This study explores youth accuracy judgments of disinformative and nondisinformative claims. Analyses are based on a nationally representative youth (16–20 years old) survey experiment conducted in the Czech Republic in 2017. When they were exposed to posts regarding refugee crisis, young people were asked to judge the accuracy of the statements accompanying the posts. Motivated reasoning of youth depended primarily on the alignment with the posts and the ideology of participants. Results of this research suggest that motivated reasoning works differently for liberals and conservatives. Perceived amount of media literacy training does not seem to affect directional motivation. General trust works as moderator of motivated reasoning and, in combination with ideology, appears to be important for understanding directional motivation when exposed to disinformation.

INTRODUCTION

The rise of fake news and disinformation is a worldwide phenomenon. Although it is most visible in the USA due to this country's importance in world politics, there is a huge rise of fake news and disinformation in Europe as well. It is even more evident in post-communist countries where we observe a particularly high activity of pro-Russian disinformation servers. This disinformation is often pro-Russian or against the EU, USA, Muslims, or refugees (Čulík, 2017). Articles about the dangers of Muslim refugees are very frequent on Czech media servers as well as on fake news sites and are massively shared by users of social network sites.

The spread of disinformation is not usually a consequence of innocent ignorance, but there are people and organizations that benefit from sharing disinformation. This particularly applies to populist-leaning politicians who benefit from their anti-refugee and anti-EU attitudes (Čulík, 2017). For example, invoking fear of Islamism and waves of refugees, who cannot be stopped because of Czech EU membership, became standard in populist speeches and Facebook posts of several politicians (Čulík, 2015).

Since such disinformation is spread all over social media sites, media literacy and political literacy soon became important topics. The public debate about media literacy and political literacy primarily goes along with expected civic education reform. The aim of awaited civic education reform should ensure that youth will get necessary information about politics and help them in making informed decisions. Similarly, media literacy training should provide young Czechs with the necessary skills to recognize disinformation and to be less prone to trust information produced by fake news servers.

In this article it is assumed that democracy works better if citizens have access to information and can make decisions based on correct information. In this sense, use and share of disinformation is seen as undesirable because citizens will not be able to make informed decisions based on real facts; they may become demotivated to be politically active, and their belief in legitimacy of democratic governance may be undermined (Delli Carpini & Keeter, 1993; Hochschild & Einstein, 2015; Kahne & Bowyer, 2017).

Using a representative sample of 16- to 20-year-old Czech high school students (N = 1207), a survey experiment that tests youth's judgments of the accuracy of truth claims was conducted. It is tied to an important public issue of (not) accepting refugees to the Czech Republic. This topic has become a very salient and frequently used political issue in the past two years.

We used an experiment setting similar to Kahne and Bowyer's (2017), trying to discover if youth assessments depend on (a) ideological alignment of the claim, (b) whether the statement is factually correct, and (c) youth's own political ideology. If it was true that the youth's approval of the claim depended on factual correctness and ideological framing, then civic education reform and training that aimed at improving political knowledge, media literacy, and general trust supposedly would help to diminish the effect of disinformation.

A THEORY OF MOTIVATED REASONING

Perception of any information is affected by the individual's value system. There is evidence that basic human values relate to subjective situation experiences and behavior (Schwartz, 1994). An individual's values and ideology are important because they act as filters on how perceived information is interpreted. This biased assessment of information is known as motivated reasoning.

The basic accepted premise is that all reasoning is motivated (Kunda, 1990). When individuals process information, they do not completely control their perceptions and are influenced by their values and goals. There are basically two broad categories of goals: (1) accuracy goals that motivate individuals to reach the correct answer and (2) directional goals that motivate individuals to use their reasoning to defend their own specific attitudes in the face of attitude-challenging information. People seek to justify conclusions that are in agreement with their attitudes (Taber & Lodge, 2006).

According to Kunda (1990), individuals have to use more cognitive effort to choose the correct arguments and process all the information more exhaustively when they are motivated by accuracy goals. Directionally motivated reasoning, which is common in processing political information, works differently. Individuals who use directional motivation tend to uncritically accept arguments that are aligned with their prior beliefs and tend to disagree or agree less with those that are contrary to their prior attitudes (Ditto et al., 1998).

Most fake news and disinformation is political and usually controversial. If individuals are exposed to ideological issues, their judgment of the truth claim is influenced by several perceptual biases. One of the information processes is sometimes called “hot cognition.” Hot cognition refers to a situation when feelings about the statement automatically come to mind and information is being processed with biases (Lodge & Taber, 2005). This process, which leads individuals to seek evidence that supports their prior beliefs, is called a confirmation bias (Kunda, 1990). Motivated reasoning is the tendency to accept new information that agrees with an individual's values and to be critical of information that does not support them. Although motivated reasoning and confirmation bias are not the same, in reality there is an interplay between these two types of biases when individuals are exposed to new information. In this article, political knowledge, media literacy training, and general trust are considered to have a potentially mitigating effect on confirmation bias among youth.

