How University Students Evaluate the Role of Social Media in Political Polarization: Perspectives of a Sample of Turkish Undergraduate and Graduate Students

Abstract

This study aimed to find out if there is a relationship between social media and political polarization in Türkiye from the perspective of Turkish students. To reach this aim, the needed data were collected through qualitative and quantitative approaches. A total of 303 valid questionnaires were analyzed. The sample consisted of university Turkish students across undergraduate, masters, and PhD levels in Türkiye aged between 18 and 50+. As well, an online focus group discussion with six Turkish students from different universities and education levels was conducted to gain a more in-depth understanding of the study’s problem. The results of the study showed that the perspectives of the Turkish students were that social media had a weak-to-moderate effect on political polarization in Türkiye. Furthermore, the results indicated that the studied sample of the Turkish students does not rely on social media platforms to obtain political news, and most of them do not follow political leaders on social media. Moreover, communication platforms did not encourage many Turkish students to express themselves, which is an indication that social media algorithms have contributed to a medium degree in creating filter bubbles through the content they suggest to users. Results have also shown that Turkish students are afraid that their posts and comments are being censored.

1. Introduction

Since social media has the role of allowing users to express their opinions, unlike traditional media, where the person is only a receiver, many studies have shown how the role of the audience has changed from “has been affected” to has affected (Aldamen 2023d) and is shifted from receiving messages to sharing content, as well as how it is used to affect the receiver’s point of view and opinions regarding many issues in societies (Aldamen 2023a).

Some users might abstain from doing so because they are afraid of facing backlash. Therefore, it leads them to fall back into a one-dimensional hemophilic environment again, creating a spiral of silence (Noelle-Neumann 1974). Posts that are shown on the users’ feed are personalized based on each user’s preferences and isolate them from content that they have shown less interest in. This means that social media algorithms can hide posts from people, news websites, or accounts that the user does not agree with or holds different viewpoints. This leads to people believing that they have full information about something when they are only getting one side of it, making it harder for people to agree on facts and limiting them to a single viewpoint. Partisan disinformation can easily be spread on social media, especially if it is coming from a partisan leader or influencer. Thus, dependency on social media to learn about events and obtain information about other parties’ ideas can increase the intensity of incompatibility between different party supporters.

According to the Turkish Statistical Institute (2021), around 56 percent of the population had access to the Internet in 2015. The fact that a significant portion of the population lacks internet access makes estimating the consequences of internet use on various outcomes, such as political awareness, easier. Even though internet content consumption is not as high as traditional media, especially TV, social media is exceptionally in demand among Turkish internet users. A total of 87 percent of adult internet consumers (those who have a smartphone) declared using social media, making Türkiye one of the top five fastest-growing countries in social media usage. Social media has earned its place as a handy tool for following local politics.

Critics commonly blame social media platforms for intensifying political polarization by creating “echo chambers” that shield people from information that contradicts their preexisting ideas. Rather than studying the consequences of social media, this line of research aims to diagnose the existing status of the political landscape on social media.

Polarization cannot be reduced to an equilibrium of responses between agreement and disagreement with survey questions (except in the limiting case of two-point scales). Polarization is inherent in the extremes and distances between reactions, not in their substantive content (DiMaggio et al. 1996). Bearing in mind the significance of polarization in modern political discourse, the literature offers startlingly little aid in defining it. Possibly the easiest place to start is by defining what polarization is not. Polarization is not disruptive incivility in political debate, even though the two may or may not be empirically connected. Consequently, polarization relates to the magnitude of disagreement, not how disagreement is communicated. When inequality within groups diminishes or increases between groups, polarization rises. In this situation, the members of each cluster are of comparable quality, yet it could be claimed that society is polarized if distinct clusters contain people with dissimilar traits (Keyifli and Akdede 2020).

Contemporary polarizations sometimes begin when a previously divided or disadvantaged segment of society becomes politically unified and determined to attain social, economic, cultural-ideological, or institutional goals. This happens when political entrepreneurs effectively highlight and trigger underlying social cleavages, bringing to the forefront, creating, or reinventing a dominant cleft around which other cleavages align. As a result, polarization is characterized as the process through which a society’s typical variety of disparities progressively coincide along a single axis, cross-cutting contrasts become reinforcing, and individuals understand and explain politics and society in terms of “us” vs. “them” (McCoy et al. 2018). The separation between elite and mass polarization is a feature of the current period. The elite polarization is a polarization within the government party, and within the party as an organization. The distribution of political attitudes among all citizens is one approach to looking at public opinion polarization, also known as mass polarization (Baldassarri and Gelman 2008).

Higher degrees of polarization can be good for society, as they predict higher levels of political engagement and voter attitudes. Political polarization, on the other hand, can be harmful to democracy, increasing power concentration and legislative deadlock (Pérez-Liñán et al. 2019), as well as decreasing citizen satisfaction (Kubin and von Sikorski 2021).

2. Türkiye’s Polarization Patterns

Within the historical process, political polarization happens on a societal level. A valuable political tool to comprehensively understand political and social transformation has always been the center–periphery approach. Although this approach has occasionally failed to explain social and political development and transition, it has nonetheless been an essential tool for studying Turkish politics. Although the center–periphery relationship has remained important in the analysis of the political structure from the classical period Ottoman to the Turkish Republic, it has experienced periodic changes (Tuncel and Gündoğmuş 2012).

