Why we believe fake news is true
A study into the effects of partisanship and cognition on the capacity of identifying fake news
Master’s Thesis Graduate School of Communication
Author Petra van Gelsdorp Student number 11032871
Master’s programme Political Communication and Journalism Word count 7497
Supervisor Alessandro Nai Date 28 June 2019
Abstract
Into what extent are people capable of identifying fake news in the Netherlands? While previous studies into that issue were mainly focused on the United States, by conducting an online experiment this study aims to find out whether people of the Netherlands are capable of identifying fake news, and into what extent cognition and partisanship have an influence on this capacity. The findings show that people are significantly less capable of identifying fake news in which a politician is presented favorably, than of fake news in which a politician is presented unfavorably. Furthermore, participants failed in identifying real news as true. Also, results suggest that conservatives have a higher capacity of identifying fake news than
liberals. Although no significant results were found of partisanship and cognition on the capacity of identifying fake news, results still indicate that people find it hard to identify fake news, which can cause serious negative implications for the democratic system.
Keywords: Fake news, partisanship, cognition, information processing, experiment
The capacity of identifying fake news
“What caused me a lot of sorrow, is that I heard from a number of doctors and medical specialists they receive signals the plug is pulled on faster among the elderly with a migrant
background.”
This is a translated version of what Dutch politician Tunahan Kuzu of the left-wing political party DENK claimed on Facebook (NOS, 2017, para.6). Without any evidence or explanation, Kuzu argued that people in the Netherlands with a migrant background would die earlier in a hospital than native citizens. It caused a lot of commotion because what he said turned to be fake; no reports were ever done about the claims Kuzu made (Volkskrant, 2017). The political party, whose electorate mainly consists of immigrants (Parool, 2018), was criticized for its action. Because the information they used was fake and could cause a skewed perception among people. For example, the statement may induce fear among immigrants who do not speak the Dutch language which could prevent them from getting access to health care (NOS, 2017). However, DENK is not an exception in Dutch politics when it comes to fake news. Also, other politicians like Geert Wilders and Thierry Baudet have spread fake news before (Bouma, 2017). It is not surprising that politicians, or other individuals, create fake news because it can work effectively. As it is cited in the paper of Pennycook and Rand (2017), a study of Silverman, Strapagiel, Shaban, and Hall (2016) indicated that fake news is more popular than real news: the most popular fake news stories had more likes, shares, and comments on Facebook than the most popular real news stories. Some even state that Donald Trump would not have been elected as president of the United States without the help of fake news (Allcott & Gentzkow, 2017).
Recently, Dutch minister Kasja Ollongren of Internal Affairs expressed her concerns about the spread of fake news in the Netherlands. In a letter to parliament, she wrote about Russians who actively produced fake news with the aim to manipulate the Dutch public
opinion (Ministerie van Algemene Zaken, 2017). Minister Ollongren calls the spread of fake news a danger for Dutch society and its democratic system. Since the US elections of 2016 Facebook received criticism for its role in the circulation of fake news (Allcott & Gentzkow, 2017; Guess, Nyhan & Reifler, 2018). With more than 2 billion active users per month, Facebook is the most popular social media platform in the world where the ‘battle of the voter’ mainly takes place (Bouma, 2017; Baten, 2018). Like in the United States, it is also quite easy in the Netherlands to spread fake news via Facebook (NRC, 2018). For example, a Dutch journalist could easily publish a political advertisement full of lies during election times. And almost half of the Dutch population have ever been exposed to fake information (NOS, 2018).
One of the biggest problems with Facebook and the spread of fake news is the way how the content is presented. Facebook is making it hard for people to recognize fake news (Albright, 2017; Tandoc, Lim & Ling, 2017). For example, when someone publishes an article, it can reach a user via their own timeline or a shared post by friends. Furthermore, the source it is not prominently displayed in the format of the Facebook post. Also, journalists are less involved in the news circulation on the platform which is why the ‘watchdogs’ have become less powerful as the ‘fourth estate’ (Deuze, 2005). Instead, some argue that fake news is taking over as ‘the fifth estate’ by challenging the traditional media with their fabricated news content. It means that anyone is able to communicate their message via Facebook to the world without any verification of critical press (Bardoel, 1996; Bouma, 2017). Dutch
politicians, like Kuzu in the example at the beginning of this introduction, are able to produce fake news, design it like real news, and send this without any form of fact-checking to
(potential) voters, faster and way more direct, and sometimes with the intention to deliberately mislead the people (Bouma, 2017). All these aspects make it difficult to find out where the story originates from and thus whether it is accurate or not. From a normative point of view,
this development can have negative implications for democracy. As Ferree, Gamson, Gerhards, and Rucht (2002) underlines, in a representative democracy, citizens need to vote for politicians based on correct information, and the media need to deliver that information. But as fake news easily finds its way through the people, important requirements in a democracy like “transparency”, “proportionality”, “expertise”, and “evaluating the content” will not be met. Therefore, it could be possible that citizens using Facebook make political decisions based on false information. However, Facebook promised to invest in fact-checkers but in most of the cases users never receive those fact-checks (Guess, Nyhan & Reifler, 2018).
The Dutch government is taking measures against fake news. Minister Ollongren decided to invest 1 million euros in a campaign to raise awareness about this topic (De nieuwe reporter, 2019). The campaign ‘Stay curious. Stay critical’ was launched on 11 March 2019 with the aim to increase media literacy among Dutch citizens. Therefore, by informing people about the existence of fake news it should make them better in identifying fake news. Results from similar studies indicate that Dutch people have difficulty in judging a source as reliable or not (Vlooswijk, 2018). However, to our knowledge, the capacity of identifying fake news among Dutch people is still not investigated, while scholars are calling for more research into the role of the audience in believing fake news (Tandoc et al., 2017).
When it comes to identifying fake news, two aspects are of importance. At first, it depends on who people are in terms of partisanship. People rather believe a fake story when their preferred political candidate is portrayed in their favor (Allcott & Gentzkow, 2017). And second, on the way how people think in terms of cognition (Bronstein, Pennycook, Bear, Rand & Cannon, 2018; Roets, 2017; Kahan 2013; Miller, Saunders & Farhart, 2015; Pennycook, Cannon & Rand, 2018). When people are partisan, cognition can make people less capable of identifying fake news (Bronstein, Pennycook, Bear, Rand & Cannon, 2018).
But if partisanship does not play an important role, cognition can make people more capable of identifying fake news (Roets, 2017; Pennycook, Cannon & Rand, 2018). However, the literature is undecided in how partisanship and cognition interact with each other on the capacity of identifying fake news. Some state that motivated reasoning is a plausible explanation (Allcott & Gentzkow, 2017; Flynn, Nyhan and Reifler, 2017; Miller, Saunders and Farhart, 2015) whereas others think that motivated reasoning might not be that much of importance as might have been assumed before (Roets, 2017; Pennycook, Cannon & Rand, 2018). In addition, studies into identifying fake news are mainly orientated on the United States, but Humprecht (2018) found that the effects of fake news differ among Western countries. Therefore, this study will investigate into what extent Dutch people are capable of identifying fake news, and into what extent partisanship and cognition have an influence. The following research question will be held central in this paper: How does partisanship
influence the way people process fake news, and to what extent does cognitive capacity moderates this effect?
By conducting an online experiment, participants were exposed to Facebook posts with Dutch prime minister Mark Rutte making false or true statements about climate change. Since Facebook is considered as the most important channel when it comes to the circulation of fake news (Allcott and Gentzkow, 2017; Pennycook, Cannon & Rand, 2018; Spohr, 2017). And because a lot of people believe fake news about the political leader of the United States, Donald Trump, was true (Allcott & Gentzkow, 2017) this study will try to find out whether the same effects occur in the Netherlands. The topic of climate change is of importance to find out the role of partisanship and cognition. When people are exposed to such complex issues, they tend to rely on these mechanisms (Kahan, 2013).
Since fake news can have serious consequences for political outcomes, the aim of this study is to find out into what extent Dutch people are able to identify fake news and into what
extent partisanship and cognition are of importance in order to provide more insight into the effects of fake news in the Netherlands.
Theoretical framework
In the literature, many terms are used to describe fake news. Tandoc et al. (2017) provided a typology in which they distinct six different genres of fake news: “news satire”, “news parody”, “fabrication”, “manipulation”, “advertising”, and “propaganda”. Because news satire and news parody are more focused on entertainment purposes, manipulation and advertising have economical grounds, and the intentions of
propaganda overlay partly with advertising, this study only focus on the type of news which is the best related to the research question: “fabrication” (Tandoc et al., 2017). In the United States, this type of fake news had consequences on the political outcome of the elections (Allcott & Gentzkow, 2017). Allcott and Gentzkow (2017) also
investigated fabricated fake news investigated in their paper, which they define as followed: “… articles that are intentionally and verifiably false, and could mislead readers.” (p.213). Like the authors did in their study about fake news, this paper also focusses on news articles in which a politician make fake statements. Tandoc et. al (2017) defined “fabrication” more in detail:
This refers to articles which have no factual basis but are published in the style of news articles to create legitimacy. Unlike parody, there is no implicit
understanding between the author and the reader that the item is false. Indeed, the intention is often quite the opposite. The producer of the item often has the
intention of misinforming. Fabricated items can be published on a website, blog or on social media platforms. The difficulty in distinguishing fabricated fake news occurs when partisan organizations publish these stories, providing some semblance of objectivity and balanced reporting. (p.143)
On certain aspects, fake news significantly differs from real news. According to Horne and Adali (2017), real news articles have longer texts and are more complicated to read. In addition, fake news “use fewer analytic words, have significantly more personal pronouns, and use fewer nouns and more adverbs.” (p.763) Fake news is partisan
(Pennycook & Rand, 2017). According to Klein and Ahluwalia (2005) partisanship refers to how strongly, favorable or unfavorable, people identify themselves with a political party. In this study, a positive message means that the politician presented in the news is portrayed as favorable. In contrast, a negative message means that the politician presented in the news is portrayed as unfavorable.
Identifying fake news
When people have to identify fake news it means that they are able to distinguish fake from real news (Allcott and Gentzkow, 2017). In addition, people fail in identifying fake news if they consider it as real (Allcott and Gentzkow, 2017; Balmas, 2014;
Pennycook & Rand, 2017; Tandoc et al., 2017). Otherwise, when people consider fake news as not real it means they successfully identified it as fake. Nielsen and Graves (2017) looked into audience perspectives of fake news. According to them, people are capable of identifying fake news but only up to a certain extent. The authors found that if people think about fake news they rather come up with flaws in journalism,
propaganda, and advertising. They do not think of fabricated fake news. In addition, Pennycook and Rand (2017) found that a single exposure to a fabricated fake news article on Facebook is enough to make people believe the fake story is true. According to Kim and Dennis (2018), it is hard for people to recognize a fake story, especially because of the design Facebook is implementing. Facebook presents every story as news whereas the source is not prominently displayed. As Kim and Dennis (2018) argue:
“This subtle framing of a Facebook story as “news” not as a “story” influences how we process it; we adopt the mindset for processing “news,” not the mindset for processing “stories.” (p.3957). Because people are not always aware of the existence of fabricated fake news, a single exposure to a Facebook post makes it more likely to believe fake news, and the interface of Facebook does not make a clear distinction between verified and unverified news, the following hypothesis is expected:
H1. People will rather fail than succeed in identifying fake news.
However, to find out which people are (not) able to identify fake news there are two elements of importance. As was already mentioned in the introduction of this paper, these elements are partisanship and cognition.
Partisanship
When people are exposed to complex issues, they can rely on their political beliefs in order to take a position towards it (Taber & Lodge, 2006). Climate change is such a complex issue in which partisan polarization significantly increased for the past couple of years (Hart & Nisbet, 2012; McCright & Dunlap; 2011). For example, liberals do have more personal concerns about climate change and are more likely to agree that humans are responsible for the climate problem, in contrast to conservatives. Hart and Nisbet (2012) explained these differences based on the theory of motivated reasoning. According to this theory, people do not make their political decisions only based on rational ideas but, rather, on their own predispositions (Taber & Lodge, 2006). As Taber and Lodge (2006) explain in their paper: “While citizens are always constrained in some degree to be accurate, they are typically unable to control their preconceptions, even when encouraged to be objective. This tension between the drives for accuracy and belief perseverance underlies all human reasoning” (p.756). In other words, people do not evaluate from scratch every time they are confronted with new political information, but they use their own political beliefs as a cue to process it.
In order to protect one’s partisan identity, people prefer to pay attention to information which is in line with their ideology (Bolsen, Druckman & Cook, 2014).
In line with this theory, scholars found that partisanship also plays a role into how people process fake news (Allcott & Gentzkow, 2017; Flynn, Nyhan & Reifler, 2017; Miller, Saunders & Farhart, 2015; Pennycook et al., 2018; Van Bavel & Pereira, 2018). Loyalty towards people’s own ideological group seems more of importance than the truth (Van Bavel & Pereia, 2018). People rather believe a fake story when their preferred political candidate is portrayed in their favor (Allcott & Gentzkow, 2017). Also, conservatives are better in identifying fake news than liberals (Allcott & Gentzkow, 2017; Pennycook et al., 2018). Allcott and Gentzkow (2017) cite in their study the work of Bakshy, Messing and Adamic (2015) and offered an explanation of why conservatives are more likely to identify fake news correctly than liberals. The authors found that conservatives are exposed to a wider range of diverse news sources on Facebook than liberals do. Because of exposure to diverse content, it could help conservatives in detecting partisan fake news. Based on the theory addressed above, it is expected that the more partisan people are, the less likely they will be able to identify fake news. However, it is expected that conservatives will do a better job of identifying fake news than liberals. Therefore, the following hypotheses are formulated.
H2a. The more partisan people are, the more likely they will believe a fake story is true, when their preferred political candidate is portrayed in their favor.
H2b. Conservatives are more likely to have a better capacity of identifying fake news than liberals.
Moderating role of cognition
The other important aspect of identifying fake news is cognition. From the perspective of motivated reasoning, scholars found that high cognition reinforces the effect of partisanship on processing political information (Kahan 2013; Miller, Saunders &
Farhart, 2015; Bronstein, Pennycook, Bear, Rand & Cannon, 2018). Thus, when people are partisan, they rather engage with information in line with their ideology. In combination with high cognitive capacity, this effect will be even stronger. Kahan (2013) found evidence for the Expressive Rationality Thesis (ERT). As he cites Kahan, Jenkins-Smoth and Braman (2011) in his paper, he explains that people who think more heuristic-driven, System 1 cognitive processing, do not have to pay much effort to engage with partisan information. But, if people do think more analytically, System 2 processing, they are even better in engaging partisan information because they have the cognitive capacity to engage with the best factual information available. In line with this, Miller, Saunders and Farhart (2015) found when people have to process fake news (e.g. conspiracy stories) cognition reinforces motivated reasoning. Also, Bronstein et al. (2018) found evidence suggesting that people with higher cognitive capacity rather want to believe a fake story is true. In addition, some scholars found evidence that cognition does lead to a better capacity of identifying fake news when partisanship does not play a role (Roets, 2017; Pennycook, Cannon & Rand, 2018). Pennycook, Cannon and Rand (2018) cite their own previous work, in which they found that cognition does lead to a better capacity of identifying fake news (Pennycook & Rand, 2017). People with low cognitive skills would be less able to make a correct judgment about false information as they are less able to remember certain news issues. Also, Roets (2017) found that people with lower cognitive skills remained biased after exposure to fake news, even when they knew the story was fake. Therefore, it is expected that cognition does reinforce the relationship between partisanship and the capacity of identifying fake news. Thus, people who are partisan and have a high
cognitive capacity are less likely to identify fake news. But, without partisanship, higher cognition will lead to a better capacity of identifying fake news. Therefore, the following hypotheses are formulated.
H3a. Cognition reinforces the effect of partisanship on the capacity of identifying fake news.
H3b. People who are not partisan and have a higher cognitive capacity will be more likely to identify fake news than people who are not partisan and have a lower cognitive capacity.
Based on the theory above, overall, it is expected that people will find it hard to identify fake news stories. Partisanship will probably make the capacity of identifying fake news worse, moderating for cognitive capacity. But, cognitive capacity as a main predictor will probably lead to a better capacity of identifying fake news. Therefore, figure 1 provides the theoretical model of this study. In the next section, more details will be provided about how this is tested.
+ +
-
Figure 1. Theoretical model of the effects of partisanship and cognition on the capacity of identifying fake news.
Capacity of identifying fake news Partisanship
Methods
To find out what into what extent partisanship and cognition are of importance for the capacity of identifying fake news, participants conducted an online experiment. More details about the sample, procedure, materials and measures will be provided in this section.
Sample
During two weeks an online experiment was conducted among Dutch participants. To keep the response rate as high as possible, multiple attempts were done to collect respondents via different channels and moments in time. Using a convenience sample, most people were recruited at the University of Amsterdam, via social media and with the help of family and friends. Therefore, no claims can be made in terms of precision for any target population (Scheepers, Tobi, & Boeije, 2016). Answers to all questions were required to finalize the experiment, in order to prevent item non-response. Everyone in the sample is entitled to vote in the Netherlands. Eventually, 194 participants agreed to participate in the study but 60 people had to be excluded because they did not finish the online experiment, 1 person did not agree with the terms and conditions of the experiment and 3 others were not allowed to vote in the Netherlands. Therefore, 130 participants are part of the final sample. All participants are aged 18 years and older and have the Dutch nationality. 61% are female and 37.5% are male. The mean age of participants is 28.33 (SD = 11.43) and the median age is 23. The total age ranges from 19 till 67 years old. Most people were highly educated. 52.2% of the sample had a bachelor’s degree, 28.7% of the sample had a master’s degree and 19.1% of the sample was lower educated (e.g. high school or vocational education). On a scale of (left) 1 till 10 (right) on political orientation most people are slightly leaning to the right (M = 5.12, SD = 1.98).
Materials and procedure
As mentioned before, an online experiment has been conducted. The experiment was created via software program Qualtrics (See Appendix). Using this type of experiment made it possible to investigate a causal relationship with limited recourses, like money and time. Therefore, this was the best method to leave confounding factors out (Scheepers et al., 2016). The experiment could easily be spread among participants and conducted at any time and place.
At the beginning of the experiment, participants had to agree with the terms and conditions. They also had to indicate that their Dutch nationality and be aged above 18 years. If not, they could not proceed with the experiment. Afterward, partisanship was measured by asking participants to indicate on a thermometer scale (0 = cold, 100 = warm) into what extent they had warm feelings towards six politicians, liberals and conservatives, of the six political parties which have the most seats in Dutch parliament (Tweede Kamer, n.d.). The three right-wing political leaders are: Mark Rutte (VVD), Sybrand Buma (CDA) and Geert Wilders (PVV), and three left-wing political leaders are: Jesse Klaver (GroenLinks), Rob Jetten (D66), Lodewijk Asscher (PvdA). On the next page, cognition was measured, based on the scale of Maksl, Ashley & Craft (2015). In the next paragraph, more details will be provided about how cognition and
partisanship were measured.
After partisanship and cognition were measured, participants were randomly assigned to one of the four different conditions, which were: positive fake news, negative fake news, positive real news, and negative real news. In each condition participants were exposed to one news article, real or fake and positive or negative, in the form of Facebook posts which was created via the website thefakenewsgenerator.com. This website offers a tool to create Facebook posts. In the positive articles, real and fake, the
political actor, Mark Rutte, was portrayed favorably. In negative news articles, real and fake, the political actor, Mark Rutte, was portrayed unfavorably. As mentioned before in the theoretical framework, Facebook was chosen as a communication channel because this is the place where fake news occurs a lot (Allcott & Gentzkow, 2017; Pennycook, Cannon & Rand, 2018; Spohr, 2017). Therefore, the experimental setting was made as realistic as possible in order to improve ecological validity. The topic of the content was climate change because this is an issue in which partisanship and cognition can play an important role in processing information. The politician presented in the posts was Dutch prime minister Mark Rutte because in similar studies conducted in the United States, political leaders were also used in Facebook posts (Allcott & Gentzkow, 2017). By also presenting a political leader in the Facebook posts, it is easier to interpret the results in the light of similar studies conducted elsewhere in the world. Based on the definition of fake and real news in the theoretical framework, the content of the Facebook posts was created. The two Facebook posts of the positive real and negative real conditions were factually correct and verified by Dutch news organizations AD (2019) and NOS (2019). Based on these real news articles, two new fake ones were made. This was done by changing a positive real news article into a negative fake article and by changing a negative real news article into a positive fake article. The two Facebook posts of the positive fake and negative fake conditions were not factually correct. Besides that, the fake posts were hard to recognize because the layouts were similar to the real news articles. The fake news posts had a more simplistic language.
After exposure to the Facebook posts, participants had to indicate into what extent they thought the message was positive, negative or real. After that, ‘Gender’, ‘Age’, ‘Education level’ and ‘Political ideology’ was measured. At the end of the study, two manipulation checks were done by asking participants which politician occurred in the Facebook posts and which topic was mentioned. Afterward, people were debriefed
and provided with information about the exact purpose of the study. Participants were still able to decide after finishing the experiment if they wanted to participate in this study or not.
Measures
As was mentioned in the previous paragraph, several measures have been done during the online experiment. Down below more details will be provided of how the capacity of identifying fake news, partisanship and cognition were measured.
Capacity of identifying fake news. To measure the ‘Capacity of identifying fake news’ participants had to indicate to what extent they thought the content of the Facebook message was true. This variable is created based on the way Alcott and Gentzkow (2017) analyzed whether people were capable in recognizing fake news. People were asked to indicate, on a scale from (0 = not true) till (100 = true) into what extent they thought the social media post was true. Overall, participants rated the four Facebook posts, real and fake, as rather untrue (M = 44.79, SD = 28.43).
Cognition. ‘Cognition’ is measured based on the scale which is used in the study of Maksl, Ashley, and Craft (2015). The scale measures automatic versus mindful thought processing, based on the cognitive model of media literacy of James Potter (2004). This scale is suitable to measure the cognitive capacity of people when they are exposed to media content (Maksl, Ashley & Craft, 2015). Participants had to rate five different statements, on a scale from 0 (strongly disagree) till 5 (strongly agree) (See Appendix). A Principal Component analysis with Varimax rotation indicated that the scale was multidimensional. Two components have been identified (Eigenvalue > 1), and there is a clear bow after these components in the scree plot. However, only the first component (Eigenvalue = 2.48) is used for analysis, because the items of this scale match better with the purpose of this study. Together, the factors of this component
explain 49.63% of the variance of the nine original items. Items 1, 2 and 5 were reversed recoded which is why all items correlate positively with the first factor. The variable “I try to avoid situations where I have to think deeply about something” has the strongest association (factor loading = .80). The reliability of the 5-item scale is also good,
Cronbach’s Alpha = .74. The scale cannot be further improved by removing items. On a scale from 1 (strongly disagree) till 5 (strongly agree), participants rather prefer
automatic thought processing than mindful thought processing (M = 2.14, SD = .70). Partisanship. To indicate whether someone is partisan, the evaluation of a politician can be measured which indicates the extent people perceive politicians as favorable or unfavorable (Klein & Ahluwalia, 2005). Therefore, a feeling thermometer is useful (Zavala-Rojas, 2014). It was created to let participants rate their feelings towards a political candidate. In this study, three right-wing political leaders (Mark Rutte, Sybrand Buma and Geert Wilders) and three left-wing political leaders (Rob Jetten, Jesse Klaver and Lodewijk Asscher) were measured, based on a scale from 0 (cold) till 100 (warm). A cold feeling will be interpreted as not supportive for the politician and therefore not partisan, and a warm feeling will be interpreted as supportive for the politician and therefore partisan. Participants had the warmest feelings towards Mark Rutte of VVD (M = 60.76, SD = 22.55), followed by Jesse Klaver of GroenLinks (M = 55.93, SD = 23.61), Rob Jetten of D66 (M = 54.22, SD = 19.79), Lodewijk Asscher of PvdA (M = 51.64, SD = 18.20), Sybrand Buma of CDA (M = 47.11, SD = 19.75). And people had the coldest feelings towards Geert Wilders of PVV (M = 18.66, SD = 21.27).
Other measures. As was already mentioned in the sample section, three control variables were measured: ‘Gender’, ‘Age’, and ‘Education level’. Also, respondents had to indicate their political ideology based on a scale from (0) left till (100) right.
Results
Randomization check
Participants were randomly assigned to ‘condition fake news negative’, ‘condition fake news positive’, ‘condition real news negative’, ‘condition real news positive’ conditions (N=130). To check if the randomization to the conditions was successful, first, a Chi square test was conducted with the conditions as the independent variables and ‘Gender’ and ‘Education’ as dependent variables. There was no effect of condition, χ2 (6) = 4.70, p = .583, on gender. Also, no effect was found of condition, χ2 (9) =7.70, p = .565, on participant’s education level. Second, a One-Way ANOVA with the conditions as independent variable and ‘Age’ as dependent variable was conducted. There was no effect of condition, F (3,125) =.30, p =.835, on participant’s age. This means that there are no differences in the conditions among gender, education level and age. Thus, the randomization for these variables was successful.
Manipulation check
To find out if the manipulation worked as intended to, participants had to rate on a scale of 0 till 100 to what extent they agreed with the following statements: “I think this message contains a positive tone”. A One-Way ANOVA with ‘Conditions’ as independent variable and ‘Rate news as positive’ as dependent variable. There is a quite strong significant effect between the conditions and the extent of how participants rate the positive tone of the Facebook post, F (3, 125) = 24.26, p <.001, η2 =.34. Participants from the negative fake condition (N = 34, M=32.53) and the negative real condition (N=35, M=30.71) rated their Facebook post as not positive. Participants from the positive fake condition (N= 37, M=62.70) and positive real condition (N= 24, M=64.43) did rate the Facebook posts as positive. There was also a check by asking participants “Can you indicate which politician you saw in the Facebook message?”, possible answers were: (1) Mark Rutte, (2) Geert Wilders, (3) Sybrand
Buma and (4) Jesse Klaver. All participants chose the right politician (N=129). To check if participants read the post well, the following question was asked: “What was the topic of the Facebook message?”, possible answers were: (1) Climate, (2) Immigration, (3) Education, (4) Terrorism. 97.7 percent of all participants chose the right answer (N= 126). This indicates that the manipulation and exposure check worked as intended to. Participants who were exposed to a Facebook post with a negative tone indicated that the post was negative, and participants who were exposed to a positive tone indicated that they saw indeed a post with a positive tone. Also, participants were aware of the content of the post which meant they knew the politician presented, Mark Rutte, and the topic, climate change.
The capacity of identifying fake news
On average, participants in the sample rated fake news not in favor of Mark Rutte as rather untrue (M = 35.86, SD = 18.19) but participants who had seen fake news in favor of Mark Rutte rather rated it as they do not know whether it was true or untrue (M = 48.26, SD = 29.22). A one sample t-test revealed that this result does significantly differ, t (72) = -1.85, p = .034, 95% CI [-25.74, .94], d = 0.04. Participants rated fake news not in favor of Mark Rutte as less true (M = 35.86, SD = 18.19) than real news not in favor of Mark Rutte (M = 48.44, SD = 24.89). A one sample t-test revealed that his result does significantly differ, t (72) = -2.04, p = .023, 95% CI [-24.88, -.28], d = 0.05. Participants rated fake news in favor of Mark Rutte as truer (M = 48.26, SD = 29.22) than real news in favor of Mark Rutte (M = 46.25, SD = 31.70). A one sample t-test revealed that this result does not significantly differ, t (61) = .26, p = .400, 95% CI [-13.65, 17.66], d = 0.07.
Figure 2. The capacity of identifying fake news of participants. Rated on a scale from 0 (not true) till 100 (true).
In hypothesis 1 it was expected that in general, people would rather fail in identifying fake news. As figure 2 shows, people indicate fake news not in favor of Mark Rutte as rather fake, but fake news in favor of Mark Rutte is not indicated as untrue. Besides that,
participants were also not capable of identifying real news as real. Therefore, it seems they are not able to distinguish fake from real news articles. This indicates that people’s capacity of identifying fake news is not that high. Therefore, hypothesis 1 is accepted.
Effects of partisanship and cognition on identifying fake news
Before conducting the analysis, three dummy variables were created for ‘Conditions’. A multiple regression analysis was conducted with ‘age’, ‘education level’, ‘gender’,
‘condition fake news negative’ , ‘condition fake news positive’, ‘condition real news 0 10 20 30 40 50 60 70 80 90 100
Fake news negative Fake news positive Real news negative Real news positive
P er ce iv ed ac cu rac y o f th e n ew s sto ry
negative’, ‘Thermometer Mark Rutte’, ‘Thermometer Sybrand Buma’, ‘Thermometer Geert Wilders’, ‘Thermometer Rob Jetten’, ‘Thermometer Jesse Klaver’, ‘Lodewijk Asscher’ (0 = cold, 100 = warm), ‘Cognition’ (0 = automatic processing, 5 = mindful processing) ,
‘condition fake news negative*need for cognition’, ‘condition fake news positive*need for cognition’ and ‘condition real news negative*need for cognition’ as independent variables and ‘Capacity of identifying fake news’ (0 = as dependent variable. The regression model with gender, education level, age, partisanship, cognition, and the extent to which someone thinks a message is real is not significant, F (16, 111) = 1.02, p = 0.446. The model can therefore not be used to predict whether someone will identify a fake or real news message as true.
However, one predictor, ‘Thermometer Rob Jetten’ was significant, b* = - 0.26, t = - 2.10, p = 0.038, 95% CI [- 0.78, -0.23]. For each additional point on the thermometer scale of Rob Jetten, the capacity of identifying fake news decreases by 0.40. Thus, the warmer feelings people had towards Rob Jetten, the less likely they thought a Facebook post about Mark Rutte was true. For this prediction the other variables are assumed to be held constant. In
hypotheses 2a it was expected that the more partisan people are, the more likely they will believe a fake story is true, especially when their preferred political candidate is portrayed in their favor. These results do not indicate partisan people rather believe a fake story is true. However, people who liked Rob Jetten indicated that in general, they rate Facebook posts as untrue, which could mean they are not capable to identify fake from real. Therefore,
hypothesis 2a is partly accepted. Hypothesis 3a and 3b that cognition as a moderator would reinforce the effect of partisanship on the capacity of identifying fake news, and that cognition as a main predictor would have a positive outcome on the capacity of identifying fake news. However, no significant results concerning cognition are found. Therefore, hypothesis 3a and b are rejected.
Liberals versus conservatives. To compare liberals with conservatives, the variable ‘Politial orientation’ (0 = left, 100 = right) was recoded into a dummy variable, in which participants who rated themselves as 3.33 or less are considered as liberal and higher than 6.66 are considered as conservative. Overall, liberals (M = 41.86, SD = 31.56) rated news stories as less true than conservatives (M = 44.16, SD = 29.39). Liberals (M = 38, SD = 32.93) believed fake news not in favor of Mark Rutte slightly less than conservatives (M = 38.17, SD = 28.12). Liberals believed fake news in favor of Mark Rutte more (M = 47.63, SD = 27.86) than conservatives did (M = 39.38, SD = 33.58). In addition, conservatives believed real news not in favor of Mark Rutte (M = 48.27, SD = 20.97) more than liberals did (M = 39.50, SD = 36.21). And conservatives (M = 70, SD = 29.44) believed real news in favor of Mark Rutte more than liberals did (M = 43.67, SD = 32.35).
Figure 3. The capacity of identifying fake news of conservatives and liberals. Rated on a scale from 0 (not true) till 100 (true).
0 10 20 30 40 50 60 70 80 90 100
Fake news negative Fake news positive Real news negative Real news positive
P er ce iv ed ac cu rac y o f th e n ew s so tr y Conservatives Liberals
A two-factor analysis of variance was carried out to assess the influence of exposure to fake or real news, in conjunction with the effect of political ideology on the capacity of
identifying fake news. There is no significant effect of political ideology on the capacity of identifying fake news, F (1,72) = .74, p = .363, Eta-squared = .01. In hypothesis 2b it was expected that conservatives were more likely to identify fake news as untrue than liberals. As Figure 3 shows, conservatives were better in identifying fake news in favor of Mark Rutte. They were also better in identifying real news as true. Although these differences are not significant, the indication that conservatives are more likely to identify fake news better than liberals is there. Therefore, hypothesis 2b is partly accepted.
Conclusion and discussion
Facebook is considered as one of the most important online platforms where political debate take place (Bouma, 2017). At the same time, one could question itself to what extent Facebook is a good facilitator for democracy. As results of this study also indicate, it is hard for citizens to identify fake news on Facebook, which is in line with studies that have been conducted in the United States (Allcott & Gentzkow, 2017; Kim & Dennis, 2018; Nielsen & Graves, 2017; Pennycook et al., 2018). A possible reason for this is the lack of trust of the people in the platform itself; 56 percent of the Dutch people distrust social media as a news source (NU.nl, 2019). Which is why it could be possible that participants in this study were more likely to think that news in general, no matter if it was real or not, on Facebook is rather untrue. Therefore, to find out if the amount of trust in social media does have an effect on the capacity of identifying fake news, future research is required. Furthermore, this study found that fake news in favor of a politician was harder to identify as untrue than fake news not in favor of a politician. In addition, participants rated real news not as true, which means that they were not capable of distinct fake news from real news. In contrast to other studies
concerning the effects of cognition and partisanship on identifying fake news, this study did not find any significant effects of these two variables. However, concerning partisanship, there are indications that conservatives do have a higher capacity of identifying fake news than liberals, which is also what is found in the United States (Allcott & Gentzkow, 2017; Pennycook et al., 2018). An explanation of why no significant results were found can have many reasons. For example, fake news in the United States does have another effect on the population compared with other Western countries due to differences in the political system and media environment (Humprecht, 2018). Therefore, it could be argued that fake news in the Netherlands does not have that much of an impact as it has in the United States. However, it would also be reasonable that effects only occur over a longer period of time like
Pennycook, Cannon, and Rand (2018) found in their study. Concerning no significance for cognition, another cause is possible. While the scale used in this study, is to determine whether people are automatic or analytical processors, Pennycook, Cannon, and Rand (2018) did a Cognitive Reflection Test (CRT). And as they mention in the limitations of their paper CRT is not necessarily a variable to measure into what extent people think analytically, but rather, a measure of cognitive ability. Which is why thinking analytically does not have an effect on identifying fake news, but the cognitive ability does. To find out how these
mechanisms exactly have an influence on the capacity of fake news need to be investigated in future research.
Like in any other paper, there are some limitations that need to be addressed. Probably the most important one of all is that the results of this paper are based on a convenience sample. Therefore, one should keep this in mind for the interpretations of the results because it is not representative of the Dutch population. However, like Pennycook, Cannon and Rand (2018) also argued in their limitations, using a convenience sample does not have to mean that the results do not suggest anything at all. Studies show that the results which occur by a
convenience sample are, up to a certain extent, similar to results which occur in a
representative sample (Mullinix, Leeper, Druckman & Freese, 2015). However, to find out if the results of this study could be an indication of how the Dutch population identifies fake news should be investigated in future research by using a representative sample. A second limitation of this paper is the scale which is used to measure cognition (Maksl, Ashley, and Craft, 2015). The scale was particularly developed for teenagers while in this study, no teenagers were part of the sample at all. Still, this scale was developed with a focus on media literacy and offers inside into how people process media content in relation to how willing someone is to think deeply about things. For future research into identifying fake news, it would be interesting to find out if significant effects would occur, using this scale only among teenagers. The third limitation is the measurement of partisanship, which was done by using a feeling thermometer. According to Greene (2002), this is a valid and effective measure to define attitudes towards political parties. However, this study asked participants to rate their partisanship based on the political party in combination with the political leader. Therefore, one could assume that this measurement leans more towards an evaluation of a politician, rather than identification or affiliation with a political party. On the other hand, it is argued that evaluations of political candidates are in a direct connection with the issues of a political party (Hayes, 2005). The fourth limitation concerns the stimuli used in this study. Although the experiment was conducted online, like Pennycook, Cannon and Rand (2018) also argued, at the same time it made it a more natural setting because Facebook posts are of course only online available. However, the source used in the post was from an American news website. No source of a Dutch fake news website was used, which could affect the validity. On the other hand, the source of a website on Facebook is not dominantly displayed which is why people probably do not pay attention to this. However, future research into the impact of the source and identifying fake news among Dutch people should find out whether other effects
occur when a Dutch (fake) news website is mentioned as a source. Besides that, the content which is used as fake news is about climate change and Mark Rutte. However, there are a lot of other issues in which motivated reasoning can play an important role (Taber & Lodge, 2006). Therefore, it is not known to what extent the topic, but also the political candidate presented in the fake news post, have an influence on the capacity of identifying fake news. Future research should take that into account as well.
Although this study was conducted in a limited amount of time, this master thesis made an attempt to gain more insight into the effects of partisanship and cognition on the capacity of identifying fake news. As the results indicate that people find it hard to recognize fake from real, it means that fake news needs to be taken seriously in the Netherlands as well. Because when we look at the United States, this can have serious implications for society. As addressed in the introduction, this could mean that it is very likely Dutch people make their political decisions based on incorrect information. Social media platforms already took measures against the spread of fake news like removing accounts, and use fact-checkers (Allcott & Gentzkow, 2017; Guess, Nyhan & Reifler, 2018). However, one could question into what extent these methods are working effectively since it is still very easy to spread fabricated fake news (NRC, 2018). According to Lazer, Baum, Benkler, Berinsky, Greenhull, Menczer & Schudson (2018), the whole system of how news is delivered and presented should be redesigned. However, until the news system is not improved yet, people have to be capable of identifying fake news as untrue, and real news as true. But that leads to a broader question, like Allcott & Gentzkow (2017) also wondered: what makes a story true, and, who will decide that?
References
AD. (2019, February 7). Rutte: betrokkenheid scholieren bij klimaatmars is fantastisch. Retrieved June 27, 2019, from https://www.ad.nl/politiek/rutte-betrokkenheid-scholieren-bij-klimaatmars-is-fantastisch~a86fb85b/
Albright, J. (2017). Welcome to the Era of Fake News. Media and Communication, 5(2), 87. https://doi.org/10.17645/mac.v5i2.977
Allcott, H., & Gentzkow, M. (2017). Social Media and Fake News in the 2016 Election. Journal of Economic Perspectives, 31(2), 211–236.
https://doi.org/10.1257/jep.31.2.211
Balmas, M. (2014). When fake news becomes real: Combined exposure to multiple news sources and political attitudes of inefficacy, alienation, and cynicism. Communication Research, 41(3), 430-454.
Bardoel, J. (1996). Beyond Journalism. European Journal of Communication, 11(3), 283– 302. https://doi.org/10.1177/0267323196011003001
Baten, M. (2018, February 9). Alle laatste cijfers over het gebruik van Facebook in 2018. Retrieved June 27, 2019, from https://www.socialmediaacademie.nl/alle-laatste-cijfers-over-het-gebruik-van-facebook-in-2018/
Bolsen, T., Druckman, J. N., & Cook, F. L. (2014). The influence of partisan motivated reasoning on public opinion. Political Behavior, 36(2), 235-262.
Bouma, R. (2017, 24 februari). Nepnieuws en sociale media als wapen in politieke
campagnes. Retrieved June 18, 2019, from https://nos.nl/nieuwsuur/artikel/2159802-nepnieuws-en-sociale-media-als-wapen-in-politieke-campagnes.html
Bronstein, M., Pennycook, G., Bear, A., Rand, D. G., & Cannon, T. (2018). Reduced Analytic and Actively Open-Minded Thinking Help to Explain the Link between Belief in Fake News and Delusionality, Dogmatism, and Religious Fundamentalism. SSRN
Electronic Journal. https://doi.org/10.2139/ssrn.3172140
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2e ed.). Hillsdale, NJ: Lawrence Erlbaum Associates
De nieuwe reporter. (2019, 9 april). Een overheidscampagne tegen nepnieuws draagt niet bij aan de bescherming van onze democratie - De Nieuwe Reporter. Retrieved June 18, 2019, from https://www.denieuwereporter.nl/2019/04/een-overheidscampagne-tegen-nepnieuws-draagt-niet-bij-aan-de-bescherming-van-onze-democratie/
Deuze, M. (2005). What is journalism? Journalism: Theory, Practice & Criticism, 6(4), 442– 464. https://doi.org/10.1177/1464884905056815
De Volkskrant. (2017, February 25). Turks-Nederlandse artsen laken opmerkingen Kuzu: “Ongefundeerd en schadelijk.” Retrieved June 27, 2019, from
https://www.volkskrant.nl/nieuws-achtergrond/turks-nederlandse-artsen-laken-opmerkingen-kuzu-ongefundeerd-en-schadelijk~b0f2e587/
Ferree, M. M., Gamson, W. A., Gerhards, J., & Rucht, D. (2002). Four models of the public sphere in modern democracies. Theory and society, 31(3), 289-324.
Flynn, D. J., Nyhan, B., & Reifler, J. (2017). The nature and origins of misperceptions: Understanding false and unsupported beliefs about politics. Political Psychology, 38, 127-150.
Greene, S. (2002). The social-psychological measurement of partisanship. Political Behavior, 24(3), 171-197.
Guess, A., Nyhan, B., & Reifler, J. (2018). Selective exposure to misinformation: Evidence from the consumption of fake news during the 2016 US presidential
campaign. European Research Council, 9.
Hart, P. S., & Nisbet, E. C. (2012). Boomerang effects in science communication: How motivated reasoning and identity cues amplify opinion polarization about climate mitigation policies. Communication Research, 39(6), 701-723.
Hayes, D. (2005). Candidate qualities through a partisan lens: A theory of trait ownership. American Journal of Political Science, 49(4), 908-923.
Horne, B. D., & Adali, S. (2017, May). This just in: fake news packs a lot in title, uses simpler, repetitive content in text body, more similar to satire than real news. In Eleventh International AAAI Conference on Web and Social Media.
Het Parool. (2018, 22 maart). Exitpoll: driekwart van de Turkse kiezers stemt op Denk. Retrieved June 18, 2019, from https://www.parool.nl/nieuws/exitpoll-driekwart-van-de-turkse-kiezers-stemt-op-denk~b1259b1b/
Humprecht, E. (2018). Where ‘fake news’ flourishes: a comparison across four Western democracies. Information, Communication & Society, 1-16.
Kahan, D.M. (2013). Ideology, motivated reasoning, and cognitive reflection. Judgment and Decision Making, 8 (4), 407-424.
Kim, A., & Dennis, A. (2018, January). Says Who?: How News Presentation Format
Influences Perceived Believability and the Engagement Level of Social Media Users. In Proceedings of the 51st Hawaii International Conference on System Sciences. Klein, Jill G., and Rohini Ahluwalia. 2005. “Negativity in the evaluation of political
candidates.” Journal of Marketing, 69, 131-42.
Lazer, D. M., Baum, M. A., Benkler, Y., Berinsky, A. J., Greenhill, K. M., Menczer, F., ... & Schudson, M. (2018). The science of fake news. Science, 359(6380), 1094-1096.
Maksl, A., Ashley, S., & Craft, S. (2015). Measuring news media literacy. Journal of Media Literacy Education, 6(3), 29-45.
McCright, A. M., & Dunlap, R. E. (2011). The politicization of climate change and polarization in the American public's views of global warming, 2001–2010. The Sociological Quarterly, 52(2), 155-194.
Miller, J. M., Saunders, K. L., & Farhart, C. E. (2016). Conspiracy endorsement as motivated reasoning: The moderating roles of political knowledge and trust. American Journal of Political Science, 60(4), 824-844.
Ministerie van Algemene Zaken. (2017, 14 november). Kamerbrief over beïnvloeding van de publieke opinie door statelijke actoren. Retrieved June 18, 2019, from
https://www.rijksoverheid.nl/documenten/kamerstukken/2017/11/13/kamerbrief-over-beinvloeding-van-de-publieke-opinie-door-statelijke-actoren
Mullinix, K. J., Leeper, T. J., Druckman, J. N., & Freese, J. (2015). The generalizability of survey experiments. Journal of Experimental Political Science, 2(2), 109-138.
Newman, N. (n.d.). Executive Summary and Key Findings of the 2019 Report. Retrieved June 24, 2019, from http://www.digitalnewsreport.org/survey/2019/overview-key-findings-2019/
Nielsen, R. K., & Graves, L. (2017). News you don’t believe”: Audience perspectives on fake news. Reuters Institute for the Study of Journalism, Oxford, Oct, 2017-10.
NOS. (2019, February 7). Rutte tegen jongeren: “We doen al veel aan klimaat, vraag niet meer.” Retrieved June 27, 2019, from https://nos.nl/artikel/2270913-rutte-tegen-jongeren-we-doen-al-veel-aan-klimaat-vraag-niet-meer.html
NOS. (2017, 23 februari). 'Beschuldiging Kuzu aan artsen raakt kant noch wal'. Retrieved June 18, 2019, from https://nos.nl/artikel/2159666-beschuldiging-kuzu-aan-artsen-raakt-kant-noch-wal.html
NOS. (2018, 14 juni). Nederlanders maken zich weinig zorgen over nepnieuws. Retrieved June 18, 2019, from van https://nos.nl/artikel/2236391-nederlanders-maken-zich-weinig-zorgen-over-nepnieuws.html
NRC. (2018, October 19). Nederlands nepnieuws via Facebook nog makkelijk te verspreiden. Retrieved June 20, 2019, from https://www.nrc.nl/nieuws/2018/10/19/nederlands-nepnieuws-via-facebook-nog-makkelijk-te-verspreiden-a2673603
Nu.nl. (2019, May 12). Meerderheid nieuwsconsumenten wantrouwt sociale media als nieuwsbron. Retrieved June 27, 2019, from
https://www.nu.nl/media/5882901/meerderheid-nieuwsconsumenten-wantrouwt-sociale-media-als-nieuwsbron.html
Pennycook, G., Cannon, T. D., & Rand, D. G. (2018). Prior exposure increases perceived accuracy of fake news. Journal of experimental psychology: general.
Pennycook, G., & Rand, D. G. (2017). Who falls for fake news? The roles of analytic thinking, motivated reasoning, political ideology, and bullshit receptivity. SSRN Electronic Journal.
Rini, R. (2017). Fake News and Partisan Epistemology. Kennedy Institute of Ethics Journal, 27(2S), 43–64. https://doi.org/10.1353/ken.2017.0025
Roets, A. (2017). ‘Fake news’: Incorrect, but hard to correct. The role of cognitive ability on the impact of false information on social impressions. Intelligence, 65, 107-110.
RTL Nieuws. (2014, 21 juli). Wat is er gebeurd met MH17? Negen vragen. Retrieved June 18, 2019, from https://www.rtlnieuws.nl/nederland/artikel/1755876/wat-er-gebeurd-met-mh17-negen-vragen
Scheepers, P., Scheepers, P. L. H., Tobi, H., & Boeije, H. R. (2016). Onderzoeksmethoden (9th ed.) Amsterdam: Boom.
Spohr, D. (2017). Fake news and ideological polarization: Filter bubbles and selective exposure on social media. Business Information Review, 34(3), 150-160.
Statista. (z.d.). Global social media ranking 2019 | Statistic. Retrieved June 18, 2019, from https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-of-users/
Strömbäck, J. (2005). In search of a standard: Four models of democracy and their normative implications for journalism. Journalism Studies, 6(3), 331-345
Taber, C., & Lodge, M. (2006). Motivated skepticism in the evaluations of political beliefs. American Journal of Political Science, 50 (3),755-769
Tandoc, E. C., Lim, Z. W., & Ling, R. (2017). Defining “Fake News”. Digital Journalism, 6(2), 137–153. https://doi.org/10.1080/21670811.2017.1360143
Tweede Kamer. (n.d.). Fracties | Tweede Kamer der Staten-Generaal. Retrieved June 26, 2019, from https://www.tweedekamer.nl/kamerleden_en_commissies/fracties Van Bavel, J. J., & Pereira, A. (2018). The partisan brain: An Identity-based model of
political belief. Trends in cognitive sciences, 22(3), 213-224.
Vlooswijk, E. (2018, 28 mei). Zien acht op de tien jongeren geen verschil tussen nep- en echt nieuws? Retrieved June 18, 2019, from
https://www.volkskrant.nl/nieuws- achtergrond/zien-acht-op-de-tien-jongeren-geen-verschil-tussen-nep-en-echt-nieuws~b4a326bb/
Zavala-Rojas, D. (2014). Thermometer scale (feeling thermometer). Encyclopedia of quality of life and well-being research, 6633-6634.
Appendix
[Introduction, informed consent] Beste deelnemer,
Graag nodig ik u uit om deel te nemen aan mijn masterscriptie-onderzoek dat wordt uitgevoerd onder de verantwoordelijkheid van onderzoeksinstituut Amsterdam School of Communication Research (ASCoR), onderdeel van de Universiteit van Amsterdam. In dit onderzoek vragen we naar uw mening over een Facebookbericht dat gaat over Mark Rutte. Het volledige onderzoek duurt niet langer dan 5 minuten.
Aangezien dit onderzoek wordt uitgevoerd onder de verantwoordelijkheid van de Amsterdam School of Communication Research, kunnen we het volgende garanderen:
1. Uw antwoorden zijn vertrouwelijk en we delen uw persoonlijke gegevens niet met derden, tenzij u hier eerst nadrukkelijk toestemming voor verleent. Om uw gegevens te beschermen, zal de opgeslagen data geen informatie bevatten die u kan identificeren en zullen alle gegevens als collectief geanalyseerd worden.
2. Deelname aan dit onderzoek is vrijwillig. U kunt op ieder moment besluiten om af te zien van deelname aan dit onderzoek. U kunt binnen 7 dagen na het onderzoek verzoeken om uw onderzoeksgegevens te laten verwijderen.
3. Deelname aan dit onderzoek brengt geen noemenswaardige risico’s of ongemakken met zich mee. Ook zult u niet met expliciet aanstootgevend materiaal worden geconfronteerd. Voor meer informatie kunt u contact opnemen met Petra van Gelsdorp per adres:
petra.vangelsdorp@student.uva.nl.
