A Tale of Cities: Urban Biases in Volunteered Geographic Information

By Brent J. Hecht, Monica Stephens. 2014

244 citations recorded by the source corpus.

Geotagged tweets, Foursquare check-ins and other forms of volunteered geographic information (VGI) play a critical role in numerous studies and a large range of intelligent technologies. We show that three of the most commonly used sources of VGI – Twitter, Flickr, and Foursquare – are biased towards urban perspectives at the expense of rural ones. Utilizing a geostatistics-based approach, we demonstrate that, on a per capita basis, these important VGI datasets have more users, more information, and higher quality information within metropolitan areas than outside of them. VGI is a subset of user-generated content (UGC) and we discuss how our results suggest that urban biases might exist in non-geographically referenced UGC as well. Finally, because Foursquare is exclusively made up of VGI, we argue that Foursquare (and possibly other location-based social networks) has fundamentally failed to appeal to rural populations.

Subfields: Epidemiology, Geography, Planning and Development

How researchers see this paper

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

In Epidemiology

In Geography, Planning and Development

Citations in the corpus

Social Data: Biases, Methodological Pitfalls, and Ethical Boundaries (2019, background): “For example, data-driven public policies may only succeed in economically advantaged, urban, and data-rich areas (Hecht and Stephens, 2014; Shelton et al., 2014), if eﬀorts are not made to improve data collection elsewhere (see the “digital divide” on section 9.4).”
Systematic patterning in phonologically‐motivated orthographic variation (2015, background): “Finally, social media has well-known demographic and geographic biases (Mislove, Lehmann, Ahn, Onnela and Rosenquist, 2011; Hecht and Stephens, 2014), making it difficult to draw robust conclusions about the linguistic profile of any particular population.”
The Geography and Importance of Localness in Geotagged Social Media (2016, methodology): “It appears that not only is there more social media VGI per capita in urban areas [20,29], but our results suggest that urban social media is also far more local.”
Identifying Platform Effects in Social Media Data (2016, methodology): “…of data comparison have been used to derive demographic information in social media data (Chang et al. 2010; Mislove et al. 2011; Sloan et al. 2013; Hecht and Stephens 2014; Longley, Adnan, and Lansley 2015; Malik et al. 2015) and to identify biases in public APIs (Morstatter et al. 2013;…”
Social Media data: Challenges, opportunities and limitations in urban studies (2019, background): “Also, there is a difference in terms of the quantity of information retrieved from LBSNs between rural and urban areas (Hecht & Stephens, 2014); and, in the specific case of Twitter, the fact that only a small portion of its users activate the geocoded function when publishing tweets is also an…”
Structural Causes of Bias in Crowd-derived Geographic Information : Towards a Holistic Understanding (2015, background): “Across multiple CdGI datasets, people who live in urban areas have been found to be more active participants in CdGI-generating activities than people who live in rural areas (Hecht and Stephens 2014; Gilbert, Karahalios, and Sandvig 2010).”
Confounds and Consequences in Geotagged Twitter Data (2015, background): “We focus on the ten largest metropolitan areas in the United States, and consider three sampling techniques: drawing an equal number of GPS-tagged tweets from each area; drawing a county-balanced sample of GPS-tagged messages to correct Twitter’s urban skew (Hecht and Stephens, 2014); and drawing a ”
The psychology of job loss: using social media data to characterize and predict unemployment (2016, background): “, Twitter users are known to skew toward urban geographies [20]), at 250,000 our set of randomly selected people is large and thus has a better chance of capturing large scale population effects.”
There's No Such Thing as the Perfect Map: Quantifying Bias in Spatial Crowd-sourcing Datasets (2015, background): “Few studies exist in this domain, with a notable exception being the recent work by Hetch and Stephens [19]: they examined three of the most common sources of volunteered geographic information (VGI [14]) - that is, Twitter, Flickr, and Foursquare and showed that all of them are strongly biased towa”
Population Bias in Geotagged Tweets (2015, methodology): “More recently, Hecht and Stephens (2014) investigated urban biases across the US.”