Polarization in Türkiye is understood through the center–periphery theory (Shils 1975). Shils’s model is based on the idea that every society has a center and a periphery. Center–Periphery Theory refers to the existence of a conflict that takes place around the positions that the center and the periphery are trying to gain or protect. After evaluating the Ottoman period and the first years of the Republican period, it can be seen that, even though the center and the periphery changed in name or shape, these groups retained their place in terms of the status, location, or culture they adopted. In other words, even if a peripheral element transitions to the center in terms of management or economy, it has always been culturally left out of the center or left; similarly, the center has always made an effort to stay in the center in a sense, even when it was in the periphery with the claim of cultural superiority (Bilgiç 2014). According to this model, the “center” is conceptualized as a social and political structure and is positioned as the determinant of the values, symbols, and beliefs on which the rulers of society are based. In other words, the center is not defined as a structure formed by places and institutions but as the center of spiritual characteristics formed by values and beliefs. Historically, political and social polarization in Türkiye developed over five eras or stages: the Ottoman Empire Era, the Republican Era, the Multi-Party Period, the 1961 Constitution, and the 2000s Period. In each period, the relationship between the center and the periphery took on a different shape; also, with different political and economic events taking place in each period, each event introduced new value systems and brought changes to the socio-political atmosphere (Kurt 2015).

The overlap between political parties and political identities is one of the most important causes of the high level of division among political party followers. Political identities, or one’s “place on earth,” are shaped through shared victimhood and glorification, as well as shared tastes and traits, and are defined in connection to who the “other” is. When political party identification and political identities collide, political polarization is more likely. Political party preferences are no longer transient and changing, and these preferences, which are shaped by political identities, may help to separate oneself from other party followers (Erdoğan 2018).

3. Political Polarization and Social Media

The media bears a significant amount of responsibility for highlighting the most pressing concerns in society. People consider online media far more effective in this respect, and they prefer it over traditional media when it comes to following up on citizen rights concerns, as traditional media must follow a winning strategy and do more in this regard (Aldamen 2017). Social media, particularly Twitter, has evolved from a mere communication tool to a crucial tool for information dissemination and aiding in identifying individuals in need. It has been widely used by official accounts, informational accounts, and citizens, aiding rescue and relief efforts during crises, demonstrating its transformative impact on individuals (Aldamen and Hacimic 2023). There are complicated interactions between social media and traditional media. However, it is obvious that social media has evolved into a tool for traditional media reporting; just consider how frequently a tweet from @realDonaldTrump follows a news item about the president. At the same time, stories created by traditional news media sites make up a large portion of the political content shared on social media (Tucker et al. 2018).

Media polarization is both a cause and an effect of political polarization. Individuals follow media outlets that represent their political ideas in this process. As a result, people have the opportunity to interpret the news through the lens of their worldview. Polarization in the media may reach a stage where people who have quite diverse perspectives on the world and find it impossible to grasp what others think surround followers of various political parties (Erdoğan 2018). The inception of social media was attributed to political polarization, according to many researchers. They claim that social media creates “echo chambers” in which a consumer and his friends are not able to view material that does not comply with their point of view (see, for example, Pariser 2011; Sunstein 2009; Mutz 2006; Hindman 2008), and that social media practices encourage this situation (see, for example, Sambrook 2016; El-Bermawy 2016; Allcott et al. 2019). More recent research claims that social media does not necessarily raise polarization (Boxell et al. 2017), but rather can minimize it by introducing users to a variety of viewpoints and therefore lead to less narrow political views (Dubois and Blank 2018; Algan et al. 2019; Barber’a 2015). This is because social media such as Twitter or Facebook spans a far bigger network than just direct connections, exposing individuals to other political perspectives through their weak ties and hence making them less likely to have extreme political views. Both points of view appear to be rational and have empirical backing (Campbell et al. 2019).

In terms of its capability of increasing polarized ideas on the Internet, social media has received considerable interest in comparison with other media outlets due to its ability to create networks between people who share the same perspectives (Hong and Kim 2016). Moreover, other researchers claim that the open structure of web 2.0 permits users to consume a diverse range of ideological perspectives, which was not available before the Internet (Garrett et al. 2011; Mutz and Mondak 2006; Hong and Kim 2016).

Echo chambers had opposite effects than expected. Members in each group leaned toward the opinions of the other group. This unexpected result happened because the group that the study was conducted on was egalitarian; there were no influencers among them. Therefore, ideas were spread among the group based on their quality rather than the person holding them. Social media platforms are centralized, meaning that a small number of people or just one “influencer” at the center is connected to the majority of normal users at the “periphery”. Ideas are filtered through, or sometimes even blocked, by a powerful social influencer on centralized networks, like many social media sites. In a centralized echo chamber, even a modest amount of partisan bias displayed by the influencer can be magnified across the entire group (Centola 2020).

Users may polarize over time if they are exposed to mostly attitudinal material because of their social and algorithmic recommendations. Online selective exposure, according to Lelkes et al. (2017), is a likely cause of their finding that broadband internet access is linked to increasing political and emotional polarization. However, among selective exposure researchers, there is still controversy and conflicting findings concerning the effect of social media on the media material people choose to consume (Beam et al. 2018).

