, & Schmitt, Citation1986; Greenberg, Citation1987) have noted that the psychosocial properties of adults in higher education are highly correlated with their peers in the workforce; as this study sought to test general psychosocial principles, findings should generalize to other populations as well.
2. Findings of H1 and H2 are reported using one-tailed significance tests. Levine and Banas (Citation2002) recommended using one-tailed tests when directional hypotheses are developed a priori, clearly reported, and when one-tailed tests are used across the entire study. Yet, the statistical significance of H1 and H2 may be moot due to the inappropriateness of interpreting main effects in the presence of an interaction effect (H3).