1 of 28

THE SOM SURVEY AND THE CITIZEN PANEL

OPEN DATA, COLLABORATION AND NOT-SO-OPEN DATA?

TILDA EKSTRÖM & MARCUS WEISSENBILDER

2 of 28

The SOM Institute

An independent research organisation at the University of Gothenburg. Since 1986, we have combined a close relation to survey research with a strong presence in the public debate.

​

​

3 of 28

Brief history

  • Started in 1986 as a collaborative survey project between researchers in political science and media studies
  • Became a centre at the Faculty of Social Sciences in 2010
  • Joined together with the Swedish Citizen Panel in 2014
  • Became a national research infrastructure supported by the Swedish Science Council (VR) in 2018

4 of 28

The purpose of the SOM Institute

  • Serve as an efficient, high-quality research infrastructure for survey-based data collections in Sweden
  • Conduct long-term research and analysis on the Swedish societal development
  • Actively communicate results outside of the university
  • Conduct research collaborations outside of academia with public agencies
  • Conduct methodological research and development on survey data collections

5 of 28

Organisation and staff

In total, 17 employees:

​

→ Research Director: Annika Bergström

→ 5 Researchers (political science, media,

psychology/economics, survey methodology)

→ 2 Survey managers and 7 Assistant survey

managers

→ 1 Communications officer, 1 Administrator

6 of 28

Two main facilities for data collections

The annual SOM surveys

  • Large cross-sectional, representative samples of the Swedish population (mainly paper surveys)

The Swedish Citizen Panel

  • Online access panel (strictly web surveys)

7 of 28

Our different types of open/available data

8 of 28

The SOM-surveys

9 of 28

How the SOM-surveys are conducted

  • Yearly national (n=26 250), regional (n=6 000) and local surveys (n=9 000)
  • A paper questionnaire is delivered to the respondent with an option to respond online
  • The questionnaires are between 12-16 pages
  • Five postal reminders, two text messages and reminders via digital mailbox are sent out during the field period (which is 100 days)
  • Respondents receive a lottery ticket after having sent in their survey
  • After the field period is ended all direct identifiers/keys are destroyed, so that individuals can’t be identified

10 of 28

Response rate over time in the national SOM survey

1986–2025 (percent)

68

52

40

​

30

​

20

​

10

​

0

50

60

100

​

90

​

80

​

70

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

11 of 28

Results from the SOM-surveys

  • Often in the form of time-series
  • However also many group comparisons

12 of 28

(Our own) annual output – the SOM surveys

  • Datasets: 3–5
  • Book series (anthologies): 1–3 each year
  • General and/or themed reports : 40+ each year
  • Notes on survey methodology 2–3 each
  • Presentations and seminars

13 of 28

Making data available

From the SOM survey each year:

  • 3-4 datasets from the annual surveys
  • 3 cumulative datasets 1986-

​

  • Available for any researcher
  • Also available for students

​

  • Free to request via Researchdata.se

14 of 28

What data is not made available?

  • While all data from the SOM surveys eventually become freely available to researchers and students there are some restrictions.
  • Data we collect during e.g. the autumn/winter of 2025 first becomes available in the spring of 2027, giving our collaborating partners a chance to do their own research.
  • Even as data is made available, for the first five years data is ”reduced”.

Open-ended questions are removed, some geographical variables (such as municipality) and other variables which increase the risk of backwards-identification are also removed or aggregated.

  • After five years a more complete dataset is added to Researchdata.

15 of 28

Collaboration within the SOM surveys

  • Researchers can collaborate with SOM to include their own questions in the survey
  • SOM assists with question wording, as well as questions of scales and other methodological aspects
  • This type of collaboration is associated with a cost
  • SOM does however not strive to make any profit, the fee is simply based on the space questions take in our survey and our costs of conducting the survey
  • See more information here:
  • https://www.gu.se/en/som-institute/collaboration/collect-data-in-the-som-surveys

16 of 28

Annual output from others using SOM data

  • Data requested by 150 students for use in student papers

​

  • Data requested by on average 30 Ph.D. students per year

​

  • Approximately 40 peer reviewed journal articles per year

​

  • Several 100 news articles referencing SOM per year

17 of 28

The Swedish Citizen Panel

18 of 28

The Swedish Citizen Panel

  • Online Access Panel (no other survey modes offered)
  • Academically administered
  • Exclusively used for research:
    • Longitudinal studies
    • Survey-embedded experiments (including conjoints)
    • Cross-sectional studies
    • Recruitment to qualitative studies

19 of 28

The Swedish Citizen Panel

  • Approximately 70 000 participants aged 16+ living in Sweden
  • Self-recruited and probability-recruited participants:
    • Self-recruited: 70%
    • Probability-recruited: 30%
  • No economic incentives are offered to participants

​

  • The large number of participants enable targeted sampling across socio-demographic groups, professions, opinions and habits

→ Tailored solutions

20 of 28

Waves

  • We conduct four large omnibus survey waves each year: in February, June, September, and December
  • Each wave includes several different studies, ranging from survey experiments to cross-sectional studies
  • In addition to these omnibus waves, separate projects and questionnaires can also be conducted between waves upon request.
  • Participation rate around 60 percent

21 of 28

Based on collaboration with other research groups

  • Researchers may apply to collaborate for a user fee
    • Cost is estimated based on the type of sample, the length of the questionnaire, and the number of requested completed responses
  • Feedback on the questionnaire is provided to researchers, including suggested revisions
  • We offer support with ethical review and data management plans
    • Technical reports are provided after each data collection
  • Approximate timeline: About 3 months from application to completed data collection
  • Approval from the Swedish Ethical Review Authority is required when collecting sensitive personal data

More information about collaboration with the Swedish Citizen Panel

22 of 28

Technical reports for each data collection

https://www.gu.se/som-institutet/medborgarpanelen/for-dig-som-forskar/panelvagor-och-tekniska-rapporter

23 of 28

Technical report example

•

https://www.gu.se/sites/default/files/2025-01/TR_SCP_MP56.pdf

24 of 28

The CORE panel

  • In addition to collaborating research projects, we also have our own longitudinal panel
  • About 4 000 panelists
  • Get identical questionnaires once or twice per year
    • The surveys include a set of societally relevant questions related to: subjective health, media use, political ideology and attitudes, trust in public institutions, social trust, and perceptions of Sweden’s economy
  • The panel has been active since 2010 (35 waves)
  • Available for researchers for a fee
    • Ethical approval from the Swedish Ethical Review Authority is required

25 of 28

Data availability

  • Data from the Swedish Citizen Panel is only shared with collaborating researchers
  • The CORE panel is available to researchers for a fee
  • In general, data are not publicly accessible due to legal restrictions related to the panel structure
  • We provide legal guidelines for researchers working with our data

26 of 28

Open and transparent research

27 of 28

All methodology and publications freely available online

​

28 of 28

Thank you for your attention!

Questions?

Contact: marcus.weissenbilder@som.gu.se