Mocht u naar aanleiding van uw deelname aan dit onderzoek toch nog klachten of
opmerkingen hebben over het verloop van het onderzoek en de daarbij gevolgde procedure, dan kunt u contact opnemen met het lid van de Commissie Ethiek namens ASCoR, per adres: ASCoR secretariaat, Commissie Ethiek, Universiteit van Amsterdam, Nieuwe Achtergracht 166, 1018 WV te Amsterdam; 020-525 36980; ascor-secr-fmg@uva.nl. Een vertrouwelijke behandeling van uw klacht of opmerking is daarbij gewaarborgd.
Ik hoop u hiermee voldoende te hebben geïnformeerd. Alvast hartelijk dank voor uw deelname.
Ik verklaar hierbij op voor mij duidelijke wijze te zijn ingelicht over de aard en methode van het onderzoek, zoals uiteengezet in de informatie op de vorige pagina.
Ik stem geheel vrijwillig in met deelname aan dit onderzoek. Ik behoud daarbij het recht deze instemming weer in te trekken binnen 7 dagen na deelname zonder dat ik daarvoor een reden hoef op te geven. Ik besef dat ik op elk moment mag stoppen met het onderzoek.
Als mijn onderzoeksresultaten gebruikt worden in wetenschappelijke publicaties, of op een andere manier openbaar worden gemaakt, dan zal dit volledig geanonimiseerd gebeuren. Mijn persoonsgegevens worden niet door derden ingezien zonder mijn uitdrukkelijke toestemming.
Als ik meer informatie wil, nu of in de toekomst, dan kan ik me wenden tot Petra van Gelsdorp, per adres: petra.vangelsdorp@student.uva.nl.
Voor eventuele klachten over dit onderzoek kan ik me wenden tot het lid van de Commissie Ethiek namens ASCoR, per adres: ASCoR secretariaat, Commissie Ethiek, Universiteit van Amsterdam, Postbus 15793 1001 NG Amsterdam; 020-5253680; ascor-secr-fmg@uva.nl. Ik begrijp de bovenstaande tekst en ga akkoord met deelname aan het onderzoek.
• Ja • Nee
Ik ben kiesgerechtigd. Dat betekent dat ik 18 jaar of ouder ben en bezit over de Nederlandse nationaliteit.
• Ja • Nee
[Partisanship - Thermometer politicians]
Hieronder ziet u zes Nederlandse politieke leiders.
Geef op een schaal van 0 (zeer negatief) tot 100 (zeer positief) aan welk gevoel u heeft bij deze politici.
Zeer negatief Neutraal Zeer positief
0 10 20 30 40 50 60 70 80 90 100
• Lodewijk Asscher - PvdA • Rob Jetten – D66
• Jesse Klaver - GroenLinks • Sybrand Buma - CDA • Geert Wilders - PVV • Mark Rutte - VVD
[Cognition – scale need for cognition]
Geef op een schaal van 1 (helemaal mee oneens) tot 5 (helemaal mee eens) in hoeverre u het eens bent met de onderstaande stellingen.
Helemaal mee oneens Helemaal mee eens
0 1 2 3 4 5
• Ik vind het niet leuk om veel te moeten nadenken.
• Ik probeer situaties te vermijden waarin ik diepgaand over iets moet nadenken. • Ik geef de voorkeur aan iets wat mijn denkvermogen uitdaagt, in plaat van iets wat
weinig denkvermogen vereist.
• Ik denk liever over complexere problemen in plaats van simpelere problemen. • Veel en lang nadenken geeft mij weinig voldoening.
[Randomizer into one of the four conditions]
Hieronder staat een Facebookbericht weergegeven waarin gelinkt wordt naar een nieuwsartikel. Lees het artikel aandachtig door.
[Treatment real news negative, source: https://nos.nl/artikel/2270913-rutte-tegen-jongeren-we-doen-al-veel-aan-klimaat-vraag-niet-meer.html
]
[Treatment fake news negative]
[Treatment real news positive, source:
https://www.ad.nl/politiek/rutte-betrokkenheid-scholieren-bij-klimaatmars-is-fantastisch~a86fb85b/]
Geef op een schaal van 0 tot 100 aan in hoeverre u dit bericht positief of negatief van toon vindt. Daarnaast wordt gevraagd om aan te geven in hoeverre u de inhoud van dit bericht nep of echt vindt.
Met nep wordt bedoeld dat de inhoud van het bericht feitelijke onjuistheden bevat, met echt wordt bedoeld dat de inhoud van het bericht feitelijk juist is.
Helemaal mee oneens Helemaal mee eens
0 10 20 30 40 50 60 70 80 90 100
• Ik vind dit een positief bericht • Ik vind dit een negatief bericht • Ik vind dit een echt bericht [Control variables]
Wat is uw geslacht? • Man
• Vrouw Hoe oud bent u?
Bent u bijvoorbeeld 34 jaar, vul dan alleen het getal ‘34’ in. …..
Wat is uw hoogst genoten opleiding? • Geen onderwijs / basisonderwijs
• Lbo/vbo/vmbo (kader- en beroepsgerichte leerweg)
• Eerste 3 jaar havo, vwo, mavo of vmbo (theoretisch en gemengde leerweg) • Mbo
• Havo en vwo bovenbouw / wo en hbo propedeuse • Hbo of wo bachelor of kandidaats
• Wo doctoraal of master
Op een schaal van 1 (links) naar 10 (rechts), wat is uw politieke oriëntatie?
Links Rechts
0 1 2 3 4 5 6 7 8 9 10
[Manipulation check]
Kunt u aangeven welke politicus u heeft gezien in het Facebookbericht? • Mark Rutte
• Geert Wilders • Sybrand Buma • Jesse Klaver
Wat was het onderwerp van het Facebookbericht? • Klimaat
• Immigratie • Onderwijs • Terrorisme [Debriefing]
De vragen die u net heeft beantwoord, maakten deel uit van een onderzoek dat bekijkt in hoeverre Nederlanders in staat zijn om nepnieuws te herkennen. Als deelnemer van deze studie werd u willekeurig toegewezen aan één nep- of echt nieuwsbericht. Hieronder ziet u de vier verschillende berichten die gebruikt worden in dit onderzoek. Daarvan zijn twee berichten nep en twee berichten echt. De berichten in de bovenste rij bevatten een positieve toon. De berichten in de onderste rij bevatten een negatieve toon.
Mocht u meer informatie hierover willen ontvangen, neem dan gerust contact op via
petra.vangelsdorp@student.uva.nl. Voor andere vragen, klachten of opmerkingen over het onderzoek kunt u contact opnemen met het ASCoR secretariaat, Commissie Ethiek, Universiteit van Amsterdam, via: Nieuwe Achtergracht 166, 1018 WV te Amsterdam; 020-525 36980; ascor-secr-fmg@uva.nl. Een vertrouwelijke behandeling van uw klacht of opmerking is daarbij gewaarborgd.
Bedankt voor uw deelname aan dit onderzoek. Mocht u nog iemand kennen die ook bereid is om deze vragenlijst in te vullen, dan zou het fijn zijn als u de onderstaande link zou kunnen delen.
https://uvacommscience.eu.qualtrics.com/jfe/form/SV_eKHkekISSHarId7
