The global refugee crisis has displaced 82.4 million people (UNHCR, 2020). Turkey is the most refugee-receiving country worldwide, with 3.9 million refugee people as of 2020, not to mention many undocumented refugees coming from Afghanistan. Among these refugees, 1.4 million refugees are under 15, and over 800 thousand are aged between 15 and 24 (UNICEF, 2019). As the statistics show, Turkey has a large number of school and university-aged refugee youth.  

Education, a fundamental human right that states are obliged to ensure, is vital for refugee people since it paves the way for social integration through learning the societal rules, customs, and values (Unutulmaz, 2019). Furthermore, it is a source of hope for both refugee families and children (Sidhu & Taylor, 2012; Thomas et al., 2015). Even though there are many positive outcomes of refugee education for both refugees and the host country, various stakeholders of the Turkish education system seem to tend to conceive refugee education as one of the principal problems of the system (Celik et al., 2021). 

Massive refugee influx into Turkey in a brief period has aroused mixed reactions from local actors. Limited research revealed that the negative reactions are more prevalent, and the local people in Turkey blame refugees for the economic issues (İçduygu & Şimşek, 2016) and an increasing number of adverse events (Çelik & İçduygu, 2018). It is evident that there is a growing xenophobic attitude among local people towards refugees in Turkey (Erdoğan, Kirişçi, & Uysal, 2021). Considering that social media can exacerbate xenophobia and discriminatory attitudes in societies (Bursztyn et al., 2019; Müller & Schwarz, 2020; Ozduzen et al., 2021), there is a need to investigate local actors'  social media discourses on refugee education.

Many people convey messages that reflect their attitudes and perceptions on social media websites as a part of their communication. Twitter is among the most popular social media sites, with 211 million active users (Statista, 2011). Its users share messages including text, links, images, and videos within a 280-character limit. These accessible and voluntarily created messages, known as tweets and posts, are suitable for data mining and analysis (Felt, 2016). Therefore, this platform is handy to investigate global and local discourses on various issues that users classify through hashtags. 

To date, only a limited number of studies have focused on the local actors’ views on refugees, and there are even fewer studies that have used social media or big data for investigating this issue. Since education is a crucial factor for the adaptation of refugees into host countries, exploring the host society discourse on refugee education helps frame the attitudes towards refugees. Focusing on the microblogging site Twitter, this study aims to profile the discourse on refugee education by local actors in Turkey on social media and seeks to address the following questions:

1. Which topics are mentioned in tweets about both refugees and education?
2. What type of sentiment (positive, neutral, or negative) occur in tweets mentioning both refugees and education?
3. What are the prevailing discourses within tweets about refugees and education?

This study aims to profile the discourse on refugee education by local people in Turkey. Social media facilitates strong interaction to understand the social structure. Social actors’ and institutions’ exclusivist discourses on social media can be given as an example to understand how current social structures are constructed and maintained over time (Reid et al., 2010). By combining sentiment and qualitative content analyses, the common topical patterns and sentiments in the tweets disseminated under various hashtags by the local actors were investigated and extracted. Most of the studies based on data obtained from Twitter have focused on tweets indexed by hashtags that help users follow and contribute to ongoing discussions. Similarly, this research mined tweets that were posted from 1 January 2019 to 31 December 2021 under 23 refugee-related hashtags, including the Turkish equivalent of #SyriansShouldGotoSyria, #MyOpinionAboutRefugees, #WeWontTransformIntoARefugeeCamp, by using the academictwitteR package of R (Barrie & Ho, 2021). From our query, we captured about 317 thousand tweets, which consisted of predefined hashtags. Because there is a risk of excluding hashtags that we are not aware of, we sorted the most used hashtags in our data set and repeated our query with 12 additional hashtags found frequently used by Twitter users but were not included in the initial query. As a result, our final data set comprises 540 thousand tweets related to 35 hashtags about the refugee issue. To find education-related tweets, we filtered tweets based on a criterion that whether the tweets include educational terms such as education, school, teacher, and student. After eliminating the duplicated tweets and the tweets containing only hashtags, our final data included about 17,000 tweets related to both education and refugees. Qualitative content analysis was used to examine the data obtained from Twitter since it is intended in this study to analyze common patterns in the texts disseminated by actors on Twitter. Pre-established themes and codes were not used to reveal unique discourses of the social actors (Yıldırım & Şimşek, 2016). Two researchers of this study coded the whole data simultaneously. One academic from the Educational Sciences Department coded the data for triangulation purposes. Suggested themes and codes were considered and finalized after the triangulation process.