Political Knowledge

The association between political knowledge and confirmation bias has already been analyzed by several scholars (Taber & Lodge, 2016; Kahne & Bowyer, 2017), and contrary to expectations the effect of political knowledge does not seem to be in favor of mitigating the confirmation bias. Taber and Lodge (2006) found out that directional motivation bias is greatest among highly knowledgeable individuals. They conducted an experiment in which they gave the same number of arguments in favor and against an individual's argument. Individuals who possessed high political knowledge were more likely to be biased because they preferred to look at a higher proportion of aligned arguments rather than looking at a balanced number of pro and con arguments as low-sophisticated respondents did.

The effect of political knowledge on confirmation bias seems positive. The reason why politically sophisticated individuals should be more affected by confirmation bias is that highly knowledgeable individuals have the cognitive capacity to successfully counter the arguments that are contrary to their own beliefs and therefore tend to have a greater bias in favor of arguments aligned with their attitudes. This, however, is not what Kahne and Bowyer (2017) found in their study of American youth. Their conclusion was that although political knowledge did not improve judgment of accuracy, media literacy education did.

Media Literacy

Media literacy education became an important topic with the rise of fake news servers and disinformation on the internet and in media in general. News has never been more accessible. The Internet and particularly social media allow people to select the stories they want to read. People may get to fake news stories due to following fake news servers, reaching the story because it was shared by their friends or as a consequence of a paid advertisement. The incentive to click on the story to read it is greater if the story title is dramatic or shocking. This phenomenon, along with general changes in the media world, such as the Internet environment that makes choosing stories far easier, leads to an increase in partisan polarization (Barber & McCarty, 2013). Research in this field also shows that reading a story aligned with an individual's beliefs reinforces attitudinal polarization (Stroud, 2011; Iyengar & Hahn, 2009; Prior, 2013). Moreover, social networks are usually ideologically homogeneous. This motivates people to share aligned politically motivated disinformation even more because they are encouraged by being surrounded by like-minded readers (Rojecki & Meraz, 2016).

Because of changes in the media environment, people are more likely to act on confirmation biases. This is due to a combination of highly polarized news environment and readership that makes individuals' judgments more likely to be driven by directional motivation (Kahne & Bowyer, 2017). The likelihood of being driven by directional motivation may be mitigated by media literacy training. If people were informed how to recognize an untruthful source, fake news, or disinformation servers, they would have either avoided such sources or read them with high caution. Kahne and Bowyer (2017), in their analysis, found that although youth who reported having media literacy learning opportunities were more likely to be influenced by accuracy motivation, they were not more likely to be influenced by directional motivation.

In this article I am going to test if media literacy training is related to directional motivation. Additionally, I am investigating the effect of trust and ideology that recent research has shown to be important (Saunders, 2017; Miller, Saunders & Farhart, 2016).

General Trust

We witness declining trust in politics and the media, which appears to be stronger in politically polarized societies (Hanitzsch et al., 2017). It in effect reduces the role of media as reliable sources and intensifies the role of other factors that shape and frame information reflecting several types of bias (Gainous & Wagner, 2013). People select sources of information depending on many aspects including their availability and accessibility (Bronstein, 2010). They are also inclined to choose a particular source of information that corresponds to their values and allows them to validate their opinions. This then leads them to follow like-minded sources that, in turn, produces something that Goldman and Mutz (2011) call the “friendly media phenomenon.” The opposite is “oppositional news hostility,” which stands for opposing views on news and exacerbates polarization (Sunstein, 2009). Social media represent an ideal platform for both phenomena. This, in combination with an effort for producing scandalous news by simplistic coverage that will attract attention, leads to a situation that current affairs may not be able to sufficiently explain and leave the audience undecided what to believe. The obvious answer would be directional motivation, that is, believing information or sources that are most aligned with a person's own opinion.

Kunda (1990) suggests that although people are influenced by directional motivation, they try to be rational and pursue justification for their desired conclusions. In other words, they try to make an “illusion of objectivity” in cases where they are motivated. People are motivated to do so when dealing with politics because as Lodge and Taber (2013) argue, citizens are rarely dispassionate when thinking about politics. Respondents who indicate that the statement that protects their attitudes is accurate will be less likely to make the illusion of objectivity if they also believe that people, politicians, and media are trustworthy (Saunders, 2017). The argument here is similar to that of Miller, Saunders, and Farhart (2016), who have shown that high trust mitigates directional motivation because individuals who believe that people, media, and institutions are trustworthy will not be motivated to maintain the “illusion of objectivity.”

Ideology

Confirmation bias makes individuals seek and perceive consistent information and leads them to strengthen their attitude. Ideology is a set of predispositions that consistently predict which disinformation individuals will believe. Ideology is not only a shortcut for making judgments about politics, but it is also a motivational stimulus for justifying and rationalizing an individual's judgment about certain information (Jost, Federico & Napier, 2013).

In accordance with Miller, Saunders, and Farhart (2016), it is expected that conservatives and liberals will differ in their strength of motivation of directional reasoning. They operationalized ideology as a party identification of either Democrats or Republicans. Their argument was grounded in meaning that directional motivation should be stronger among those whose party is in opposition. A similar argument was made earlier by Uscinski & Parent (2014) claiming that conspiracy theories are being used as coping mechanisms by groups to manage their anxieties about perceived threats and loss of power. Investigating letters to the New York Times editor, they found that the prevailing ideology of conspiracy theories depended on the political affiliation of a president. There were always more conspiracy theories that were aligned to ideology of a current president. Results of their research suggest that conspiracy theories were always more popular among groups that had a reason to feel disempowered and endangered.

