I've seen a little bit of discussion in the media about Facebook being used in selection and recruitment settings. Assuming an employer can gain access to a person's Facebook profile, have any scientific studies been conducted to assess the following questions:
I had a read through the article by Kluemper et al (2012) mentioned in the answer by John Pick. The following summarises some key points.
After discussing the broader context of using social networks to measure personality, Kluemper et al (2012) cited the findings of a couple of existing studies:
Karl, Peluchette, and Schlaegel (2010) found that social network users high in conscientiousness were less likely to post problematic content (e.g., substance abuse, sexual content), while Amichai-Hamburger and Vinitzky (2010) found a relationship between self-rated extraversion and number of Facebook friends.
The study by Kluemper et al (2012) involved two studies. Each study involved getting ratings on around 250 participants including self-report Big 5 personality and then getting three raters to read through the participant's Facebook profile and then fill out a measure of Big 5 personality. Self-other correlations for the same personality factors ranged from .16 to .44. This supports the idea that Facebook profiles can be used to get at least an approximate measure of personality. It should also be mentioned that these self-other correlations are a fair bit lower than those obtained by Costa and McCrae (1998) when using the NEO, and a single knowledgeable other-rater (e.g., in the r=.5 to .6 range).
Meta-analyses have generally found a positive but weak correlation between Big 5 personality and job performance (e.g., Barrick & Mount, 1991). In combination with the results from Kluemper et al, this may suggest that personality measures extracted from Facebook profiles might predict job performance. However, given the relatively low self-other correlations, this still requires further empirical testing.
Furthermore, extracting Big 5 personality using three raters reflects only one rating strategy. Thus, there are many other useful measures and measurement strategies that could be explored.
The paper referred to in the OP's media link is free online. It reviews the literature and presents its own new studies as well:
My takeaway: Yes, Facebook activity can be used to predict the Big Five personality traits, which, in turn, can predict job performance, but HR people aren't necessarily qualified to make a valid assessment. That, and the usual "but more research is warranted".