Growing political involvement caused by social media has indirectly contributed to political polarization, even though there are no direct effects of social media utilization (Lee et al. 2018). Active social media users were more inclined to contribute to political processes, which resulted in them adopting more radical political beliefs over time than those who did not. Discussing political differences can widen the cleavage in the viewpoints of those participating in political debate and aggravate polarization, according to some studies (Lee et al. 2018).

4. Echo Chambers and Polarization

The group polarization theory indicates the proclivity of a group to make decisions that are riskier than the average individual decisions made by members before the group meeting. Subsequent social psychology studies revealed that a similar tendency applies more broadly to changes in attitude and opinion following a debate (Proietti 2017).

An echo chamber can operate as a mechanism to reinforce an existing view inside a group and, consequently, push the entire group toward more extreme perspectives. Echo chambers have been identified as a growing result of human characteristics such as selective exposure, contagion, and group polarization, according to certain research. However, the effects of echo chambers, as well as their very existence, have lately been called into doubt. Users can communicate in a variety of ways on different platforms, ranging from retweets and mentions on Twitter to likes and comments in groups on Facebook, resulting in quite varied social dynamics (Cinelli et al. 2021).

The concepts of echo chambers and filter bubbles represent the widespread public apprehension that the use of social media will reduce the number of pieces of information users view or absorb online, blocking a common pattern of knowledge transfer. The concern is that social media algorithms, when paired with a desire to engage with like-minded individuals, will build an environment in which users will be exposed to pleasant, opinion-reinforcing material at the expense of more diverse, opinion-challenging information.

Researchers’ capacity to examine the presence or development of echo chambers and filter bubbles is hampered by a lack of consensus over their conception and quantification, despite decades of interest in the topic (Kitchens et al. 2020). Indeed, there are elements of popular social media sites that may encourage information-limiting settings. People in online social networks, for example, communicate more frequently with like-minded people, just as they do in physical interactions. Bakshy et al. (2015) revealed that more than 80 percent of Facebook connections had the same party affiliation in a survey of 10.1 million U.S. Facebook users with self-reported ideological affiliation (Kitchens et al. 2020). Individuals prefer to consume opinion-reinforcing news sources over opinion-challenging news sources, even within this already confined option set, according to selective exposure theory (Frey 1986). Garrett (2009) found evidence for this trend in a field experiment with 727 online news consumers, who reported interest in reading online news items that they judged to be supportive of their current perspective and reluctance to read stories that challenged their existing opinion.

Scholars have long been concerned that algorithmic filtering might restrict the variety of information sources that people are exposed to, engage with, or consume. Social media networks have been found in several pieces of research to be information-expanding. Users can utilize social media to find new information sources, thus broadening the range of ideas, opinions, and information to which they are exposed (Kitchens et al. 2020).

Because the majority of Turkish social media users who observe political headlines choose media and people who share their existing partisan beliefs, most of the Turkish audience members live in echo chambers, responding primarily to sources and individuals within their party groups. Confirmation bias is widespread among news consumers, which is worsened by partisanship and low-quality news production throughout the Turkish digital media industry (Kirdemir 2020). This polarization is not curable with social media. A Twitter account is used by one third of all internet users. Only 15 percent of them routinely post political thoughts on Twitter, while almost half of all users never do so. Furthermore, 60 percent of Twitter users claim to follow individuals who share their viewpoints. The situation is no different for Facebook users, who account for 87 percent of all internet users. Only 7 percent of Facebook users express their political views, while 56 percent never do so. Two-thirds of Facebook users say their Facebook friends share their political viewpoints (Erdoğan 2018). Online and offline echo chambers, as well as “selective exposure to information,” are blamed for the problem. People in divided cultures prefer to access information from sources they consider to be politically aligned with their views or from sources that give information they already believe to be trustworthy. The entire isolation of knowledge from the outside is the consequence of such a communal process. Because of the focus on echo chambers, it was assumed that exposure to other or opposing viewpoints would reduce polarization and foster more moderate views and compromise (Kirdemir 2020).

Partisanship and polarization are interwoven with misinformation, fake news, and other types of lies across the Turkish digital news ecosystem. The overall reality is depicted either as a conflict between a prevailing and rising country and internal and external players who want to destroy it or as a total disaster in which an inept government conspires against its inhabitants. As a result, polarization, poisonous remarks, personal assaults, and organized campaigns target opposing individuals and political parties throughout online social networks (Kirdemir 2020). It is more and more likely that one of the main objectives of internet propaganda, which is frequently spread by automated social media accounts known as “bots,” is to ensure that some traditional media news pieces are viewed more frequently than others (Sanovich et al. 2018).

Most of the previous research regarding this topic is concentrated on the U.S. population, whereas this study uses a sample of Turkish students, which contributes an international perspective. This study has significant implications for the developing discipline of computational social science as well as continuing initiatives to minimize political division on the Internet.

However, several researchers have contradicted this viewpoint, claiming that exposure to sources from the opposite party may increase rather than decrease political and social division. Receiving a consistent message from the other side may have the unintended consequence of increasing division among politically active liberal and conservative groups in the United States. Although the topic has yet to be answered, it is reasonable to conclude that the free flow of information and competing viewpoints across political groups do not inevitably help reduce polarization (Kirdemir 2020).

The perception that the opposition parties, which are the main elements of the structural opposition in democratic systems, do not have the potential to be an alternative to power in the eyes of society has paved the way for rapid tension and polarization in the political field in Türkiye (Tuncel 2014).

5. Methodological Framework of the Study

The Turkish society is more technologically evolved than other societies; thus, Turkish students use numerous media tools for various gratifications, such as learning, education, work, and business, information access, cultural and social interests, connecting with networks and families, establishing friendships, learning new skills, self-expression, conducting business, and finding employment (Aldamen 2023c).

Because Turkish students are skilled at using various technical tools prior to the COVID-19 pandemic, the results produced by these tools for them were deemed superior to those produced by other students (Mohammad and Aldamen 2023). Thus, the rising amount of usage of social media among students’ age groups dictated the decision to choose students for this study. In addition, this age group is eligible to vote, which means that their opinions will have a significant effect on any upcoming election, referendum, etc.

The gap that the study tries to cover is the lack of comparative research on social media, particularly in the area of news consumption and polarized political opinion from the perspective of students. Consequently, the study aims to know if using social media is linked to having a polarized political opinion. This will be achieved by examining to what extent social media is creating echo chambers among students.

As a result, it aims to shed light on how Turkish students deal with social media platforms and how they are affected by them in terms of political polarization.

Depending on that, the study hypothesized that social media has an effect on political polarization among Turkish students.

Depending on the above hypothesis, the study poses this main question:

Is there a link between social media platforms and political polarization in Türkiye among Turkish students?

The sub-questions are the following:

• Is the use of social media linked to the level of political polarization among Turkish students?
• Do the top four platforms used in Türkiye (Facebook, Instagram, Twitter, and YouTube) contribute differently to creating echo chambers?
• Can social media be a solution to the spiral of silence for students? In other words, do they feel more comfortable sharing their thoughts about them?
• Does social media have a polarizing or depolarizing effect? i.e., does it help users to be more open to the ideas of the supporters of their opposing party?
• Do social media algorithms contribute to creating filter bubbles through the content they suggest to the user?
• How much do Turkish students depend on social media to form their political news?
• The study has the following limitations:
• Spatial limits: The study includes all Turkish university students enrolled in the Republic of Türkiye, ages 18–50+, at the educational levels of undergraduate, Master’s, and PhD.
• Time limits: Both quantitative and qualitative tools of the study were conducted between 26 April and 1 June 2022.

5.1. Data Collection Methods

Fieldwork data were gathered quantitatively and qualitatively via a base of 303 survey questionnaires from a sample of Turkish students as well as a focus group discussion conducted with six Turkish students.

Quantitative Method: A questionnaire consisting of 25 questions (multiple-choice, Likert scale, and an open question) was conducted between 26 April and 16 May 2022, using Google Forms. The participants were Turkish university students in Türkiye aged between 18 and 50 and above, including undergraduates, graduates, and PhD levels.

A questionnaire was used because it is considered to be an effective way to gather a variety of information from a large group of people, often known as respondents (Roopa and Rani 2012).

The study used a questionnaire to obtain data from respondents who were picked from different cities, ethnicities, backgrounds, education levels, and universities, both public and private, genders and ages, and from multiple scientific and humanities departments.

The first few respondents were chosen by convenience sampling, and they were asked if they knew other Turkish students with similar ideas or concerns who might be the right choice to participate in the study. They received the link, and then someone who knew other potential students sent the URL of the online questionnaire to other students, asking them to send it to their colleagues, which resulted in the snowball sampling technique. This snowball technique allowed us to effectively reach more potential participants because we were acquainted with the first participants (Naderifar et al. 2017). The first participant was a law student who was interested in political topics and shared his political opinions and views on his Instagram account.

Since this study aims to measure the effect of social media, the targeted participants were students who were either active on social media or use it daily.

The following parts were covered in the questionnaire:

• The students’ demographic data.
• The user’s behaviors and usage of social media platforms.
• How algorithms affect the suggested posts on their newsfeed.
• A comparison between face-to-face communication and communication through social media in terms of expressing political views.
• The perspectives on whether social media has a polarizing or depolarizing effect.

5.1.1. Validity and Reliability Procedures

Distribution Pre-Test: It was carried out through the distribution of the questionnaire to a pilot sample of only 70 participants to make sure the questions were clear and understandable to the students. After collecting the answers, some required modifications were made to some questions regarding the Turkish language used to make the questions clear.

Reliability Procedures: To guarantee the reliability of the research tool, Cronbach’s alpha (), or so-called internal consistency, was employed. Questions in cases where the treatment was strong or positive were eliminated, and questions in cases where the coefficient was weak or negative were kept.

Table 1. Reliability Test.

	Corrected Item-Total Correlation	Cronbach’s Alpha If Item Deleted
1.  The news accounts or pages I follow are compatible with my political ideas.	0.328	0.463
2.  The political views of the accounts I follow on Twitter are compatible with mine.	0.169	0.501
3.  I use Twitter to share my political views.	0.195	0.495
4.  My Facebook friends’ political opinions match mine.	0.229	0.486
5.  I use Facebook to share my political views.	0.178	0.500
6.  The political views of the accounts I follow on Instagram are compatible with mine.	0.473	0.420
7.  I use Instagram to share my political views.	0.216	0.490
8.  The political views of the accounts I follow on YouTube are compatible with mine.	0.532	0.395
9.  I use YouTube to share my political views.	0.248	0.497
10.  News that appears on my newsfeed that is suggested by the social media platform is usually compatible with the ideas of the party I feel close to.	0.045	0.529
11.  I engage in political conversations using my social media account with people from the party I feel most distant from.	0.048	0.520
12.  I engage in political conversations using my social media account with people from the party I feel the closest to.	0.344	0.464
13.  Social media has helped me to be more open to the ideas of the party I feel most distant from.	0.118	0.519
Cronbach’s alpha	N of Items
0.515	14

shows the value of the reliability coefficient for each axis of the questionnaire.