In this paper, ideology is not understood as party identification but, rather, as a set of values that is operationalized as an average position on issues related to refugees. Those who are against equal rights and the same treatment for refugees as for Czechs are considered conservative. Due to the fact that there are almost no refugees in the Czech Republic and Czech authorities grant equal rights to minorities, the conservatives are considered to be ideological losers who feel endangered by potential refugee arrival.

Additionally, according to Adorno et al. (1950), right-wing ideology is a manifestation of traits such as dogmatism, aversion to complexity, and closure in argumentation. The cognitive style using such dispositions produces closed-mindedness toward evidence that is against the factual premises of arguments, which are against conservative values (Kahan, 2013; Jost et al., 2003; Jost, Nosek & Gosling, 2008; Feygina, Jost & Goldsmith, 2010). It is expected that directional motivation will work differently, depending on the respondent's ideology, and that the types of arguments will have different effects on conservatives and liberals when testing accuracy reasoning.

Hypotheses

This section discusses the hypotheses dealing with the effects of political knowledge and media literacy training on directional motivation. In accordance with Taber and Lodge (2006) and Kahne and Bowyer (2017), the expectation is that the effects of directional motivation will be greatest for individuals with ideologically aligned posts, the highest political knowledge, and those reporting the most media literacy training.

H1: Ideologically aligned posts will be more likely to be judged as accurate than unaligned ones.

H2: The difference in the accuracy rating of participants who are assigned to ideologically aligned posts and those assigned to unaligned posts will be greater for participants with high levels of political knowledge.

H3: The difference in the accuracy ratings of participants who are assigned to ideologically aligned posts and those assigned to unaligned posts will be greater for participants with reported frequent media literacy trainings.

Although Kahne and Bowyer (2017, p. 13) note that their approach of aligned posts is similar to that of other studies testing partisan or ideological orientation of motivated reasoning (Taber & Lodge, 2006), they only separate ideologically aligned from unaligned posts but do not assume that motivated reasoning of liberals and conservatives could work differently.

However, there are studies (Miller, Saunders & Farhart, 2016; Saunders, 2017; Kahan, 2013) suggesting that conservatives and liberals are affected by motivated reasoning differently. In this study the conservatives are considered to be the group that may feel endangered and disempowered. As discussed earlier in the paper, the desire to engage in motivated reasoning should be stronger for conservatives because for them the policy implications are much more unpleasant.

H4: Conservatives will be more affected by directionally motivated reasoning than liberals.

H5: General trust will moderate disinformation accuracy judgment for conservatives but will decrease or be unchanged for liberals.

METHODS

The analysis is based on Czech Youth Survey (2017) data. It is a nationally representative survey of youth (16–20 years old). The survey was conducted by paper-pencil method in schools during regular school days. Schools were selected by stratified random sampling method. Strata were geographic location, size of the city, and type of school (gymnasium, vocational school with graduation, vocational school without graduation). The fieldwork was administered by a private survey company. The dataset contains answers from 1,207 respondents from 72 schools. Ninety-five percent of the sample does use their Facebook account, and therefore the topic of this study is highly relevant to them. Due to hierarchical structure of the data (pupils are nested in schools), all models presented in this article are multilevel ordinal logit with random intercept. It is controlled for age, gender, family income, size of hometown, and parental education in all presented models.

Experimental Design

The Czech Youth Survey questionnaire includes questions regarding political participation and political attitudes of the youth. Apart from that, there was also an experimental question. Respondents were exposed to the experimental question when they had finished with about one-third of the questionnaire. The design was developed and tested by Kahne and Bowyer (2017). The aim was to prepare visual illustrations with catchy phrases that are likely to circulate online through social media that represent the main source of fake news and disinformation (Cohen et al., 2012).

The actual questions were inspired by Kahne and Bowyer (2017). In their survey, they had six posts about taxation to which youth were assigned. In our study, there were six groups illustrating pros and cons of accepting refugees to the Czech Republic (CR). Respondents were randomly assigned to one of these six groups. The treatment was always either a political cartoon or a graph with a short declarative expression on the topic of (not) accepting refugees. Figure 1 displays each of the posts. For purposes of this article, descriptions were changed from Czech to English in this figure. These posts were manipulated to vary in their political ideology (liberal-conservative) and type of argument, which was either emotive, disinformation, or evidence-based. Varying these two aspects was done for testing directional motivation.

FIGURE 1.

“Posts” seen by experimental groups by topic, ideology, and type of argument.

FIGURE 1.

“Posts” seen by experimental groups by topic, ideology, and type of argument.

