Research, part 2

Last week I described some research issues I'd been working on here at the Amsterdam School of Communication Research (ASCoR) pertaining to online self-disclosure, impressions, and liking in computer-mediated communication. These issues prompted my visit here, to learn from Jochen Peter and Patti Valkenburg and their colleagues what insights and approaches we might compare or put together in order to untangle some issues in this research area. That work, as I mentioned, is moving forward with several new studies and some new ideas. But I also mentioned that I'd been probing my hosts on the matter of how and why they do things the way they do in their study of online communication. That has led to some fascinating ideas so far, and opened my eyes to new possibilities in my own work and some possible synthetic ideas.

  

Science pauses for the papparazzi

In terms of how each of us does things—research that is—I tend to do lab experiments, in which we ask a sixth of the participants to pursue goal A (vs. B), while communicating via medium X (vs. Y or Z), to examine how their communication changes and what effect it has on their attitudes and feelings toward their partners. (The other 5/6ths of the participants are allocated to the alternative combinations.)  Part of experimental processes involves being able to rule out competing explanations (other than factors A/B and channels X/Y/Z) for the outcomes we’re interested in. In contrast, although Jochen and Patti do experiments, too, they primarily use survey techniques rather than lab procedures: asking lots of people whether they are greater or lesser on characteristic 1 and 2, whether they communicate with partner type i or ii, on topics of greater or lesser intimacy, how much they use medium X or Y, with what naturally-occurring perceptions they have about the qualities of these media, and how these things affect psychological outcomes.

Why, I recently asked, do you deal with participants’ perceptions of media, rather than use different media in different contexts and see what they do? Sure, the perceptions should follow the situations and actions; but if you just ask people about their perceptions and activities, how do you know if they really know? What situation might they have been thinking about when they answered the survey? Why deal with these interesting but elusive free-floating characteristics?

Answer one, I learned, has to do with what lab experiments often miss (or questionably assume), and answer 2 has to do with how those missed characteristics can reverberate through the communication process.

Patti explained that one thing experiments can miss is that some people, in reality, do not get exposed to every media situation. Some people seek certain media and content, while others avoid it. (That’s answer 1.) The people who seek it may be different than those who do not, and, the way they are different from others may also affect how they react to media messages (and that’s the beginning of answer 2).

In experiments, everyone has an equal chance of being exposed to Condition X, but in the world, a lot of people wouldn’t go near it. And the lab reactions of people who we expose to Condition X, but who would not normally go near Condition X, don’t tell us as much about the people who would and do go near it.

These ideas are depicted and explained in Patti and Jochen’s new article in the Journal of Communication, “The Differential Susceptibility to Media Effects Model.”  It provides a really fascinating re-shuffling of the literature and how we can think about how people deal with media differently than we mostly have done.

Figure 1 from Valkenburg & Peter, 2013, p. 226

So I reflected somewhat cynically an idea that Mike Shapiro pointed out a few years ago: From experimental lab research, I know a lot about what would happen (if the average person had Goal A using Medium X), but not much about what does happen.  Others know a great deal about what does happen, but they can’t explain what causes it (using the most rigid definition of causal explanation, by experimentally controlling participants’ exposure to factors that provide rival explanations).

We’ve started discussing how to bridge these approaches. We’ve begun identifying paradigm cases where researchers took “random” variables – factors that are theoretically defined as perceptions or individual differences which might vary normally but randomly among people – and rendered them “active” (subject to controlled experimentation) for the sake of theory-testing. In a situation such as that, if participants’ perceptions line up with the specific experimental conditions we control, great; that’s a good way to test a theory of perceptions and or through conditions.

What if perceptions do not line up with objective conditions? Now that would be interesting, too. Maybe the perceptions that we theorized to be important don’t actually mediate the processes we imagined they do. Or maybe the conditions that we theorized to cause certain outcomes, actually stimulate different perceptions, and there is some new intervening reaction that makes a difference when people do A or B, in ways we did not contemplate. Or maybe we have not been measuring perceptions, or implementing conditions, as well as we thought we did. Or all of the above. Any of these conclusions suggest more work to do to unwrap the mysteries of how people use media (in or out of relationships) and how media affect them. We have more thinking and writing to do on these ideas, but if we can synthesize a good framework it may enhance our own research, and maybe influence the thinking and research by other students like us.

