How does it work?
Applicants can submit proposals (together with a substantive supervisor) for projects in the social sciences (e.g., psychology, sociology, economics, behavioural science) for which a data-related problem needs to be solved.
Examples of such solutions could be writing a high-quality data analysis pipeline, performing a causal analysis with simulation-based robustness checks, creating a software package, etc.
Who can apply?
Early-career social scientists (i.e., PhD candidates, early-career postdocs, pre-PhD researchers who have finished their studies) in any domain of social sciences. SoDa fellows currently in a PhD or Postdoc position can pause/extend their current appointment for the duration of the programme and resume it after the SoDa fellowship is concluded.
Deadlines and timeline
SoDa fellowships are open on a rolling basis and the call is open throughout this financing period but is subject to availability. The deadlines for submission in the current period are the following:
- 30 September 2024
- 12 January 2025
Call for Proposals
Current SoDa Fellows
Gabrielle Martins van Jaarsveld (EUR, ESSB): Large Language Model (LLM) Driven Text-Mining to Understand and Support Students SRL Processes. Year: 2024
This project focuses on better understanding how students with differing self-regulated learning (SRL) levels use intervention carried out by a text-based conversational agent (CA) by utilising the Large Language Model (LLM) assisted text-mining. For this project, OpenAI’s ChatGPT models will be used to carry out the text-mining analyses. This fellowship will result in a scientific publication, and a professional publication containing guidelines for how and when to effectively fine tune and use an LLM as a text-classifier.
Kristina Thompson (WUR): Using text analysis to build a Dutch historical disease database based on newspaper sources. Year: 2024
An important use of an infectious disease indicator involves the timing of exposure and a granular indicator of the disease environment in specific times and places is critical to undertaking such research. For the nineteenth- and early twentieth-century Netherlands, such an indicator of the disease environment is lacking and a dataset of disease mentions based on historical newspapers could help overcome the data scarcity. This project aims at creating such a data set and overcoming methodological challenges inherent in working with such a data source (e.g., representativity).
Joris Broere (SCP): Predicting the impact of policy on ‘brede welvaart’ outcomes. Year: 2023
This fellowship focuses on the implementation of “brede welvaart” (broad welfare) as the new standard for evaluating governmental policy. Brede welvaart is a set of 14 indicators that cover a wide domain of policy areas such as health, environment, well-being, and social cohesion which are more difficult to evaluate than traditional monetary indicators. The fellowship will focus on building a suitable prediction model for predicting policy alternatives on one of these indicators and will attempt to determine causality for predicting policy alternatives on one of these indicators.
Nadya Ali (Tilburg University): Poverty, prenatal maternal health and child health in the Netherlands. Year: 2023
Nearly 20% of the world’s population lives in poverty, half of which are children and children born into poverty are much more likely to live in poverty as adults. Using a network analysis approach this project aims to identify poverty-related environmental factors that affect maternal and child health. The output of the traineeship will be a scientific paper, a novel network model of poverty and its effects on maternal and child health, and a dashboard which will showcase the poverty-related environmental factors that are relevant to maternal and child health (such as nutrition, substance use, stress, living situation), with a map overview of the Netherlands.