Respondents assigned to Groups 1, 2, and 3 had a statement that presented a liberal position about accepting refugees to the CR (“CR should accept refugees!”); respondents assigned to Groups 4, 5, and 6 were exposed to a conservative argument about accepting refugees (“CR should not accept any refugees!”). Posts varied not only in their ideological position but also in the type of argument. Groups 1 and 4 had an emotive appeal. Refugee groups had either a liberal argument: “Refugees coming to the Czech Republic do not endanger our security and do not increase crime rate” or a conservative argument: “The number of refugees in the EU is rising sharply.” No emotive cartoon showed any evidence on behalf of the policy position. Groups 2 and 5 used the same two sentences as did Groups 1 and 4, respectively, but the cartoon was replaced with graphs that presented real data accompanied by indication of trustworthy source.

The last pairs of experimental conditions were designed to test whether disinformation affects young people's judgments of the accuracy of truth claims. Groups 3 and 6 were exposed to the same cartoons as Groups 1 and 4. The difference was that the cartoons of former groups were presented with a disinformative first sentence. The disinformation was so huge that it was evident that it could not be true. Group 3 was exposed to “90% of refugees coming to the Czech Republic are mothers and children”; Group 6 read that “the share of refugees in the population of the Czech Republic was over 30% in 2015.”

The purpose of these conditions was to test if youths' judgment is affected by inclusion of either an emotive, evidence-based, or substantially disinformative statement. In total these conditions also test the liberal and conservative versions of these three types of arguments.

DEPENDENT AND INDEPENDENT VARIABLES

Judgment of Accuracy Respondents as Dependent Variable

The dependent variable for all models presented in this article is the respondent's judgment accuracy of the claim. Respondents were asked to say how strongly they (dis)agree with the claim that “I think this comment is accurate” presented under the picture. Their answer options were “strongly disagree; disagree; agree; strongly agree.”

Alignment with Political Beliefs

To find out whether a post was aligned with a respondent's beliefs or not, we used some ready answers to a set of questions regarding attitudes toward refugees. A set of questions is better than a single question because of better reliability and accuracy of measurement. The set of questions was as follows: (1) “Refugees should have the opportunity to continue their own customs and lifestyle.” (2) “Refugees should have all the same rights that everyone else in the country has.” (3) “Refugees who live in a country for several years should have the opportunity to vote in elections.” (4) “Immigrant children should have the same opportunities for education that other children in the country have.”

Respondents were supposed to answer on a scale: “definitely agree; rather agree; rather disagree; definitely disagree.” Answers to these questions were transformed into an index. Respondents whose index was lower than 2.5 (prevailed answers were positive about refugees) were coded as ideologically aligned if they were assigned to Group 1, 2, or 3. Ideologically aligned were also considered respondents whose index was higher than 2.5 (answers were on average negative about refugees) and were assigned to Group 4, 5, or 6. The rest of the respondents were coded as ideologically unaligned. Ambivalent respondents (those who were in the middle) were excluded from the analysis.

Political Knowledge

Political knowledge is measured as an additive index of correct answers to five questions regarding contemporary Czech politics, the Czech political system, foreign politics, and the unemployment rate, and are supposed to be a complex measure of how respondents are politically informed. The questions always had four possible answers and a “don't know” option. Questions measured knowledge of how the parliament works, what is the actual unemployment rate, last election results (regional elections), which parties compose current government, and the number of membership states in the EU.

Media Literacy Training

Kahne and Bowyer (2017, p. 15), in their article, use media literacy training as a proxy. They argue that “the central rationale for media literacy instruction is that it should increase individuals' attentiveness to the accuracy of information as well as the skills to assess accuracy.” The same measurement of media literacy training was used also in this study. Media literacy training is measured as the sum of responses to questions of how often respondents have “Discussed how to tell if the information you find online is trustworthy” and “Discussed the importance of evaluating the evidence that backs up people's opinions” in school classes.

Political Ideology

Ideology is operationalized as the average of answers to a set of questions used to operationalize alignment with political beliefs regarding refugees. Those who are in favor of refugees are considered liberal. Respondents who oppose refugees' rights and refuse equal treatment of refugees and Czech citizens are considered conservative.

General Trust

In accordance with Miller, Saunders, and Farhart (2016), the trust index is supposed to capture general trust of a respondent. In this article the motivation was to include items that are as similar as possible to those used by Miller, Saunders, and Farhart (2016). General trust is operationalized as an index derived from responses indicating the extent of trust of (1) people in general, (2) local government, (3) regional government, (4) parliament, (5) national government, 6) news on the Internet, (7) news on TV, (8) news in printed newspapers and journals (alpha = .77). Each item had four response options (not trust at all, rather not trust, rather trust, definitely trust) and a “do not know” option. The “do not know” option was put in the middle as it does not represent either trust or distrust. The higher the index, the higher the general trust.

RESULTS

Judgments of Truth Claims Using Directional Motivation

In order to test the first hypothesis, answers from respondents exposed to ideologically aligned and unaligned posts are compared. Posts in each pair of experimental conditions (Groups 1 and 4, 2 and 5, 3 and 6) are equivalently accurate; therefore, differences in ratings between ideologically aligned and unaligned are the consequence of directional motivation.

Figure 2 graphically presents the share of people agreeing and disagreeing with the accuracy of statements divided by type of argument. Accuracy of the statement was agreed on by an average of 40% of those who were exposed to an unaligned argument and 51% of respondents who saw an aligned statement. Although the differences are not as huge as in the US study, there are statistically significant differences (p < .001 for a two-tail t-test) between those assigned to ideologically aligned and unaligned posts. This is true across all three types of arguments. As expected, directional motivation has an effect on judgments of accuracy and, similarly to Kahne and Bowyer (2017), H1 is confirmed. Contrary to expectations, disinformative posts did not receive a lower share of accuracy judgments than emotive or evidence-based posts. We actually witness completely opposite results to our expectations. The share of positive responses to judgment of accuracy was 55% (disinformation), 52% (evidence-based), and 49% (emotive). However, the differences are so small that the t-test does not prove statistically significant differences.

FIGURE 2.

Ratings of posts' accuracy by topic, ideological alignment, and type of argument.

FIGURE 2.

Ratings of posts' accuracy by topic, ideological alignment, and type of argument.

Effects of Political Knowledge, Media Literacy Training, and Trust on Judgments of Truth Claims

This article tests effects of political knowledge, media literacy training, and general trust on motivated reasoning, which is driven by an individual's motivation to reach preferred conclusion (directional goal). Hypotheses H1 to H3 deal with a question about which factors affect directional motivation. Results of regression models are presented in table 1.

TABLE 1.

Results of multilevel ordinal logistic regression models with random intercept

Model 1Model 2Model 3Model 4Model 5
Female .20 .20 .33* .33* .32* 
Parental education −.03 −.04 .03 .03 .03 
Household income −.08 −.07 −.10 −.12 −.12 
Evidence−based posts .01 .01 .16 .16 .17 
Disinformation −.07 .07 .02 .36 1.63 
Unaligned posts −.69* −2.70*** −.53 −.54 −.24 
Aligned posts − conservative   3.86*** 3.93*** 5.32*** 
Political knowledge −.04 .08 .12 .12 .11 
Media literacy −.09 −.10 −.06 −.11* −.11* 
General trust .08 −.35 .23 .32 .49 
Unaligned posts* Political knowledge −.02 −.02 −.17   
Aligned posts – conservative* Political knowledge   −.34*   
Unaligned posts* Media literacy .03 .01 −.01   
Aligned posts − conservative* Media literacy   .05   
Unaligned posts*General trust  .92*** .34 .27 −.15 
Disinformation*Political knowledge  −.28**  −.33** −.33** 
Disinformation*Media literacy  .08  .13 .14 
Disinformation*General trust  −.03  −.13 −.67 
Unaligned posts*Disinformation  −.02  −.17 −1.14 
Aligned posts − conservative*Disinformation    −.45 −4.69* 
Aligned posts − conservative*General trust   −.67 −.69 −1.30** 
Unaligned posts*General trust*Disinformation     .40 
Aligned posts − conservative*General trust*Disinformation     1.87* 
Observations 921 921 921 921 921 
Number of groups 72 72 72 72 72 
ll −1253 −1243 −1185 −1181 −1178 
Model 1Model 2Model 3Model 4Model 5
Female .20 .20 .33* .33* .32* 
Parental education −.03 −.04 .03 .03 .03 
Household income −.08 −.07 −.10 −.12 −.12 
Evidence−based posts .01 .01 .16 .16 .17 
Disinformation −.07 .07 .02 .36 1.63 
Unaligned posts −.69* −2.70*** −.53 −.54 −.24 
Aligned posts − conservative   3.86*** 3.93*** 5.32*** 
Political knowledge −.04 .08 .12 .12 .11 
Media literacy −.09 −.10 −.06 −.11* −.11* 
General trust .08 −.35 .23 .32 .49 
Unaligned posts* Political knowledge −.02 −.02 −.17   
Aligned posts – conservative* Political knowledge   −.34*   
Unaligned posts* Media literacy .03 .01 −.01   
Aligned posts − conservative* Media literacy   .05   
Unaligned posts*General trust  .92*** .34 .27 −.15 
Disinformation*Political knowledge  −.28**  −.33** −.33** 
Disinformation*Media literacy  .08  .13 .14 
Disinformation*General trust  −.03  −.13 −.67 
Unaligned posts*Disinformation  −.02  −.17 −1.14 
Aligned posts − conservative*Disinformation    −.45 −4.69* 
Aligned posts − conservative*General trust   −.67 −.69 −1.30** 
Unaligned posts*General trust*Disinformation     .40 
Aligned posts − conservative*General trust*Disinformation     1.87* 
Observations 921 921 921 921 921 
Number of groups 72 72 72 72 72 
ll −1253 −1243 −1185 −1181 −1178 

Model 1 includes interactions of disinformation, ideological alignment and political knowledge, media literacy training, and general trust. The prediction that directional motivation will have greater influence on individuals with high political knowledge (H2) and the most media literacy training (H3) if they were exposed to aligned argument is not supported by the Czech data.

