Experiences of Everyday Racism and Media-Mediated Racism in the (Political) Public Sphere

Research question/goal: 

The project investigates media-mediated racism. Media mediation includes both news media coverage and communication on social networks (SN). We distinguish three dimensions: 1) explicit hostility towards groups of ethnically/culturally defined “others”, 2) implicit discriminatory biases in language use, and 3) emotionality of intergroup relations.
The questions each address one dimension:
1. What forms of explicit racial disparagement and hostility are found in the news media and on social networks, and to what extent?
2. What forms of implicit racial discrimination are found in the news media and on social networks, and to what extent?
3. What forms of emotional coloration do journalists and SN users employ to construct relations between ethnically/culturally defined groups?
In relation to racism in SN and in everyday life, various experiences of racism will be explored:
4. To what extent do people of immigrant background experience racism in online and offline contexts and are questions about a person’s country of origin perceived as racism?
Finally, we are interested in the geographic context of racism in SN:
5. Is racist language influenced by users’ geographic context?
The following methodological approach is used to answer these research questions:
The extent and target groups of attacks are measured by combining named entity recognition, sentence structure analysis, and semantic analysis. A dictionary is created to detect racist language.
The measurement of racial bias uses word embeddings (Stanford GLoVE Word Embeddings), which reconstruct the meaning of individual words or phrases through their co-occurrence with other words in the text.
The emotionality of intergroup relationships is also analysed using named entity recognition as well as contextually validated dictionaries for positive and negative emotions.
To survey immigrants’ experiences of racism, we conduct an online survey with individuals who have immigrant backgrounds. For this purpose, we use an online non-probability access sample, which allows us to target this population (e.g., with the provider Respondi). Recruiting the sample online is reasonable because we explicitly seek respondents who may have had experiences with racism in online contexts. We aim for a sample size of approximately 2,000 respondents to gather a broad range of experiences, achieve better estimate frequencies of experiences, and identify differences across immigrant groups.
The locations of Twitter users will be determined through a mixture of geo-tagging, text analysis, and network analysis. Using this data, a map of racist Twitter use, of a sample of the Twitter population and of political actors, will be created and processed in an app.

Current stage: 

The project is composed of two sub-projects. The first sub-project examines the extent to which members of the German Bundestag are confronted with racism on Twitter and how this is influenced by various factors. The second sub-project investigates the occurrence of explicit and implicit racist group stigmatizations in German media reporting. The results of these two sub-projects were presented at various conferences and each was summarized in a short report intended for the public, which are currently being revised. Both project teams plan to present further results at international congresses and to submit additional manuscripts.

Fact sheet

2020 to 2022
Data Sources: 
Survey data; social network data, content analysis of media reports
Geographic Space: