Understanding the Demographics of Twitter Users

By Alan Mislove, Sune Lehmann, Yong-Yeol Ahn, Jukka-Pekka Onnela, J. N. Rosenquist. 2011

920 citations recorded by the source corpus.

Every second, the thoughts and feelings of millions of people across the world are recorded in the form of 140-character tweets using Twitter. However, despite the enormous potential presented by this remarkable data source, we still do not have an understanding of the Twitter population itself: Who are the Twitter users? How representative of the overall population are they? In this paper, we take the first steps towards answering these questions by analyzing data on a set of Twitter users representing over 1% of the U.S. population. We develop techniques that allow us to compare the Twitter population to the U.S. population along three axes (geography, gender, and race/ethnicity), and find that the Twitter population is a highly non-uniform sample of the population.

Subfields: Artificial Intelligence, Statistical and Nonlinear Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistical and Nonlinear Physics

Citations in the corpus

Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019, background): “In general, both early surveys from the Pew Research Center and academic studies (Hargittai, 2007; Mislove et al., 2011) show that the demographic composition of major social platforms diﬀers both with respect to each other, and with respect to the oﬄine or Internet population (see e.g., Duggan,…”
"How Old Do You Think I Am?" A Study of Language and Age in Twitter (2013, methodology): “Mislove et al. (2011) analyzed the US Twitter population using data from 2006-2009.”
Gender Inference of Twitter Users in Non-English Contexts (2013, background): “The use of name-gender associations are problematic when non-English content is considered because databases of anglophone name-gender associations are no longer useful (Mislove et al., 2011).”
Tampering with Twitter’s Sample API (2018, background): “Some platforms are more attractive to specific sociodemographic groups than others [37] and specific services of platforms can appeal differently to certain groups, e.”
Inferring Gender from the Content of Tweets: A Region Specific Example (2012, background): “Mislove et al. (2011), basing their work on a set of over one billion tweets collected between 2006 and 2009, looked at the geographic distribution, gender, and ethnicity of Twitter users in the United States.”
Characterizing the Demographics Behind the #BlackLivesMatter Movement (2015, background): “Looking at organizations, accounting for about 5% of profiles, we see a 3 times higher tweeting rate than for individuals.”
Who Does What on the Web: A Large-Scale Study of Browsing Behavior (2012, background): “Likewise, a demographic study of the Twitter population (Mislove et al., 2011) used self-reported name and location information to reveal that Twitter users are a highly biased sample of the U.S. population in terms of gender, race, and location.”
Online and Social Media Data As an Imperfect Continuous Panel Survey (2016, background): “[43] documents the demographic skew on Twitter along with other studies that focus on inferring unknown demographics and attributes of social media users [44,45,46,47], but do little in joining this information with activity and content.”
Confounds and Consequences in Geotagged Twitter Data (2015, methodology): “We induce distributions over author gender in much the same way (Mislove et al., 2011).”
External validity of sentiment mining reports: Can current methods identify demographic biases, event biases, and manipulation of reviews? (2014, background): “Furthermore, given names and family names can reflect gender and ethnicity [14].”
The Demographics of Social Media Users in the Russian-Language Internet (2017, background): “One of the first papers on Twitter demography published by A. Mislove et al. [13] asked the following questions: (i) Who are Twitter users?”
Using Twitter to Examine Smoking Behavior and Perceptions of Emerging Tobacco Products (2013, background): “African-American, Hispanic, younger, and urban populations are in fact overrepresented on Twitter relative to the general population [3].”
Who Makes Trends? Understanding Demographic Biases in Crowdsourced Recommendations (2017, background): “Mislove et al (Mislove et al. 2011) provided one of the ﬁrst studies in this space, by looking at the gender and racial demographics of Twitter users, and analyzing how the demographics vary across different US states.”
A Critical Review of Online Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2018, background): “…representativeness of online social and behavioral data, covering major classes of social data biases such as population biases (distinct platforms are characterized by different demographic composition of users) [16, 24], or behavioral biases (users behave differently in different contexts) [23].”
A Tale of Cities: Urban Biases in Volunteered Geographic Information (2014, background): “In research that also helped to motivate this paper, Mislove et al. (Mislove et al. 2011) These papers are not specifically interested in the rural/urban divide, which is significantly more complex than raw populations and distance from significant cities (e.g. see the census discussion below).”
Obfuscating Gender in Social Media Writing (2016, neutral): “Since gender is not known for users in either dataset, it is inferred from users’ ﬁrst names, an approach commonly employed in research on gender classiﬁcation (Mislove et al., 2011).”