Popular media has been giving significant coverage to a recently published article on the relationship between narcissism–a personality trait that captures admiration or love of self–and use of social network sites (SNSs). This line of research is not new (see here and here[pdf], for example), but it brings more attention to an ongoing discussion regarding Gen Y, technology, and the “downfall” of modern society. More broadly, it considers long and ongoing questions related to the potential negative impacts of new technology on individuals, groups, and relationships.
While none of the studies I’ve referenced above establish causality, media stories I have read suggest that sites such as Facebook are, in fact, making today’s youth more narcisstic. Take, for example, the headline, “Facebook Feeds Narcissism, Survey Finds” from a CNN article about the recent study. In response to this, I (first) get an exasperated look on my face and (second) direct the authors to the following xkcd strip:
As to my exasperation, first let me comment on a pet peeve of mine. As I have been in media in one form or another now for more than a decade, I understand that catchy titles draw in readers; however, bad reporting of data and not acknowledging limitations of studies often leads to gross exaggeration of research findings. Academic studies are typically written in a way that make them inaccessible (both literally and cognitively) to the average person. This is a major reason we have journalists: to transcribe technical findings into a format that nearly anyone can understand. Due to either a lack of knowledge on the part of the writer, the desire to write a more “interesting” story, or sheer laziness, this process often ends like a game of “telephone,” with the final story only somewhat resembling the original. And while I don’t expect your average newspaper or broadcast news writer to understand the technicalities of a newly published stem-cell study (heck, I would barely understand that), distinguishing between correlation and causation is essential to accurate reporting. Furthermore, it is essential that when reporting significant findings (such as those which could have an impact on policy down the line), writers acknowledge potential limitations to the findings, such as issues of generalizability, small sample sizes, or small effect sizes.
Okay, back to the findings. The most recent study found positive correlations between various components of a user’s profile and narcissism. Unfortunately, the coding of profile components appear to be seriously flawed for a number of reasons. First, coding was only performed by the author, which prevents inter-coder reliability from being established and potentially introduces researcher bias: because the author is also the coder, she could unconsciously have coded content to reflect the findings she wanted. Another concern is that the author provides no details regarding if she created a code book or, assuming she did, how she established criteria for coding. For example, she states that use of positive adjectives such as “nice” and “funny” in the About Me section of the profile were coded as indicators of narcissism, which I find to be a bit of a stretch. Likewise, she coded the use of photo editing software for the profile picture to be an indicator of narcissism, which raises two concerns from me: (1) photo editing software is often not easy to detect, and (2) in cases when it is, it may be serving an artistic purpose rather than an egotistical one. Because the author had no interaction with participants regarding why they made decisions regarding content choices, it is impossible to make assertions as to the reason behind these choices.
A second point related to this and the other studies I have seen on narcissism and SNSs relates to the technology itself. SNSs are centered on sharing information; they are designed with the intent to simplify the process through which users can post content to an audience. Facebook prompts users to post status updates with the query, “what’s on your mind?,” which is a direct request for information about the individual. Therefore, we need to be especially careful in creating operational definitions of what constitutes narcissistic content in order to make sure we are measuring what we set out to measure and not merely capturing standard practices on the site. Furthermore, research should work to establish behavioral norms on the site — what a researcher perceives of as “narcissistic,” such posting a specific type of photo, may in fact be the norm for that given group.
A third consideration to consider when interpreting the results of this and other studies on narcissism and SNSs is the choice of population. Early research on SNSs (and, to be honest, the vast majority of current research as well) tends to employ college student populations, most likely because college students were the primary users of the site and college students are a convenient sample for academic research. College students are, in many ways, in a four-year transition from youth to adulthood, a period that J.J. Arnett refers to as “emerging adulthood.” It is unsurprising that people at this age are self-absorbed — not only are they trying to figure themselves out, but they’re apt to try a variety of ways to fit in, which probably requires a bit of self-promotion.
So is “Generation Me” a more apt name for Gen Y (as at least one book has suggested)? Or have advancements in technology merely made young people’s narcissistic tendencies more public than previously possible? While only research can answer this question, one study provides some initial insights: a recent meta-analysis of studies conducted between 1976 and 2006 found no relationship between cohort and egotism, individualism, self-enhancement, or self-esteem. In other words, kids today are just as caught up in themselves now as they were when their parents were there age; they simply couldn’t share their thoughts on how awesome they were with the rest of the world as easily.
In looking forward, an important next step is to conduct more research with non-college student populations and to identify ways of studying new technology adoption and use over time so as to address questions related to the types of changes in beliefs and behaviors that social media may be effecting. I’ll be presenting research in November at the National Communication Association’s annual conference on how adult users (ages 25-55) negotiate the tensions between using Facebook to obtain social capital benefits and concerns related to making those disclosures. I also hope to employ multiple methodologies (both qualitative and quantitative) in some future studies so as to achieve both breadth and depth of findings. More on that as it develops, but I should probably stop talking so much about myself, lest you think I’m narcissistic.