SAMF Data Lab 2023-24 Digital Infrastructure Research Grant Awards
Over 300000 DKK in Funding Awarded to SAMF Research Projects-125000 DKK to be Awarded in Final Round
12 SAMF research projects were awarded a total of 300000 DKK funding based on the following criteria:
"How will your project create or support digital research infrastructure within the social sciences, OR fund data intensive social science research that wouldn’t be possible otherwise? How will your project benefit the broader social science research community?"
List of Projects:
Birgit Bräuchler & Taufik Nurhidayatulloh
Anthropology
DIGINEX Project
Through a combination of online and offline ethnographic research, this project focuses on the role of digital media in raising environmental awareness in Indonesia and aims at gaining insights into how digital media usage can promote environmentally sustainable behaviour in Indonesia.
Hjalmar Bang Carlsen & Morten Axel Pedersen
SODAS
AInterviewer
This project aims to develop an open-source and local GDPR compliant digital research tool automating the collection of qualitative interview data. AInterviewer makes it possible to instruct an algorithm, drawing on generative AI, to ask open-ended follow-up questions to respondents akin to a classic semi-structured interview.
Jakob Demant
Sociology
ManuScrape
ManuScrape is a researcher and student tool for conducting large scale netnographic work and structured online observations, including recording the visual elements of interactions, combining qualitative netnographic observational notes, and integrating quantitative background data. ManuScrape provides three core functionalities that no other software offers, which is screenshots and scroll shots (extended screenshots), anonymization before data storage (ensuring GDPR compliance), coding and netnographic notes and export to NVivo and Excel.
Alice el-Wakil
Political Science
Representative Claims in Referendums
This project seeks to provide the first systematic evidence of the phenomenon of official campaigners making claims to represent others in referendums and ballot initiatives, which is often thought of as ‘direct’ democratic devices that are at odds with representation. The project seeks to learn about which political identities, values, or interests are invoked in making such representational claims and whether those claims are accepted by the relevant audiences.
Yevgeniy Golovchenko & Jan Stockbruegger
Political Science
Sanction Evasion and Covert Maritime Networks (SECM)
The project studies the impact of sanctions on Russia’s oil trade following Russia’s invasion of Ukraine using Automatic Identification System (AIS) shipping data. By using a Geographic Information System approach (GIS) coupled with Social Network Analysis (SNA) the project seeks to descriptively explore sanction evasion and causally test the effects of sanctions on maritime trade.
Marilena Hohmann, Roberta Sinatra & Jacob Aarup Dalsgaard
SODAS
Bias in AI-suggested Scientific Literature
The main objective of this project is to systematically compare the references in AI-assisted writing with those in human-authored publications. Thereby, the project investigates whether AI-generated references are biased, e.g. are fewer, older, and less diverse, which could have implications for researchers using AI. Therefore, the findings of the project could also shape policies on the ethical use af AI in academic settings.
Pia Ingold
Psychology
Be Attractive and Succeed in the War for Talent? Antecedents and Consequences of Employer Images in Job Ads
The project studies how relevant branded employer images in job ads are for recruitment success and perceptions of applicants, and what factors predict how organizations depict themselves as employers. Thereby, the project aims to generate insights that have direct relevance and applicability to recruitment practices by explaining how much projected employer images matter for organizations’ recruitment success, and to what extent applicants vary in their perception of branded images.
Yani Kartalis & August Lohse
SODAS
TextcompareR: An R package to Seamlessly Compare Texts Utilizing Latest Advances in Large Language Models
The project aims at building an R package that will allow users to seamlessly utilize the latest advances in Large Language Models to compare texts in their respective field and produce comparability scores. The project contributes to the broader social science research community by developing a tool for comparison of texts and by creating a multilingual training dataset, including embeddings, that will be made public and can be used for further training of similar tools/models.
Mikkel Haderup Larsen
Sociology
How Different Minority Status Factors Influence People’s Perception of Fair Social Benefit Rates
This study seeks to shed further light on how people’s perception of social benefit rates changes with different types of minority status factors such as citizenship, length-of-residence, and ethnicity. Unlike earlier studies of the phenomenon, this study will disentangle the effect from the different minority status factors, and it will examine if a prior work history acts as a moderator e.g., by mitigating the effect of a minority penalty.
Lau Lilleholt Harpviken & Ingo Zettler
Psychology
Funding for the continuation of The Danish Personality and Social Behavior Panel (POSAP)
POSAP is a worldwide unique research panel that combines well-validated social science questionnaires and experiments with register data from Statistics Denmark. It comprises 14.070 participants, largely representative for the Danish adult population who since 2021 have been invited to one measurement occasion per year.
Eya-Mist Rødgaard & Kamilla Miskowiak
Psychology
Improvement of the Internet Cognitive Assessment Tool (ICAT)
By improving the ICAT to make it suitable for large-scale online-only studies, the ambition is to recruit large samples of participants with psychiatric diagnoses through the Danish health registers and from these construct a research database of individuals who consent to be contacted regarding future studies. Based on the participants ICAT results, the database will make it possible for future cognitive studies to target the most relevant individuals and by using the registers as a basis for the database it will be possible to address selection bias, which will help researchers gain a better understanding of biases in psychiatric research studies.
Stefan Voigt
Economics
Unified Bayesian Model Average Method for Empirical Asset Pricing
Designing a unified BMA method is the first step of a larger research agenda that seeks to deal with two key challenges regarding empirical asset pricing: 1) that there is a lack of a standard path for empirical analysis and 2) that researchers face a “zoo” of empirically “priced” sources of risk as stock returns. The larger research agenda aims to achieve three primary goals: i) to quantify the relevance of methodological uncertainty on the estimated stochastic discount factors, ii) the development of a BMA method to shield against misspecification in portfolio choice and risk assessment, and iii) to identify relevant factors to guide the relevant economic drivers which dictate asset prices.
Data Lab and Funding:
This grant program is funded by the UCPH Data Science Program, a strategic initiative to improve researchers’ digital skills and usage of digital data sources. The Data Lab is hosted at the Copenhagen Center for Social Data Science (SODAS), which brings together researchers from all departments at the Faculty of Social Sciences.
If you have any questions, please write to Data Lab Team Leader Joseph Burgess: datalab@samf.ku.dk.