The previous table indicates that the stability coefficient, Cronbach’s alpha, is 0.515, which is acceptable for conducting the study. As in practice, values between 0.5 and 0.7 are acceptable (Hansjosten 2015).

5.1.2. Qualitative Method

A focus group discussion was conducted in the Turkish language with six Turkish students from different education levels on the first of June for around 60 min via Zoom. The characteristics of the participants are shown in

Table 2. Focus group discussion participants’ demographic data.

Participant	Gender	Education	Age
1	Female	Undergraduate	25
2	Male	Master’s degree	25
3	Female	Master’s degree	25
4	Male	Master’s degree	28
5	Male	Master’s degree	25
6	Male	Undergraduate	22

. The main aspects of the focus group discussion were finding out why Turkish students do not show interest in expressing themselves politically on social media and why they do not trust the news they find on social media.

The main aim of this focus group discussion was to elaborate further on the results of the questionnaire and ascertain a better explanation of specific points. The axes of the focus group discussion are shown in

Table 3. Focus group discussion axes.

Number	Axis
1	The respondents’ demographic data
2	Trust in social media
3	Interest in social media
4	Self-expression on social media

.

6. Qualitative and Quantitative Results

The questions were answered by extracting arithmetic averages, standard deviations, frequencies, and percentages according to the questions related to them, and the following scale was used in some questions: (1–2.33 low, 2.34–3.67 medium, and 3.68–5 high).

6.1. Quantitative Results

The study community is composed of Turkish university students from all provinces of Türkiye. The questions were answered by extracting arithmetic averages, standard deviations, frequencies, and percentages according to the questions related to them. The demographic data were drawn by frequencies and percentages as follows:

It is noted from

Table 4. The participants’ demographic data.

Variable	Level	Number	Percentage
Gender	Male	169	55.80%
	Female	134	44.20%
	Total	303	100.0%
Education	Undergraduate	271	89.40%
	Masters	26	8.60%
	Ph.D.	6	2.00%
	Total	303	100.0%
Age	18–28	199	65.70%
	29–39	56	18.50%
	40–50	37	12.20%
	51	11	3.60%
	Total	303	100.0%

that the study sample is divided according to the gender variable into (169) males and (134) females, while the sample according to educational qualification consisted of (271) undergraduates, (26) master’s holders, and (6) Ph.D. students. The table also showed that the sample according to the age variable consists of (199) of those aged (18–28), (56) of those aged (29–39), (37) of those aged between 40–50, and (11) of those aged 51 and over.

Table 5. The cities in which the sample members live.

City	Number	Percentage
Bitlis	82	27.06%
Istanbul	40	13.20%
Bursa	19	6.27%
Izmir	19	6.27%
Eskişehir	13	4.29%
Ankara	12	3.96%
Denizli	11	3.63%
Mersin	10	3.30%
Malatya	9	2.97%
Amasya, Gaziantep, Kirklareli	6 for each	1.98% for each
Adana, Antalya, Sinop, Van	5 for each	1.65% for each
Diyarbakir, Kayseri, Şanliurfa	4 for each	1.32% for each
Afyonkarahisar, Yalova	3 for each	0.99% for each
Bingöl, Kocaeli, Kirşehir, Konya, Kütahya, Mardin, Şirnak, Trabzon	2 for each	0.66% for each
Adiyaman, Aksaray, Aydın, Balıkesir, Batman, Çorum, Elaziğ, Hatay, Kahramanmaraş, Kars, Kilis, Muğla, Muş, Ordu, Rize, Samsun	1 for each	0.33% for each
Total	303	100%

notes that the majority of the study sample was from Bitlis and Istanbul, with percentages from the sample of 27.06 percent and 13.2 percent, respectively, followed by Bursa and Izmir with 6.27 percent. Thereafter, 4.29 percent of the participants were from Eskişehir, and between 1 and 12 participants were from the other provinces, with their percentages ranging between 0.33 percent and 3.96 percent.

6.1.1. User’s Behaviors and Usage of the Social Media Platforms

Table 6. Used social media platforms.

Social Media Platform	Number	Percentage
Instagram	238	It is not applicable from 303 here because the participants were able to use more than one answer
YouTube	173
Facebook	132
Other: (Tiktok 4, Telgram 7, Whatsapp 2, Linked in 3, Reddit 2, Ekşi Sözlük 1)	19

shows that Instagram is the most used application among Turkish students, with 238 participants saying that they use it. In second place is YouTube with 173 users, followed by Facebook with 132. Other apps such as Telegram, Tiktok, WhatsApp, LinkedIn, Reddit, and Ekşi Sözlük were also among those used by Turkish students.

Table 7. Number of times of checking social media accounts per day.

Variable	Number	Percentage
1–4 times	61	20.1%
5–10 times	109	36.0%
More than 10 times	133	43.9%
Total	303	100%

notes that 43.9 percent check their accounts more than 10 times, followed by those who check their accounts 5–10 times, with a percentage of 36 percent.