Take the concept, anonymity. Sometimes researchers conceive of it as not knowing who said what online. Others use it to refer to not being seen visually online. In others’ research, it is the idea that people cannot connect your online statements to your offline self. Could be any, could be all, could be that most online communication is not very anonymous at all by some definitions. Schouten, Valkenburg, and Peter argue that it is the perception of the relevance of anonymity that might be most important. Which of these conceptions affects the relevance of anonymity? What about contexts, such as fostering your SecondLife vs. answering your university email -- don't these contexts differ in the relevance of anonymity? We could test which contexts and/or interfaces affect the perception of anonymity’s relevance, and learn what differences in definitions really count, in terms of users’ perceptions and behavior.

There was another surprise. Just as I was reading about Patti and Jochen’s new model, they were reading a model of interactive media my colleagues and I suggested not long ago, too. It was largely influenced by our study of new media and interpersonal goals, and Prof. Charles Atkin’s communicatory utility theory in the 1970s. Chuck proposed that interpersonal goals lead us to consume media in order to gather content to bring up and discuss in our interpersonal conversations. We extended the argument by saying that different interpersonal goals should affect the way we seek online (mediated) information, what we selectively report to others that we found, what we remember of it in this biased way, and that ultimately our own perceptions of things may change by the goal-driven information-seeking we used media to do.

Patti was the first to discover the similarity among our positions: Precursors to our media information-seeking (social, developmental, or dispositional) affect in which media we seek information and how we seek information from them, and these same factors affect how we process (report, retain, recall) the information. Patti noted, “On p. 187 of your chapter you write: ‘We posit that the specific interpersonal goals that prompt an individual's media consumption shape attention to variations in the content and features of the topical information one consumes, affecting its interpretation and recall.’”  

They had a model of media effects. We had a model of interpersonal interaction in a new media environment. These works have a lot in common. Our task now is to see where these models mesh and where they diverge, and whether the points of divergence can synthesize to expand the model’s reach and frame a rather huge range of communication, in a logical and comprehensive way.

Just as Prof. Peter Neijens and I had discussed in February, modern communication is such that mass media researchers can’t ignore interpersonal dynamics; people use media to tweet, chat, comment, and criticize messages that arrive via the mass media among each other, back and forth, in ways that affect the impact of mass media messages in serious ways. Interpersonal communication researchers often need to consider media more, since relationships initiate on dating sites, maintain themselves through Facebook, and coordinate themselves through texting and email, etc. If media properties affect the way relational messages play out (I think they do), they can't be ignored. But they're complicated. As Laura Stafford wrote, “Media scholars and relational scholars from many domains can inform each other. Insular inspection only serves to constrain our understanding of the increasing complexities of the ways relationships and media interact” (p. 96). 

We’ll see what we can come up with together in Amsterdam and beyond. Only you may not see too much more about it until Reviewer B says okay.

(Can someone put references in blogs? Hope so because in this work you never get too far by yourself.)

• Atkin, C. K. (1973). Instrumental utilities and information seeking. In P. Clark (Ed.), New models for mass communication research (pp. 205-242). Beverly Hills, CA: Sage.
• Schouten, A. P., Valkenburg, P. M., & Peter, J. (2007). Precursors and underlying processes of adolescents’ online self-disclosure: Developing and testing an “Internet-attribute-perception” model. Media Psychology, 10, 292-315.
• Shapiro, M. A. (2002), Generalizability in communication research. Human Communication Research, 28, 491-500.
• Stafford, L. (2005). Maintaining long-distance and cross-residential relationships. Mahwah, NJ: Lawrence Erlbaum Associates.
• Valkenburg, P. M., & Peter, J. (2013). The differential susceptibility to media effects model. Journal of Communication, 63, 221-243.
• Walther, J. B., Tong, S. T., DeAndrea, D. C., Carr, C., & Van Der Heide, B. (2011). A juxtaposition of social influences: Web 2.0 and the interaction of mass, interpersonal, and peer sources online. In Z. Birchmeier, B. Dietz-Uhler, & G. Stasser (Eds.), Strategic uses of social technology: An interactive perspective of social psychology (pp. 172-194). Cambridge, England: Cambridge University Press.