Online Seminar Talk
Three Talks
Sofia Beatriz Gil, Jisu Kim, Asli Ebru Sanlituerk
Laboratory of Digital and Computational Demography, June 21, 2022
Does Twitter Mirror the European North-South
Family Ties Division? A Comparative Analysis of Southern and Northern European Users’ Family Tweets
Sofia Beatriz Gil, MPIDR
Abstract
Research based on geographical distance and frequency of contact between family members shows that the strength of family ties differs between Northern and Southern Europe. But little is known about how family ties are reflected in peoples’ conversations in social media. This is despite the relevance of social media for users’ daily expressions of emotions and thoughts based on their immediate experiences. This work studies whether the use of Twitter in Europe mirrors the North-South division in the strength of family ties by analyzing whether there is a difference in family related tweets between users from Northern and Southern European countries. The work relies on a longitudinal database derived from Twitter collected between January 2012 and December 2016. We perform a comparative analysis of South- and North-European users’, first, by using data mining techniques to analyze the frequency Twitter users refer to family. Second, we use multilevel models together with the ICWL-22 software to analyze the association between family ties and tweeting about family, time of the tweet sentence, and the emotion conveyed. Preliminary results show that: Twitter use reflects the North-South division in attachment to close vs extended family; and that type of tie is associated with the time focus and emotion conveyed in the tweets.
Online Social Integration of Migrants: Evidence from Twitter
Jisu Kim, MPIDR
Abstract
As online social activities have become increasingly important for people’s lives and well-being, understanding processes of integration in online spaces is crucial for providing a more complete picture of integration processes. In this work, we carefully curate a high-quality dataset to quantify new social connections of immigrants in the United States based on Twitter data. In order to isolate the effect of individual migration events, we complement the dataset of migrants with one of comparable non-migrants using a propensity score matching technique. The results provide evidence of migration events leading to an expansion of friends network on Twitter in the destination country. This effect seems to be most prominent for migrants who have been in the destination country, the United States, for more than three years. Regarding their characteristics, we identified that male migrants, between 19 to 29 years old, who actively post more tweets in English after migration tend to have more local friends than other contrasting groups of these characteristics after migration. Interestingly, we also observed that migrants in their later period of their migration years tend to again add more origin-based friends on Twitter. Lastly, different from the friends network, which is under the control of the migrant, we did not find any evidence for a role of the migration event in expanding Twitter users' followers network in the destination country. While following users on Twitter in theory is not a geographically constrained process, our work shows that offline (re)location plays a significant role in the formation of online networks.
Emigration of Academic Scientists falls with Development in Low-Income Countries but Rises in Richer Ones
Asli Ebru Sanlituerk, MPIDR
Abstract
Does economic development increase or reduce emigration of academic scientists? The answer to this question holds the key to understanding how the global competition for talent affects countries at different levels of development. Leveraging bibliometric data indexed by Scopus for over 36 million journal articles and reviews published from 1996, we developed a novel and global database of scholarly migration, where migration events are inferred from changes in authors’ institutional affiliations over time. Statistical analysis using Generalized Additive Mixed Models reveals that emigration rates initially decrease as GDP per capita increases. Then the trend reverses as countries get richer. This U-shaped pattern is the opposite of what has been found for overall, population-level, emigration rates, and calls for new theoretical frameworks to understand the heterogeneous responses of migration to development. As we develop our models further, we aim at identifying the main determinants and consequences of observed dynamics.