As we notice in

Table 8. One-way analysis of variance according to the number of times of checking social media accounts per day.

	Sum of Squares	df	Mean Square	F	Sig.
Between Groups	16.609	3	8.304	11.934	0.000
Within Groups	208.753	300	0.696
Total	225.361	303

, there are differences in the answers of the sample members due to the variable number of check times.

Table 9. Time spent on social media.

Variable	Number	Percentage
Less than 1 h	46	15.2%
1–3 h	136	44.9%
More than 3–5 h	80	26.4%
More than 5 h	41	13.5%
Total	303	100%

indicates that 44.9 percent check their accounts for more than 1–3 h, followed by those who check their accounts for 3–5 h with a percentage of 26.4 percent.

Table 10. One-way analysis of variance according to the number of check-in hours variable.

	Sum of Squares	df	Mean Square	F	Sig.
Between Groups	45.218	3	15.073	25.017	0.000
Within Groups	180.144	299	0.602
Total	225.361	302

shows that there are differences in the answers of the sample members due to the variable number of hours of checking.

Table 11. The effect of the top four used social media platforms (Facebook, Instagram, Twitter, and YouTube) in creating echo chambers differently.

n	Question	Mean	sd	Scale
1	The news accounts or pages I follow are compatible with my political ideas	3.03	1.12	Moderate
2	The Political Views of The Accounts I Follow on Twitter are Compatible With Mine	2.08	1.55	Low
3	I use Twitter to share my political views	1.35	1.47	Low
4	My Facebook Friends’ Political Opinions Match Mine	1.1	1.45	Low
5	I use Facebook to share my political views	0.77	1.27	Low
6	The Political Views of The Accounts I Follow on Instagram are Compatible with Mine	1.84	1.55	Low
7	I use Instagram to share my political views	1.01	1.07	Low
8	The Political Views of The Accounts I Follow on YouTube are Compatible With Mine	1.73	1.53	Low
9	I use YouTube to share my political views	0.82	1.01	Low

notes that the effect of the four social media platforms (Facebook, Twitter, Instagram, and YouTube) was low.

6.1.2. Face-to-Face Communication or Communication through Social Media in Terms of Expressing Political Views

Table 12. Engagement and expression on social media.

n	Question	Mean	sd	Scale
1	I engage in political conversations using my social media account with people from the party I feel most distant from.	2.12	1.24	Low
2	I engage in political conversations using my social media account with people from the party I feel the closest to.	2.23	1.26	Low
3	I express my political views on social media more comfortably than in face-to-face communication.	2.37	1.26	Low

shows that social platforms were not a solution to the spiral of silence, as their responses to this area were low.

6.1.3. Social Media Has a Polarizing or Depolarizing Effect

Table 13. Do you follow the accounts of leaders of the party you feel closest to.

Answer	Frequency	Percent
Yes	129	42.6
No	174	57.4
Total	303	100%

indicated that 57.4 percent of Turkish students do not follow the accounts of the leaders of the party to which they feel close.

It is noted from

Table 14. Do you follow the accounts of leaders of the party you feel furthest from?

Answer	Frequency	Percent
Yes	132	43.6
No	171	56.4
Total	303	100%

that 56.4 percent of Turkish students do not follow the leaders of the party they feel furthest from, while 43.6 percent do.

Table 15. One-way analysis of variance according to how much the students feel different from their furthest party.

	Sum of Squares	df	Mean Square	F	Sig.
Between Groups	19.098	10	2.122	3.014	0.002
Within Groups	206.263	293	0.704
Total	225.361	303

shows that there are differences in the answers of the sample members due to the variable number of how much the students feel different from their furthest party.

Table 16. Does social media have a polarizing or depolarizing effect? In other words, does it help users to be more open to the ideas of the supporters of their opposition party?

n	Question	Mean	sd	Scale
1	Social media has helped me to be more open to the ideas of the party I feel most distant from.	2.38	1.27	Moderate
2	Social media made me agree with the ideas of my opposing party on certain incidents (e.g., internal affairs, economy, refugees, etc.)	2.43	1.26	Moderate
3	I trust the political news on social media and use it to learn about other political parties.	2.98	1.30	Moderate
4	Social media makes my relationship with politics stronger.	2.90	1.28	Moderate
5	I can share my criticisms against political parties that I feel distant from on social media without being sure of their accuracy.	1.98	1.16	low

pointed out that the effect of social communication on polarization was medium and low, in all sections of the field.

Table 17. Ideologically, how differently do you see yourself from the supporters of the party you feel farthest from on a scale of 1–10?

Level	Frequency	Percent
1	6	2.0
2	3	1.0
3	6	2.0
4	6	2.0
5	22	7.3
6	18	5.9
7	34	11.2
8	61	20.1
9	29	9.6
10	118	38.9
Total	303	100%

notes that 38.9 percent had a high degree of disagreement with the party furthest from them at 10.

6.1.4. Algorithms Affect and Create Filter Bubbles through the Content They Suggest to the User

Table 18. Do social media algorithms contribute to creating filter bubbles through the content they suggest to the user?

n	Question	Mean	sd	Scale
1	News that appears on my newsfeed that is suggested by the social media platform is usually compatible with the ideas of the party I feel close to.	2.95	1.18	Moderate

indicates that social media algorithms contribute to creating filter bubbles through the content they suggest to users to a medium degree.

