How Demanding is Social Media? Predicting Twitter and Facebook Use (early data)

For the upcoming Eastern Communication Association conference in Cambridge, Mass this April, our West Virginia University communication technology research team (myself, Assistant Professor David Westerman and Doctoral candidate CJ Claus) is pleased to announce that our study "How Demanding is Social Media: Understanding Social Media Diets as a Function of Perceived Costs and Benefits – a Rational Actor Perspective" has been recognized as a Top Three Research Paper by ECA's Communication Technology Interest Group. Here is the paper abstract:

"Using the rational actor perspective (Markus, 1994a) as a guiding frame, this exploratory study examined individuals’ social media diet (i.e., amount, frequency, and duration of use) as a function of task load and expected goal attainment. Surveys were distributed (N = 337) focusing on Facebook and Twitter for informational and relational purposes. Increased task load – conceptualized as a cognitive cost – directly negatively influenced Twitter use but only indirectly influenced Facebook use as a function of perceived benefits. Across conditions, perceived self-efficacy was negatively associated with perceived task load and positively associated with goal attainment, and goal attainment was a significant correlate of increase social media usage. Interpreted, we see that a transparent technology such as Facebook (cf. Clark, 2003) has no cognitive costs associated with its use, while an opaque technology such as Twitter seems to have a salient cognitive cost element. Further, we found that older users of Facebook were more likely to judge the channel as more cognitively demanding and themselves as having lower self-efficacy in using it. Finally, results indicated that for both Facebook and Twitter, males perceived both channels as more cognitively demanding than females. Theoretical and practical explanations and applications for these findings are presented."

We can't share the paper with you just yet (still being prepared for publication consideration) but the main findings are highlighted below.

Our hypotheses:

1. Goal attainment will be positively related to usage of social media 
2. Perceived task load will be negatively related to usage of social media 
3. Self-efficacy will be positively related to perceived benefits of using social media 
4. Self-efficacy will be negatively related to perceived costs of using social media 

The hypothetical research model:

The results for Facebook users:

For Facebook, you'll notice that the presented model is not the same as the hypothetical one. This is because the link between Task Load Index (this was our measure of cognitive demand, developed by Hart & Staveland, 1988) and one's social media diet (their frequency, amount, and duration of social media use; a scale adapted from Kreek, McHugh, Schluger, & Kellogg, 2003) was not significant (.-09), suggesting that there is no association between the two variables - in other words, for Facebook users there does not seem to be a salient cognitive demand for the technology. However, there was a significant association between cognitive demand and goal attainment. It seems that the best predictor of Facebook usage is rather simple - if I can use the program to meet my goals. In other words, Facebook is not really seen as being difficult to use, but it might not be overly-useful for all situations and it is the latter that predicts the program's usage.

...and the results for Twitter users:

For Twitter users, our model holds up just as predicted, with acceptable fit indices: chi-square(2) = 3.13, p = .209, CMIN/df = 1.57, CFI = .982, RMSEA = .059. Unlike Facebook, cognitive demand is seen as a barrier to Twitter usage, while the ability to use the program to attain one's goals (in our study, information- and relationship-based) was a positive predictor of usage. Furthermore, Twitter was seen as significantly more demanding than Facebook overall - though we should not that the mean scores of cognitive demand for both programs were both in the lower quartile of our 21-point measurement instrument (Facebook: M = 5.02, SD = 3.40; Twitter: M = 5.97, SD = 3.90, t(315) = -2.33, p = .020); as well, Facebook was used significantly more by participants in our survey (Facebook: M = 6.80, SD = 2.77; Twitter: M = 4.06, SD = 3.31, t(333) = 8.21, p < .001; 12-point scale).

A note on gender and age:

A rather compelling finding was the association between gender and age as related to social media. An excerpt from our paper:

Across experimental conditions, males also self-reported significantly higher levels of perceived task difficulty at using social media than did females (males: M = 6.12, SD = 3.79; females: M = 4.25, SD = 3.10, t(315) = 4.50, p < .001) and lower social media self-efficacy (males: M = 4.53, SD = .85; females: M = 4.79, SD = .743, t(334) = -2.83, p < .001). Age was also positively associated with perceived task difficulty (r = .215, p = .005) and negatively correlated with social media diet (r = -.197, p = .010); these age findings are particularly relevant given the relative restriction of range in the age variable in our study in which 95 percent of our sample was between the ages of 18 and 22 (minimum = 18, maximum = 28, mode (n = 56, 32%) = 20). Notably as our analyses are separated for Facebook and Twitter usage, we tested the influence of age and sex on our study variables within both conditions. Unexpectedly, we found separate patterns of influence for both variables. In the Facebook sample, age was significantly correlated with perceived task load (r = .215, p = .005) and social media diet (r = -.197, p = .010) and sex was significantly correlated with perceived task load (r = -.184, p = .015). In the Twitter sample, age was significantly correlated with perceived social media self-efficacy (r = -.173, p = .030) and sex was correlated with both self-efficacy (r = .180, p = .023) and perceived task load (r = -.312, p < .001). 

What's your take on all of this? We'll provide our answers at ECA 2012, and maybe we'll see you there?

References:

Clark, A. J. (2003). Natural-born Cyborgs. New York; Oxford University Press.

Hart, S. & Staveland, L. (1988). Development of NASA-TLX (Task Load Index): Results of empirical and theoretical research. In P. Hancock & N. Meshkati (Eds.), Human Mental Workload (pp. 139-183). Amsterdam: North Holland.

Kellogg, S., McHugh, P., Bell, K., Schluger, J., Schluger, R., LaForge, K., Ho, A., & Kreek, M. (2003). The Kreek-McHugh-Schluger-Kellogg scale: a new, rapid method for quantifying substance abuse and its possible applications. Drug & Alcohol Dependence, 69, 137-150.

Markus, M. L. (1994a). Finding a happy medium: Explaining the negative effects of electronic communication on social life at work. ACM Transactions on Information Systems, 12, 119-149. doi:10.1145/196734.196738