The interaction of ideologically unaligned posts and trust is statistically significant and positive, indicating that when individuals are exposed to ideologically aligned posts, the effects of directional motivation are greatest among those with the least trust (figure 3). Similarly, the interaction of disinformative posts and political knowledge is statistically significant and negative (figure 4). This result suggests that youth with high political knowledge are less likely to be affected by directional motivation when they are exposed to disinformative posts. Results for media literacy training are not that convincing. The interaction of media literacy training and disinformation was not significant in any models tested.

FIGURE 3.

Interaction of general trust and posts alignment (based on Model 2).

FIGURE 3.

Interaction of general trust and posts alignment (based on Model 2).

FIGURE 4.

Interaction of political knowledge and disinformation (based on Model 2).

FIGURE 4.

Interaction of political knowledge and disinformation (based on Model 2).

The next step proceeded with Model 3, which does not use a variable of simple ideological (un)alignment as previous models do but also takes into account the direction of ideological alignment (unaligned vs. liberal vs. conservative). Results of Model 3 indicate a negative interaction effect of political knowledge on conservatives but not on liberals or those who are ideologically unaligned.

Next, there was a test of the expectation that conservatives will have greater motivation to be influenced by a disinformative argument than liberals (H5). This would be consistent with the belief that disinformation is for ideologically motivated losers. The negative interaction effect of disinformation on ideology suggests that conservatives are less likely to agree with the posts if they are exposed to disinformation.

The results of Model 5 are supportive of H5 that directional motivation has greater influence on conservative participants with the least general trust. The difference between Model 4 and Model 5 is given by added three-way interaction in the latter. Statistically significant results of three-way interaction imply that the combination of ideology, general trust, and disinformation is the key for understanding directional motivation among Czech youth. The three-way interaction of disinformation, ideology, and trust tests if disinformative posts and trust will interact with ideology to moderate accuracy of reasoning. The test has mixed results (figures 5 and 6).

FIGURE 5.

Interaction of general trust and ideology for nondisinformative posts (based on Model 5).

FIGURE 5.

Interaction of general trust and ideology for nondisinformative posts (based on Model 5).

FIGURE 6.

Interaction of general trust and ideology for disinformative posts (based on Model 5).

FIGURE 6.

Interaction of general trust and ideology for disinformative posts (based on Model 5).

The results graphically presented in figures 5 and 6 suggest that conservatives with more trust are less likely than others to judge posts that align with their prior beliefs as accurate, regardless of the posts' real accuracy. The nondisinformative posts negatively affect conservatives' willingness to agree with the statement if they score high on general trust. Contrary, the disinformative posts have a positive effect on conservatives with high trust. Ideologically, neither unaligned nor liberals were affected by the type of posts regardless of their level of general trust.

DISCUSSION

This study of Czech Youth contributes to the contemporary research on motivated reasoning in three ways. First, models with the same design as in Kahne and Bowyer's (2017) study were analyzed to test the motivated reasoning theory in a different environment. The analysis has shown mixed results. Although evidence that youth are guided by directional motivation was found, descriptive statistics and t-tests did not prove that disinformation has a considerable effect on youth's judgment of accuracy, which is contrary to Kahne and Bowyer's (2017) finding. In opposition to Kahne and Bowyer (2017), the results of this study have not supported the argument of role of political knowledge in directional motivation. Contrary to their results, Czech youth with the most media literacy training were not more likely to use directional motivation when making judgments about posts.

Second, Kahne and Bowyer's (2017) model of directional motivation was extended by incorporating the types of ideology. Third, the role of general trust on directional motivation was examined. The results of this analysis support findings of previous studies (Saunders, 2017; Miller, Saunders & Farhart, 2016) that general trust is an important moderator of motivated reasoning. Respondents with high trust who were exposed to an argument aligning with their prior attitudes were less likely to report the statement as accurate.

The results of this paper suggest that future studies on motivated reasoning should consider taking into account not only respondents' alignment with the posts but also with ideological orientation of respondents, and that applies to youth too. Based on models in this article, directional motivation does not seem to work the same way for conservatives and liberals. Conservatives were found to be less likely to agree with the posts if they are exposed to disinformation. There is no support for such effect on liberals. As trust was found to be an important moderator for directionally motivated reasoning, it has to be emphasized that trust is an important moderator primarily for conservatives.

Additionally, high trust has a significant negative impact on conservatives, in comparison to liberals, which suggests that conservatives with low trust are more likely to be affected by directional motivation than liberals. However, the role of trust depends also on the type of argument. High trust among conservatives causes mitigation of positive effect on directional motivation if they are not exposed to a disinformative argument about refugees. On the contrary, disinformative posts made highly trusting conservatives more likely to agree with a statement. Although the ideology does not have a clear effect, the important finding is that directional motivation works differently for liberal and conservative youth.

The connection between the extent of media literacy training and type of posts was not statistically significant. This may be explained by the fact that the media literacy trainings at schools may not be sufficient for recognizing disinformation or that the operationalization was not ideal. Furthermore, it is important to acknowledge that participation in media literacy training does not make people necessarily media literate. Having better media literacy measurements would be more helpful in this sense.