6.2. Qualitative Results

The focus group discussion was transcribed and some important extracts were drawn from the discussion, as

Table 19. The main extracts of the focus group discussion.

	Extract	Participant
1	“Because there is a lot of fake news on social media and it spreads very quickly. In addition, the news is made informally without specifying the source, so the news on social media is not very reliable.”	5
2	“I follow the celebrities I like, fashion, food, sports, etc.”	3
3	“Critics are also afraid of what would happen to them if they were comfortable expressing themselves. Fear that it will affect their work and life in the future.”	1
4	“No, because I don’t use social media to express my thoughts.”	4
5	“I would watch the celebrity because political leaders often repeat the same thing, then if I watch the politician on another broadcast, I am sure I will hear the same things.”	5
6	“I would prefer to follow the political leader because what they say and do has more impact on our daily life.”	4

(Source: Authors).

shows.

7. Discussion of the Results

The majority of the Turkish students expressed that they felt distant from their opposing party at a maximum level. The social distance that political party supporters feel toward “the most distant” political party followers is one of the most important indices of political division (Erdoğan 2018). As the questionnaire’s data show, 38.9 percent of the participants had a high degree of disagreement with the party they felt was the furthest away. This gives us an indication that the more students used social media, the further they felt from their opposing party, thereby increasing the level of political polarization.

The results showed that Turkish students do not trust the news they see on social media. Regarding that issue, the participants in the focus group mainly stated three reasons for it. First, they believe that the news on social media focuses more on speed than credibility, as it does not specify a source, as it is understood from Extract 1. They also believe that social media is full of disinformation and that the news there is not objective. Social media is not solely used for keeping up with political developments; there can be other uses as well. When the participants were asked about the subjects they follow, their answers included subjects like music, food, fashion, sports, and entertainment, as the participant in Extract 2 stated. Moreover, the majority of the Turkish students conveyed that they follow news pages on social media that are compatible with their preexisting ideas, as 70.7 percent of the participants agreed to the statement “the news accounts or pages I follow are compatible with my political ideas” on different levels. That gives us an indication that social media in general creates echo chambers, meaning that they create environments where people around them reinforce their opinions and they do not have to face any different or challenging ones.

As well, most of the students said that they would not share their thoughts on a fake account. Extract 4 is an example from the focus group discussion when students were asked whether they use fake accounts or not. The quantitative results showed that about half of the Turkish students do not follow political leaders on social media, regardless of whether they are from the party they support or not. In order to explore that more, the students were asked: “If a celebrity and a political leader you liked were on a live broadcast at the same time, which would you watch and why?”.

The focus group discussion showed the same results: half of them said they would watch the celebrity, while the other half said they would watch the political leader, which was clearly stated in Extracts 5 and 6.

After the comparison among platforms, the results showed that Twitter is the one that contributes the most to creating echo chambers, with 51.7% of its users saying that the political views of the accounts they follow on Twitter are compatible with theirs on different levels. YouTube, on the other hand, is the least popular; most YouTube users either do not have an account on it or follow channels that hold different views from theirs. Instagram and Facebook came second and third place, respectively. The different nature of each platform justifies the result. However, all of the top four platforms used in Türkiye (YouTube, Instagram, Twitter, and Facebook) had a low effect.

The questionnaire answers indicated that the vast majority of the Turkish students do not prefer to share their thoughts on social media, regardless of the platform. The focus group discussion showed that the main reason behind that is that they are afraid that their posts and comments are being censored, which can affect their future, as the participant in Extract 3 clarified.

According to the spiral of silence of Noelle-Neumann (1974), minority groups are not usually willing to speak out about their opinions because they are afraid of being isolated and cut off from their social connections. Individuals will rather agree with the consensus stated in their social surroundings than offer an opposing opinion or viewpoint. The spiral of silence theory states that people refrain from revealing their political views if they believe they are in the minority (Erdoğan 2018).

The same thing applies to social media as well, in which minority groups, including vulnerable groups such as, for example, refugees, hide their opinions, views, and preferences when they think that they fall within a minority group. They fear social isolation through social media (Aldamen 2023b). Individuals who adopt an opinion that is supported by the majority in a given environment feel entitled to express their opinions freely, whereas others who have the opinion of the minority refrain from doing so. Individuals feel pushed to observe their environment due to their fear of social isolation. The results of this study have shown that a number of students abstain from expressing their opinions on social media, as Figure 1 shows.

Being in the minority makes it understandable that some people prefer to hide their opinions, views, and preferences. They prefer to be silent; maybe it is because they do not want to be attacked by the majority when their opinion does not fit public opinion. When people encounter aggressive comments, they prefer to keep silent and not be attacked by the majority when their opinion does not fit the public’s opinion. The silence deepened as the minority quieted down and the majority grew. This theory explains how public opinion is formed in a rapidly changing media environment. Therefore, the inclination of one group to express their opinion and the other does not generates a spiraling process where some opinions are heard and others are muted, irrespective of their representation in a community (Matthes 2015). People with extreme beliefs won’t blend in with the political scene. These people have views that are wholly unrelated to the general consensus at the moment.

