Schweizerischer Verein für ausgewogene Berichterstattung
SVFAB has developed the world’s only methodology to examine public broadcasting systematically, AI-assisted and reproducibly against its legal mandate. We offer this analysis as a service — for organisations pursuing the same standard in their country.
SVFAB already operates a comprehensive data pool with nearly 30 million items from 150 countries — updated daily. Data is available for most countries and broadcasters. Partners define the research question — we deliver the analysis from the existing pool.
Standard — Analysis from the pool
Time period, topics, criteria, frequency — freely definable. The data is there. You choose what we analyse.
Extension — Special corpora
For content not included in the standard pool — government documents, regional broadcasters, historical archives — we expand the corpus specifically. The partner delivers the material or defines the sources, SVFAB integrates and analyses.
Why centralised? Because only this guarantees comparability. The same methodology, the same criteria, the same scoring — across national borders. Like a measurement standard, not like a toolbox.
Every broadcast is analysed against 18 criteria: 17 quality criteria, 16 of them with a value from 0 (unremarkable) to 10 (strongly one-sided), and the political classification based on CHES. Criterion 9 has been merged into criterion 4; criterion 15 is reported descriptively. Every criterion with a value is classified as “clearly countable” or “requiring interpretation”; the composite score is the mean of all applicable criteria with a value. The criteria in detail:
Clearly countable — what is counted appears as such in the text:
1. Expert selection — Who appears as an expert, and is counter-expertise missing?
2. Source selection — Which side do the cited sources come from?
3. Time distribution — Who gets how much airtime?
5. Presentation of numbers — Are numbers reported accurately and in context?
13. Question asymmetry — Do all sides face equally tough questions?
17. Prejudgement — Is someone presented as guilty while proceedings are pending?
Requiring interpretation — before counting, it must be decided what counts:
4. Completeness — What is missing despite being relevant?
6. Guilt by association — Is someone discredited through association?
7. Timing — Where are incriminating and exculpatory elements placed?
8. Selective outrage — Is outrage applied asymmetrically?
10. Framing — Is the perspective announced, and does the introduction hold up against the report?
11. Word choice — Which terms are used for which side?
12. Moderation behaviour — How does the presenter act?
14. False balance — Are established facts and unsupported claims treated as equivalent?
16. Care and human dignity — Are identifiable individuals disparaged or exposed?
18. Source attribution — Are central claims attributed to a verifiable source?
Descriptive — without a value:
15. Agenda setting — Which preliminary question is never asked?
Chapel Hill Expert Survey (CHES) — international research standard:
19. Political classification — Where do the parties stand on the left-right spectrum?
Tool: Source Traffic Light
Assesses the credibility of every source appearing, in six dimensions; the result feeds into criterion 1 (expert selection).
Switzerland — SRG
Since 1968 · >1.2M items · >30 formats in 4 language regions (SRF, RTS, RSI, RTR)
Tagesschau, 10vor10, Arena, Rundschau, Kassensturz, DOK, Schweiz aktuell, Club, Echo der Zeit, Infrarouge, Mise au point, Temps présent, Telegiornale, and more.
Germany — ARD / ZDF
Since 2020 · >150,000 broadcasts · 4 formats
Tagesschau (ARD), heute journal (ZDF), Markus Lanz (ZDF), Frontal (ZDF)
Worldwide — 150 countries
Systematic analysis of public broadcasters on 5 continents, in 70 languages. Nearly 30 million items in the data pool, updated daily.
Regulatory authorities — Media supervision, quality control, programme monitoring: empirical basis for supervisory decisions.
Courts — Documentary evidence in programme complaint proceedings and concession disputes.
Legislation — Parliamentary committees and media spokespersons: empirical basis for inquiries, reforms and draft legislation.
Licensing — Review of mandate fulfilment as a basis for concession decisions and performance agreements.
Control and oversight — Systematic monitoring of mandate fulfilment over time series — not individual cases, but patterns.
Civil society — Organisations that want to examine public broadcasting in their country for balance — documentarily, not politically.
Research — Universities and researchers who want to use the methodology or the data pool for academic projects.
Published party programmes of all represented parties as the reference line. No political presupposition. Full source disclosure. SVFAB is not a political camp — we work with organisations that take balance seriously, regardless of which party benefits.
Interested?
We welcome enquiries from organisations, institutions and individuals worldwide.