Source: the internet – testing digital news and information literacy in Germany


The full study is available in German. This is a translation of the executive summary.

During pandemics, economic crises, and election campaigns, it becomes particularly clear that functioning democracies depend on well-informed citizens. How well people are able to understand, classify, and question news can influence whether people become susceptible to populists, lose trust in institutions, or share disinformation millions of times with friends and family. News and information literacy of society has gained in importance in recent years and has become a critical factor for democracies. Because of a radical transformation in the media system, journalists and media institutions have lost their influence as gatekeepers. More than ever, citizens are now on their own. For every single piece of news, they have to decide for themselves each time whether a source or piece of information is trustworthy. And whether to read it, like it, or even share it. 

It can be assumed that this profound change will continue to intensify in Germany — as in many other European societies — in the coming years and will lead to a number of political and social challenges. In order to be able to respond to this, federal and state governments, education and media policymakers, schools and educational institutions, and public broadcasters first need a more precise picture of the situation. The crucial question is how well citizens are currently able to cope with the changes in our media system and where people of different age groups have strengths or weaknesses. How well is the population able to judge the reliability of sources or to recognise, classify, and verify information at all, apart from the traditional newspaper on the Internet? How well can PR content, disinformation or opinion pieces be recognised and differentiated? And how competent are people in identifying incomplete news or conflicts of interest among sources and authors?

Until now, there has been a lack of reliable data on these important information and news skills in the German population – and thus the basis for a targeted media education policy. Although studies and surveys on "media literacy" already exist, they either focus only on students and primarily on their general PC skills, or they are based on surveys and self-reporting, which do not provide a reliable measure of literacy. For this reason, we developed a news literacy test together with a group of experts, which was carried out in the fall of 2020 with a representative sample of the German-speaking population with Internet access in Germany aged 18 and over. For this purpose, 4,191 Internet users aged 18 and over were surveyed and tested nationwide using online interviews (Computer Assisted Web Interviews - CAWI).

The test uses questions and tasks to address the entire spectrum of digital news literacy, i.e. the ability to navigate in digital media environments, assess the quality of news and content, check information, discourse skills and knowledge of how digital publics work. It is one of the world's first tests of information and news literacy, representative of an entire population.

A selection of the test results at a glance

1) Differences between disinformation, information, advertising, and opinion are sometimes difficult to recognise.

Respondents sometimes find it difficult to distinguish between different forms of communication, i.e., between advertising, information, disinformation, and opinion. For example, 56% mistook an advertisement for information, despite existing advertising labelling. Only 23% correctly recognised that it was advertising. Misinformation on Facebook also caused problems for respondents: A post containing misinformation was recognised by only 43% of respondents as such, while 33% also incorrectly categorised the post as information. The distinction between opinion- and fact-oriented posts is also critical. This is particularly difficult in the case of journalistic reports on political decisions. One-third of respondents considered a commentary (op-ed) to be factual reporting, while another 15% were not sure.    

2) Whether a source is trustworthy is often correctly assessed. Conflicts of interest are recognised less frequently.

Respondents were relatively good at assessing the neutrality or trustworthiness of sources. At least 59% were successful in this regard over the course of multiple questions. However, despite further information, it was often difficult to identify the specific conflicts of interest. For example, 65% of respondents recognised that the managing director of an air travel portal is not a neutral source as the author of an article on the subject of flying. However, only half of the respondents were also able to name the specific conflict of interest in this case.

3) Labelling strategies of social media platforms on disinformation are hardly effective so far.

Across the study, it emerges again and again that platform-specific notices are, in part, not very effective. Whether it's the Facebook label for fact-checking a false news story or the Wikipedia reference on YouTube, to the financing of a state broadcaster: only one-fourth of the respondents identified the label as a helpful reference or were able to classify the information within the post correctly. 

Similar problems can be seen with labels on news sites. Only 7% of respondents recognised the reference to an advertisement as an advertising label. And just under one-third of respondents identified the marking of an opinion piece as a "column" a helpful hint.

4) People doubt journalism's independence from politics. 

The idea that deals are commonly being made between the media and politics is widespread: A quarter of the population shares "Lügenpresse" (engl. lying press, press of lies) accusations. 25% agree with the statement that the media and politics work hand in hand to manipulate the public's opinion (another 28% say partly). 24% believe that the population in Germany is systematically lied to by the media (another 30% say partly). Only half of survey participants also know that news about a federal minister can be published without the ministry's approval.

In particular, the journalistic independence of public broadcasting is misjudged. Only just over half of those surveyed were able to answer correctly that members of the Bundestag cannot decide what the radio reports on. As many as 22% believe there is leverage by political actors in editorial decisions, while another 24% said "I don't know”. 35% of respondents also think that public broadcasting is under the control of the Minister of State for Culture and the Media (40% say "I don't know").

5) Just under half pass the test, only 22% of respondents achieve high overall competency scores.

To enable comparability of the results, we developed a scoring system for the test. Respondents could score a maximum of 30 points if they answered all questions correctly. The average score was 13.3 points, less than half the possible score. One third of respondents were in the medium range. Only 22% achieved high or very high competence scores and, at 46%, most respondents were in the (very) low digital news and information competence range.