Web 2.0 offers more platforms and opportunities to express opinions that disagree with popular opinion, but it still has a role in telling what popular opinion is. Social media can also play a role in enforcing the majority’s opinion (Aldamen 2023b).

Many people are exhausted by the amount of political content they have to face on social media, even though they appreciate its capacity for political information sharing and interaction. They feel uncomfortable when they argue about politics on social media with other users who do not hold the same views as theirs. When they are faced with a dispute and do not believe their friends or followers sympathize with them, Facebook and Twitter users are less likely to speak up (Hampton et al. 2014).

The results have shown us that students are afraid that their posts and comments are being censored, which can affect their future, as participant 1 in Extract 3 stated. Hence, social media did not offer a platform for people to speak out. We saw this when participants were asked whether they felt more comfortable expressing their political thoughts on social media than face-to-face, and 106 participants (35.3 percent) were negative about it.

When the participants were asked about whether social media has helped them be more open to others’ ideas or agree with the ideas of the opposing party on some internal issues, their answers showed a medium tendency toward social media being helpful in that matter, yet the numbers are not high enough to consider social media as a solution to the polarization issue. This is a good indication that social media can have a positive contribution in making people more open to each other’s ideas in certain cases, but since the effect was only moderate, to solve the polarization situation, other social and political aspects should be taken into consideration.

It is noted that social media algorithms contribute to a medium degree to creating filter bubbles through the content they suggest to the user. Therefore, the algorithms of the social media platforms show users content that is compatible with their views, restraining them to a medium degree from being open to opposing ones.

Social media algorithms also play a role in this process, as they work on customizing the suggested posts based on the users’ prior searches and interests, which limits the viewpoints that a social media user can view. Social media can play a positive role in depolarizing users. As the participants showed, they can agree on some points with the supporters of their opposing party. Depending on the answers to the questionnaire, the majority of the sample of the Turkish students do not trust the news they see on social media, and they do not follow political leaders. The reasons behind that were that they believe that the news is full of disinformation, that many articles are presented without a trustworthy source, that they are not objective, and that the news on social media relies more on speed than credibility.

The results showed that Turkish students do not rely on social media platforms to obtain political news, and most of them do not follow political leaders on social media. When students were asked about their trust in the news on social media, the participant in Extract 1 stated that they do not.

The results show that 57.4 percent of Turkish students do not follow the accounts of the leaders of the party to which they feel closest, while 56.4 percent of the Turkish students do not follow the leaders of the party to which they feel furthest. This can be for several reasons, such as a lack of trust in political news on social media, political leaders not being successful in addressing the students on social media, and a lack of interest in politics on social media and using it for other causes.

Social media does have an effect on political polarization and can play a role in increasing it. Having said that, this effect is low to moderate. This can be due to several reasons, mainly the lack of trust in the media and politicians, as well as the alternative uses and contents of social media such as art, entertainment, and sports. The results approved the hypothesis that social media has an effect on political polarization among Turkish students even if that effect is low or moderate, which can be explained to several reasons, above all the lack of trust in the media and politicians, but also the alternative uses and contents of social media such as art, entertainment, and sports.

As the results show, all the samples use social media at a high level in terms of time spent on them and the number of checks. In addition, the majority feel very distant from their opposing party. Most of the students stated that they do not express their political opinions on any of the top four platforms. When discussed in the focus group, the participants explained that they do not want to face backlash or possible problems.

8. Conclusions

Social media has become an integral part of our daily lives, especially among university students. However, the use of social media in political discourse has raised concerns regarding its impact on polarization. In Türkiye, social media has played a crucial role in shaping political views and opinions among university students. The widespread use of social media has enabled university students to engage in political discussions and express their views more freely than ever before. However, this freedom to express opinions has also led to increased polarization within society. Many university students in Türkiye are influenced by the content they see on social media platforms. While some use it as a means of building bridges and connecting with people from different backgrounds, others tend to reinforce their own beliefs by following like-minded individuals and engaging in discussions within echo chambers. This has led to a notable decline in civil discourse and increased hostility between opposing political groups.

Furthermore, the spread of disinformation and fake news on social media has exacerbated this issue. University students in Türkiye need to be aware of the impact of social media on polarization and actively work toward preventing its negative effects. It is important for university students to critically examine the content they consume and engage in civil discourse with individuals who have different perspectives. Social media have no great effect on political polarization but can play a role in increasing it. The Turkish students do not trust the news they see on social media, and they do not follow political leaders. The lack of trust in media and politicians, as well as the alternative uses and contents of social media such as art, entertainment, and sports, could affect that role. Students are afraid that their posts and comments are being censored, which can affect their future.

9. Limitations and Future Research

Despite the study’s framework, which employed mixed-methods research to collect and assess both quantitative and qualitative data, the research methodology has its limitations. Because circumstances vary over time, the same questionnaire and focus group discussions would provide different answers today and later or even after the Türkiye’s presidential elections happened in 2023.

Furthermore, the results of this study are not generalized to Turkish society overall, as the sample consisted only of an educated group, which is the Turkish student population in Türkiye. The study purposely elicited the views of individuals trained at an advanced level in critical thinking.

The viewpoints of Turkish students of different genders, qualifications, ages, and residence towns could aid in the direction of future study. This study tries to provide a starting point for exploring the relationship between social media and political polarization among Turkish students. Future research might be conducted to thoroughly examine a broader sample of citizens’ perspectives and compare the results to the findings of this study.