The theory of motivated reasoning should be replicable and work similarly in any environment. For this reason, the same design as in Kahne and Bowyer's (2017) study was used in this research. Comparing their results with analysis of this article, we found some substantial differences. The question is, what does make the results different? One of the possible answers may be the extent of polarization and saliency of issues in the USA and the Czech Republic. The USA has a strong polarized bipartism, and the CR has ideologically fragmented multipartism with parties that mostly are not clearly ideologically defined. A second possible explanation can be in the different age of respondents. Kahne and Bowyer (2017) had participants up to the age of 30, but the Czech Youth Survey involved high school students up to the age of 20. Adolescence is a crucial period for the development of political attitudes and behavior patterns. Older adults have usually finished their education, gotten a job, moved out of their parents' home, and become independent. As a consequence, older adults have different interests than high school students. Adults are also more interested in politics, and stable in their values and attitudes. The age difference between respondents may partly explain the discrepancies in results between younger (Czech) respondents and older respondents from the USA.

CONCLUSION

Increasing political polarization and dramatic changes in the media world contribute to the rise and circulation of fake news and disinformation in the media and on the Internet in general. People who are not able to recognize disinformation and who spread disinformation can undermine the democratic process; citizens should be able to make informed decisions based on true information. One of the conclusions reached in this study is that trying to increase political knowledge or provide media literacy training does not seem to be the most effective way to fight disinformation. Political knowledge and media literacy training may improve accuracy motivation but do not affect directional motivation. Results of this analysis suggest that general trust is a more important moderator of directional motivation.

Although highly trustful people are in general more likely to agree with statements, the role of high trust is not conclusively either “good” or “bad.” High trust makes people less likely to be driven by directional motivation. This could be surprising because common sense may suggest that more-trusting individuals will be more likely to endorse statements that are aligned with their beliefs. However, we know from previous studies that trust is negatively correlated with belief in conspiracy theories (Swami, Chamorro-Premuzic & Furnham, 2010), and believing in conspiracy theories is an extreme form of motivated reasoning. As suggested by Kunda (1990, p. 483) or Miller, Saunders, and Farhart (2016, p. 829), liberals and conservatives probably do not “muster up the evidence necessary” to be driven by motivated reasoning if they believe that people, media, and political institutions are trustworthy.

Results of this article do not imply any characteristic or combination of characteristics that will universally help fight disinformation. Taking into consideration the results of this article, further research regarding disinformation and fake news may focus more on the role of civic education and consider critical thinking as a useful tool against disinformative posts. Critical thinking is different from political knowledge because it is not about being informed but, rather, about having the ability to make reasoned judgements and better understand if certain groups have a reasonable evidence to feel endangered. Additionally, media literacy training and media literacy itself are two distinct things. Variables capturing media literacy were not available in the dataset that was used, and thus this may be a good direction for future research as well.