6) Younger generations more competent than older ones - but depending on education.

Digital news literacy declines with age: the older the respondents, the lower their literacy scores. Or vice versa: the younger, the more competent. In addition to age, school education also plays a key role. When the two are considered together, it becomes clear how relevant the level of education is for news literacy, particularly among 18 to 39-year-olds: Highly educated respondents between 18 and 39 are particularly news literate, while the least news literate respondents are people under 40 with a low level of education. In general, the following applies across all age groups: the higher the formal school education, the higher the trust in journalism and politics and the higher the competence scores.

7) Digital news literacy is also linked to basic democratic attitudes.

In addition to education and age, the respondents' basic democratic attitude is also an important factor of digital news and information literacy. In our model, the basic democratic attitude includes the willingness of citizens to inform themselves about politics, an appreciation of independent journalism, a certain basic trust in democracy and the media, and the ability to tolerate other opinions. People who tend to reject these attitudes also show a lower level of news and information literacy. 

8) Digital news literacy particularly low among AfD supporters.

Supporters of different parties perform differently in our test: FDP (liberal Free Democrats) supporters achieve the best results, closely followed by the Greens. They are followed by supporters of the Left Party and the SPD (Social Democrats). CDU (Christian Democrats) supporters are pretty much in line with the overall average. AfD (far-right) supporters are in last place. The large difference between the FDP, the Greens, and the AfD suggests that (only) party preference is not the decisive factor here, but that education, age and/or fundamental attitudes — for example toward a supposed cronyism between the media and politics — have an influence on news competence. 


The overall analysis of the data shows: Internet users already have some basic knowledge to navigate competently through a media environment that is comparatively new to many people. For example, more than half of the respondents were able to recognise when a source was not neutral or not trustworthy. And most of them knew that an unfamiliar video should not be forwarded unseen. In general, the participants showed great interest in sharing accurate information and not spreading false news themselves. 

However, none of this should obscure the fact that respondents as a whole scored mediocre to poorly in almost all areas of this test and often lacked specific knowledge and skills. In this respect, the results of this survey are also critical because they show that citizens have been left alone for far too long to navigate increasingly complex media environments. 

Better digital education for schools and adults is needed.

The systematic neglect of digital skills is particularly evident in education policy – there is an urgent need to catch up in both school and adult education. Digital news and information skills are still not a part of curriculums. Especially in secondary schools, dimensions of media literacy that are related to political education and the craft of journalism have apparently been largely neglected in recent decades. This is particularly dangerous because, according to the available data, young people with a low level of school education form the socio-demographic group that has the lowest scores and also shows a particularly low level of trust in politics and the media. We cannot yet foresee what further social conflicts this polarisation could lead to. The need for education is also high among adults and older people. In the test, digital news literacy declined with age – and significantly so. For this reason, there is an urgent need to also include digital news and information literacy more systematically in adult education and, for example, in vocational training programs.

Transparent journalistic products are needed.

Overall, however, the poor test results do not only point to an educational problem. It becomes equally clear that it has become more difficult for citizens to recognise reliable news and to distinguish it from other forms of communication. This is due first and foremost to the media products themselves. Our results not only show how important it is to communicate the principles of the journalistic craft to readers in an understandable and transparent way. It also shows that journalistic products must provide much better support for users in classifying different forms of communication. For example, the separation of commentary and reporting is regulated in the German State Media Treaty, as is the labelling of advertising. However, the corresponding labels on news sites seem to be either not easily recognisable or not comprehensible. There is a need for improvement here. Especially those who find it more difficult to classify news articles, quickly get the impression of financially dependent and/or politically and opinion-driven journalism. This can lead to a decline of trust in the media.

Better platform architectures are needed.

The same applies to social media platforms, which also make it difficult for citizens to deal responsibly with news and information. In the social networks, users receive fragments of information in an even more unsorted manner, making competent navigation all the more difficult. The current trend of marking problematic posts, providing users with additional information, or even blocking accounts and content altogether is welcomed. But it does nothing to change the basic functional logic of the platforms, which favours rather than limits disinformation. In addition, respondents obviously had problems correctly recognising and classifying additional information. Without plausible and clearly visible labels, transparency about the platform architecture and design decisions that really support users in competent use, fact-check labels alone are not effective. 

The study is accompanied by a freely accessible self-test: (only available in German). The test is generally based on the questionnaire used in our study, but it does not collect any data itself. It enables users to test for themselves how news-literate they are. It encourages users to think critically about their own media consumption. In a playful approach, interested people can test their knowledge in five areas of competence and at the same time receive important information on how to navigate in the digital media world. The result shows how users performed in comparison to the overall result in Germany.

The newstest is freely accessible and may be linked, embedded and used in teaching.

The project "Digital News and Information Literacy" is supported by the Federal Agency for Civic Education/bpb, the Federal Government Commissioner for Culture and the Media (BKM), the Media Authority Berlin-Brandenburg (mabb) and the State Media Authority NRW (LFM NRW).

March 22, 2021

Dr. Anna-Katharina Meßmer
Alexander Sängerlaub
Leonie Schulz