New data sources for social indicators: the case study of contacting politicians by Twitter

Figures & data

Figure 1. Active citizenship as a variable included in the quality-of-life framework suggested by Eurostat (2016) and the dimensions and variables of the Active Citizenship Composite Indicator (ACCI) suggested by Hoskins and Mascherini (2009).

Figure 2. Research method for analysing the potential of Twitter data as a source for the survey indicator ‘contacted a politician to express an opinion’.

Figure 3. Workflow of Twitter data processing to obtain the measurement of contacts with politicians for the selected cities.

Figure 4. Temporal evolution of the data collected: Tweets from citizens mentioning a politician or municipal representative of Madrid, Seville, and Cádiz.

Figure 5. Spatial distribution of the individuals who contacted a politician or municipal representative of Cádiz, Seville, and Madrid via Twitter during the data collection period (excluding the contacts from Spain).

Figure 6. Temporal evolution of the indicator ‘contacted a politician’ from 2011 to 2015 in Cádiz, Seville, and Madrid, using the CIS surveys as data source.

Table 1. Indicator ‘contacted a politician to express an opinion’ in Cádiz, Seville, and Madrid. Values from the 2015 CIS survey, and values obtained using Twitter data for a 42-day period of analysis extrapolated to one year.

	CÁDIZ		SEVILLE		MADRID
	Contacts	%	Contacts	%	Contacts	%
Official data source (CIS)	0	0.0	17,417	2.5	73,750	2.3
Unofficial data source (Twitter)	1,205	1.0	27,239	3.9	99,798	3.2

Table 2. Comparison of the methods to obtain the citizen engagement indicator: based on Twitter data and based on traditional data from official surveys.

Evaluation criteria	Unofficial data source (Twitter)	Official data source (CIS survey)
Accuracy	Twitter data present several biases. Twitter users are not representative of the whole population, the information offered by the users might be false, and the tweet streams from the public API might be filtered with unknown proprietary algorithms	The survey sample was produced using a defined method to obtain statistically representative measurements of the target universe. Associated metadata are offered related to the technical details of the survey design and data collection. Particularly, the barometers used in this study present a sample error of ±2%
Scope and coverage	The Twitter universe might not be representative of the whole population. Demographics of Twitter users in our data collection are unknown	The coverage of the data is defined by the target universe, i.e. Spanish population of both genders over 18 years of age
Spatial dimension	Twitter data can offer accurate geoinformation when tweets are generated from devices equipped with GPS. However, most geolocated tweets in our study were obtained using the geolocation method. Thus, the tweets were mainly geolocated to a city level, with an accuracy of 84%	CIS surveys are designed to offer indications of the situation in Spain, considering the different regions and ranges of city sizes. Despite that the microdata is useful for obtaining indications at a local level for cities of a medium or large size. In small cities, the sample size is insufficient to be representative
Temporal dimension	Using Twitter’s Stream API, real-time data can be downloaded at any time, for a period of time that can be defined depending on the purpose of analysis. When collecting data from the past, the Twitter’s Search API limits acquisition to tweets from the past seven days	Data collection is done in periods of approximately 10 days, though the specific indicator asks about contacts with politicians during the past 12 months
Timeliness	Replication of the method can be done at any time. Results from the method can be obtained very quickly, depending on the number of tweets that need to be geolocated	CIS administers official surveys on citizen participation approximately once a year. The results are published a few months later
Coherence	The method ensures data comparability over time and space. However, the consistency of results is unknown due to the unclear provision of tweets from the public Streaming and Search APIs	Periodically programmed data collection allows long time series analysis. Data are comparable across the data collected, and with previous releases. The results are comparable with other studies considering the same spatial and temporal dimensions
Cost	The method implies a substantial effort in design and development of the process. Discovery of the Twitter user profiles of the politicians selected may be time-consuming, depending on the popularity of the politician. After that, the method can be applied and replicated with very little effort. To reproduce the method at a different location would require low to medium effort, depending on the difficulties encountered in identifying the Twitter user profiles of the local politicians	CIS surveys costs are not documented, but considering the labour of the statisticians, the surveyors, the data management and publication, it can be assumed to be several thousand euros. Nonetheless, ‘contacted a politician’ is but one of the 45 questions included in the questionnaire, thus overall costs should be divided across all questions posed
Flexibility	The proposed method is easily adjusted to the objectives of the analysis	CIS surveys on citizen participation usually include fixed questions, which makes comparisons over time easier
Method feasibility	The feasibility of the method is dependent on access to Twitter public API, the existence and ease to discover Twitter profiles from politicians or municipal representatives, and knowledge of the local political system (which is to be verified through a follow-up research)	Access to the CIS surveys microdata requires registration in the CIS online system, and agreement to the terms and conditions.

Eurostat. 2016. “Quality of Life Indicators.” Accessed July 6. http://ec.europa.eu/eurostat/web/gdp-and-beyond/quality-of-life/data/overview.

 Google Scholar

Hoskins, Bryony L., and Massimiliano Mascherini. 2009. “Measuring Active Citizenship through the Development of a Composite Indicator.” Social Indicators Research 90 (3): 459–488. doi:10.1007/s11205-008-9271-2.

 Web of Science ®Google Scholar