Regarding the local discourses on refugee education in Turkey, three main themes emerged from the data as preliminary findings; (1) Turkish Context, (2) Intensification, and (3) Refugee Education. The Turkish Context theme focuses on the economic and political situation in Turkey, and it has four codes pre-existing educational problems, unemployment, economic crisis, and refugees as “guests.”. Although there is a high need for teachers in different regions in Turkey, there aren’t enough teachers assigned to schools because of the economic crisis. However, Syrian teachers may be assigned to schools for various EU projects, creating hate speech among unassigned Turkish teachers. Secondly, since Turkey does not give refugee status to the people who came from Middle Eastern countries in the scope of the Geneva Convention, refugees in Turkey were acknowledged as guests, religious fellows, and sisters/brothers in the first years of the war. Therefore, this issue creates an ambiguous process for locals, and local actors frequently expressed in their tweets that they do not want ‘guests’ anymore in their country. Intensification has three codes hate speech, gender-based discrimination, and denial of racism. For example, local actors in Turkey assert that male refugees should be soldiers and fight for their countries. It isn’t considered racism, but as a debt for their countries. The Refugee Education theme has three codes: the need for education, misinformation about grants, and distortion on refugee education. Locals agree that refugees should be educated for social rules and adaptation to Turkey. Locals didn’t get clear information about the sources of these grants, and they think that they afford all the needs of refugees with their taxes. In conclusion, this research provides a significant understanding of local actors views’ on refugee education in Turkey. This research is conducive to improving refugee education since it references the local discourse.

Bursztyn, L., Egorov, G., Enikolopov, R., & Petrova, M. (2019). Social media and xenophobia: Evidence from Russia. (NBER Working Paper No. 26567). National Bureau of Economic Research. http://www.nber.org/papers/w26567 Celik, S., Kardaş İşler, N., & Saka, D. (2021) Refugee education in Turkey: Barriers and suggested solutions, Pedagogy, Culture & Society. https://doi.org/10.1080/14681366.2021.1947878 Çelik, Ç., & İçduygu, A. (2018). Schools and refugee children: The case of Syrians in Turkey. International Migration, 57(2), 253-267. https://doi.org/10.1111/imig.12488 Erdoğan, M. M., Kirişci, K., & Uysal, G. (2021). Improving Syrian Refugee Inclusion in the Turkish Economy. World Refugee & Migration Council. https://wrmcouncil.org/wp-content/uploads/2021/09/Turkey-Syrian-Refugees-WRMC-final.pdf Felt, M. (2016). Social media and the social sciences: How researchers employ Big Data analytics. Big Data & Society, 3(1), 1-15. doi:10.1177/2053951716645828 İçduygu, A. & Şimşek, D. (2016). Syrian refugees in Turkey: Towards integration policies. Turkish Policy Quarterly, 15(3), 59-69. Müller, K. & Schwarz, C. (2020). From hashtag to hate crime: Twitter and anti- minority sentiment. SSRN. http://dx.doi.org/10.2139/ssrn.3149103 Ozduzen, O., Korkut, U., & Ozduzen, C. (2021). ‘Refugees are not welcome’: Digital racism, online place-making and the evolving categorization of Syrians in Turkey. New Media & Society, 23(11), 3349-3369. https://doi.org/10.1177%2F1461444820956341 Reid L, Sutton P and Hunter C (2010) Theorizing the meso level: the household as a crucible of pro-environmental behaviour. Progress in Human Geography 34(3): 309–327. https://doi.org/10.1177%2F0309132509346994 Sidhu, R., & Taylor, S. (2012). Supporting refugee students in schools: What constitutes inclusive education? International Journal of Inclusive Education, 16(1), 39–56. https://doi.org/10.1080/13603110903560085 Statista. (2011). Twitter: 2017-2021. Statista Research Department. https://www.statista.com/statistics/970920/monetizable-daily-active-twitter-users-worldwide/ Thomas, R. L., Chiarelli-Helminiak, C. M., Ferraj, B., & Barrette, K. (2015). Building relationships and facilitating immigrant community integration: an evaluation of a cultural navigator program. Evaluation and Program Planning, 55, 77–84. UNHCR (2020) Refugee data finder. Retrieved from: https://www.unhcr.org/refugee-statistics/ UNICEF. (2019). Education of children under temporary protection in Turkey: Statistical report. Retrieved from: https://www.unicef.org/turkey/media/10716/file/Educational%20statistics%20for%20children%20UTP%20.pdf Unutulmaz, K. O. (2019). Turkey's education policies towards Syrian refugees: A macro‐level analysis. International Migration, 57(2), 235-252. https://doi.org/10.1111/imig.12476 Yıldırım, A., & Şimşek, H. (2016). Nitel Araştırma Yöntemleri. Seçkin Publication.