Age-related Differences in Online Social Networking
Department of Psychology and Brain
School of Informatics
Systematic developmental changes in personality have been
observed over a large portion of the human lifespan.
Furthermore, in separate studies, these traits have been shown to
be predictors of local social network structure centered on an
individual (i.e., ego-centered networks). We are currently
exploring a sample of over 35,000 users in an online social
networking community, ranging in age from 15 to 55, to
investigate age-related differences in individual ego-centered
networks. Our data will allow us to explore structural holes,
reciprocal interactions, transitivity, and network similarity across
ages for multiple attributes (e.g., age and location).
Categories and Subject Descriptors
J.4 [Computer Applications – Social and Behavioral
Sciences] – sociology, psychology. H.3 [Information Systems –
Information Storage and Retrieval – Information Search and
Retrieval] – search process.
Social Network Analysis, developmental factors, preferential
attachment, social computing, MySpace.com
Systematic developmental changes in personality have been
observed over a large portion of the human lifespan. Narcissism,
associated with a lack of empathy and a need for admiration,
shows a steady monotonic decline from age 15 to 50 (Foster et al.,
2003). Even over the span of college life, significant and
systematic changes are observed in four of the Big Five
personality trait dimensions, with individuals on average
becoming more agreeable, conscientious, and open, but less
neurotic. In a large web-based survey, Srivastava et al. (2003)
has shown that these personality trait changes span a period
ranging from at least age 20 to 60.
Recent work by Kalish & Robbins (2005) indicates that
similar psychological traits are significant predictors of ego-
centered social network structure. Personality traits associated
with neuroticism, locus of control, group focus, and extraversion
correlated in a predictable fashion with distributions of triadic
relationships in self-reported social networks (see also Klein et al.,
2004). Other work has shown that narcissists are generally highly
extraverted (Bradlee & Emmons, 1992).
Observed changes in personality over the lifespan and a
significant tendency for personality traits to be associated with
social network structure suggest a strong likelihood that online
ego-centered networks (networks centered around an individual,
‘the ego’, and constrained to the egos immediate relationships,
‘the alters’) will show predictable and systematic changes across
The emergence of large online social networks such as
MySpace and Facebook present the opportunity to investigate
these age-related differences in social network structure in a non-
invasive way. While online data is potentially subject to its own
set of constraints, web based investigations have been found to
offer substantial robustness (due to large sample sizes) while
simultaneously reproducing the results of psychological
personality studies using non-web-based methods (Gosling et al.,
To investigate the age-related differences in online
social networking, we selected a large online social community of
over 57 million registered users and, using crawler-based
technology, randomly sampled individuals of different age-
classes. Our data allows us to investigate relations based on both
friendship claims as well as recent communications shared
between users. Our ongoing research seeks to investigate the
correlation between these differences and the social network of
individuals, based on criteria such as structural holes, transitivity,
reciprocal interactions, local clustering, and individual similarity
across non-network attributes (e.g., age and location).
MySpace is a very large online social networking community that
allows its registered users to create a publicly viewable profile, to
which connected friends can post comments. We designed a
web-crawler to randomly sample individuals from this
community. Our initial crawl captured 35,411 egos from the
MySpace network, and stored their age and sex, as well as their
city, state and country – where the data was available. The age
distribution is shown in Figure 1. This crawl yielded friend and
comment networks with over 500,000 edges.
We then took a representative sample of all 14-34 years olds from
the ego list as our seed, and crawled each of their alters’ profiles.
This provided us with a full ego-centered network, with the age
and sex attributes for all actors, and city, state and country for
Using this dataset, we are performing an exploratory analysis in
order to better understand the nature of network, and most
importantly the nature of the connective tendencies of egos based
on their developmental differences.
Figure 1. Age distribution of initial crawl.
2.2.1 Friend and Comment Degree
Our data allows us to examine and compare two relational
networks, claimed friendships (which are reciprocal and can be
created by ‘poking’ another user by inviting them to be a friend)
and directional relations based recent posts to another users page.
These relations will be explored for sex and age related degree
differences as well as second order structural features using triadic
analyses, such as structural holes and transitivity.
2.2.2 Similarities based on Attributes
Nonstructural attributes include age, sex and geography, will be
compared with one another and with structural attributes (such as
degree-related factors) to test for age and sex related differences in
local network similarity.
Our thanks to Rob Goldstone and Filippo Menczer of Indiana
Bradlee, P. M. & Emmons, R. A. 1992, Locating narcissism
within the interpersonal circumplex and the five-factor
model, Personality and Individual Differences 52, 163-176
Foster, D, Campbell, W. K., & Twenge, J. M., 2003,
"Individual differences in narcissism: Inflated self-views
across the lifespan and around the world, Journal of
Research in Personality 37, 469, 486
Gosling, S. D., Vazire, S., Srivastava, S., & John, O. P.
(2004). Should we trust Web-based studies? A comparative
analysis of six preconceptionsabout Internet questionnaires.
American Psychologist 59, 93–104.
Kalish, Y., and Robins, G. (2005) Psychological
predispositions and network structure: The relationship
between individual predispositions, structural holes and
network closure. Social Networks 28, 56-84.
Klein, K. J., Lim, B., Saltz, J. L., & Mayer, D. M. (2004)
How do they get there? An examination of the antecedents
of centrality in team networks. Academy of Management
Journal 47(6), 952-966.
Srivastava, S., John, O. P., Gosling, S. D., & Potter, J.
(2003). Development of personality in early and middle
adulthood: Set like plaster or persistent change? Journal of
Personality and Social Psychology 84,1041–1053.