REFERENCES

REFERENCES
Adorno
,
T. W.
,
Frenkel-Brunswik
,
E.
,
Levinson
,
D. J.
&
Sanford
,
R. N.
(
1950
)
The Authoritarian Personality
.
New York
:
Harper and Row
.
Barber
,
M.
&
McCarty
,
N.
(
2013
) Causes and consequences of polarization. In:
Mansbridge
,
J.
&
Martin
,
C. J.
(eds.)
Negotiating agreement in politics: Report of the Task Force on Negotiating Agreement in Politics
.
Washington, DC
,
American Political Science Association
, pp.
19
53
.
Bronstein
,
J.
(
2010
)
Selecting and using information sources: Source preferences and information pathways of Israeli library and information science students
.
Information Research
.
15
(
4
). Available from: http://informationr.net/ir/15-4/paper447.html.
Cohen
,
C. J.
,
Kahne
,
J.
,
Bowyer
,
B.
,
Middaugh
,
E.
&
Rogowski
,
J.
(
2012
)
Participatory Politics: New Media and Youth Political Action
. Available from: https://ypp.dmlcentral.net/sites/default/files/publications/Participatory_Politics_Report.pdf.
Čulík
,
J.
(
2017
).
Why is the Czech Republic so hostile to Muslims and refugees?
Europe Now
,
9
February
. Available from: https://www.europenowjournal.org/2017/02/09/why-is-the-czech-republic-so-hostile-to-muslims-and-refugees/.
Čulík
,
J.
(
2015
)
Anti-immigrant walls and racist tweets: the refugee crisis in Central Europe
.
The Conversation
,
24
June
. Available from: https://theconversation.com/anti-immigrant-walls-and-racist-tweets-the-refugee-crisis-in-central-europe-43665.
Delli Carpini
,
M. X.
&
Keeter
,
S.
(
1993
)
Measuring political knowledge: Putting first things first
.
American Journal of Political Science
.
37
(
4
),
1179
1206
. Available from: doi:.
Ditto
,
P. H.
,
Scepansky
,
J. A.
,
Munro
,
G. D.
,
Apanovitch
,
A. M.
&
Lockhart
,
L. K.
(
1998
)
Motivated sensitivity to preference-inconsistent information
.
Journal of Personality and Social Psychology
.
75
(
1
),
53
. Available from: doi:.
Feygina
,
I.
,
Jost
,
T. J.
&
Goldsmith
,
E. R.
(
2010
)
System justification, the denial of global warming, and the possibility of “system-sanctioned change.”
Personality and Social Psychology Bulletin
.
36
(
3
),
326
338
. Available from: doi:.
Gainous
,
J.
&
Wagner
,
K.
(
2013
)
Tweeting to Power: The Social Media Revolution in American Politics
.
New York
,
Oxford University Press
.
Goldman
,
S.
&
Mutz
,
D.
(
2011
)
The friendly media phenomenon: a cross-national analysis of cross-cutting exposure
.
Political Communication
.
28
(
1
),
42
66
. Available from: doi:.
Hanitzsch
,
T.
,
van Dalen
,
A.
&
Steindl
,
N.
(
2017
)
Caught in the nexus: a comparative and longitudinal analysis of public trust in the press
.
International Journal of Press/Politics
.
23
(
2
),
3
23
. Available from: doi:.
Hochschild
,
J.
&
Einstein
,
K. L.
(
2015
)
Do Facts Matter? Information and Disinformation in American Politics
.
Norman
,
University of Oklahoma Press
.
Iyengar
,
S.
, &
Hahn
,
K. S.
(
2009
)
Red media, blue media: Evidence of ideological selectivity in media use
.
Journal of Communication
.
59
(
1
),
19
39
. Available from: doi:.
Jost
,
J.
,
Federico
,
C. M.
&
Napier
,
J. L.
(
2013
) Political ideologies and their social psychological functions. In:
Freeden
,
M.
(ed.)
Oxford handbook of political ideologies
.
New York
,
Oxford University Press
, pp.
232
250
.
Jost
,
J. T.
,
Glaser
,
J.
,
Kruglanski
,
A. W.
&
Sulloway
,
F. J.
(
2003
)
Political conservatism as motivated social cognition
.
Psychological bulletin
.
129
(
3
),
339
375
. Available from: doi:.
Jost
,
J. T.
,
Nosek
,
B. A.
&
Gosling
,
S. D.
(
2008
)
Ideology: Its resurgence in social, personality, and political psychology
.
Perspectives on Psychological Science
.
3
,
126
136
. Available from: doi.org/10.1111/j.1745-6916.2008.00070.x.
Kahan
,
D.
(
2013
)
Ideology, motivated reasoning, and cognitive reflection
.
Judgment and Decision Making
.
8
(
4
),
407
424
.
Kahne
,
J.
&
Bowyer
,
B.
(
2017
)
Educating for democracy in a partisan age: Confronting the challenges of motivated reasoning and disinformation
.
American Educational Research Journal
.
54
(
1
),
3
34
. Available from: doi:.
Kunda
,
Z.
(
1990
)
The case for motivated reasoning
.
Psychological Bulletin
.
108
(
3
),
480
. Available from: doi:
Lodge
,
M.
, &
Taber
C. S.
(
2013
)
The Rationalizing Voter
.
New York
,
Cambridge University Press
.
Lodge
,
M.
&
Taber
,
C. S.
(
2005
)
The automaticity of affect for political leaders, groups, and issues: an experimental test of the hot cognition hypothesis
.
Political Psychology
.
26
(
3
),
455
482
. Available from: doi:.
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
. Available from: doi:.
Prior
,
M.
(
2013
)
Media and political polarization
.
Annual Review of Political Science
.
16
,
101
127
. Available from: doi:.
Rojecki
,
A.
&
Meraz
,
S.
(
2016
)
Rumors and factitious informational blends: the role of the web in speculative politics
.
New Media & Society
.
18
(
1
),
25
43
. Available from: doi:.
Saunders
,
L. K.
(
2017
)
The impact of elite frames and motivated reasoning on beliefs in a global warming conspiracy: the promise and limits of trust
.
Research and Politics
.
4
(
3
),
1
9
. Available from: doi:.
Stroud
,
N. J.
(
2011
)
Niche News: The Politics of News Choice
.
New York
,
Oxford University Press
.
Schwartz
,
S. H.
(
1994
)
Are there universal aspects in the structure and contents of human values?
Journal of Social Issues
.
50
,
19
45
. Available from: doi:.
Sunstein
C. R.
(
2009
)
Going to Extremes: How like Minds Unite and Divide
.
Oxford
,
Oxford University Press
.
Swami
,
V.
,
Chamorro-Premuzic
,
T.
&
Furnham
,
A.
(
2010
)
Unanswered questions: apreliminary investigation of personality and individual difference predictors of 9/11 conspiracist beliefs
.
Applied Cognitive Psychology
.
24
(
6
),
749
761
. Available from: doi:.
Taber
,
C. S.
&
Lodge
,
M.
(
2016
)
The illusion of choice in democratic politics: the unconscious impact of motivated political reasoning
.
Political Psychology
.
37
,
61
85
. Available from: doi:.
Taber
,
C. S.
&
Lodge
,
M.
(
2006
)
Motivated skepticism in the evaluation of political beliefs
.
American Journal of Political Science
.
50
(
3
),
755
69
. Available from: doi:.
Uscinski
,
J. E.
&
Parent
,
J. M.
(
2014
)
American Conspiracy Theories
.
Oxford
,
Oxford University Press
.