Centre for Social Data Science

Research groups and projects

Good Digital Society Group

The Good Digital Society group is a multidisciplinary collective of researchers inspired by the question of how to bring about a good society for all in the digital era. It acts as a hub for doctoral researchers connected to CSDS as well as for individual research projects involving smaller teams. The collective is lead by Krista Lagus and hosted at the Centre of Social Data Science, Faculty of Social Science.

CSDS Research blog  features news from our projects and ongoing PhD projects, as well as highlights of completed MA theses.

Members

  • Krista Lagus

  • Maria Valaste

  • Kimmo Vehkalahti

  • Adeline Clarke

  • Leila Gharavi

  • Nea-Mari Heinonen

  • Johanna Laiho-Kauranne

  • Markus Petteri Laine

  • Viivi Pentikäinen

  • Noora Sopula

  • Marcella Zoccoli

Alumni

  • Titi Gävert
  • Maria Litova
  • Tuukka Oikarinen
  • Joni Oksanen

Projects

FinnSurveyText - Survey open answers analysis

Open-ended questions are an important way to obtain unexpected points of view from surveys. Previously, conducting qualitative analysis on Finnish open-ended survey responses has been time-consuming and difficult. Therefore, many researchers in survey statistics have been more confident analysing responses to closed questions.

This project examines how survey open answers in Finnish can be analysed fruitfully by combining the perspectives of survey statistics and the methods of modern language technology. So far, we have published an R tool for analysis of survey responses. Currently we are working on extending it to many other languages as well as on creating a graphical user interface for ease of use.

The analysis of voices expressed in small languages is crucial for the continued development of a diverse global world that is able to hear the concerns of its people, regardless of language and culture.

This project has been funded by the Research Council of Finland as part of the DARIAH-infrastucture for Social Sciences and Humanities.

LATA: Trust in Algorithms in the Age of AI

The project “Trust in Algorithms in the Age of AI” (LATA), led by Kimmo Vehkalahti, examines trust in algorithmic governance and the so‑called algorithm aversion, in which people lose trust in machine decision-making more readily than in human decision-making when an error occurs. Although algorithms often make administration more efficient and save costs, citizens’ mistrust and their demand for near‑perfect error‑freeness slow down their use. The project investigates whether this tendency is inherent to humans or whether it changes as people gain experience, for example through the spread of large language models and generative AI.

Perma+H SOM - Wellbeing in Communities and Contexts

The “Wellbeing in Communities and Contexts”, led by Krista Lagus, combines the PERMA+H wellbeing framework with advanced computational language analysis to study how wellbeing is expressed across languages, communities, and contexts. We use methods including vector embeddings, language models, and other computational approaches to identify patterns and relationships in large collections of text. Self-Organizing Maps (SOMs) provide an exploratory and visualization method for examining these patterns and their variation across communities and contexts.

Neurodiversity at Work

“Neurodiversity and Work” investigates the workplace experiences of neurodivergent adults in Finland and how workplaces can better support neuroinclusive working environments. The doctoral project by Noora Sopula combines qualitative interviews with a national survey to examine the experiences, challenges, strengths, and support needs of neurodivergent employees and jobseekers. The research uses these findings to develop evidence-based, practical recommendations for more genuinely neuroinclusive workplaces. See project blog