Analysis on the Refugees Representation in the Spanish Digital Press University of Granada, España Moldes-Anaya, Sergio; Khadour, Isabel. . REFERENCES CONTACT sergiomoldesanaya@gmail.com INTRODUCTION OBJETIVES MATERIALS AND METHOD RESULTS CONCLUSION The current refugee crisis that is affec0ng Europe is tes0ng the resolu0on capacity of the governments of each of the countries of the bloc. Countries like Italy or Austria are undergoing a process of radicaliza0on of migra0on restric0on policies, the demonstra0on of this, is the recent example of the refusal of the Italian government to rescue immigrants who remain adriC in the Mediterranean. Spanish stereotypes towards refugees are based, among other things, on the informa0on that media dissemina0on offers. In order to approach the research problem a discourse analysis will be carried out to verify how the representa0on of refugees in the Spanish media is and what are the stereotypes to which both representa0ons refer. The corpus analyzed consist in the related news published in the digital edi0on of two of the most widely read newspapers in Spain: "El País" and "ABC", from April 2015 to September 2017. On the one hand, cri0cal discourse analysis (CDA) will be used to study the linguis0c representa0on of refugees in both media and, on the other hand, the analysis of con0ngency tables to detect concordances intra and inter observer in the representa0on of the same in newspapers through Kappa od Cohen analysis. The objec0ve of the study will be to evaluate the representa0on of the refugees in both media, aCer analyzing a series of previously established discursive categories that will be evaluated by external judges. In accordance with the criteria of external judges and subsequent analysis of the opinions, the expected results should show that the representa0on of refugees varies significantly in both media, while in one mass media there will be related to immigra0on control and migra0on policies in another, categories related to Human Rights and Interna0onal Coopera0on will predominate. DIGITAL NEWSPAPER “El País” Observer A ! Observer B ! Observer C ! Observer D ! Observer E ! Observer F ! A-B .373*** B-A a .373*** C-A a .280*** D-A a .356*** E-A a .081* F-A a .344** A-C .280*** B-C .253*** C-B a .253*** D-B a .782*** E-B a .082* F-B a .241*** A-D .356*** B-D .782*** C-D .281*** D-C a .281*** E-C a -.006 F-C a .369*** A-E .081* B-E .082* C-E -.006 D-E .081* E-D a .081* F-D a .271*** A-F .344*** B-F .241*** C-F .369*** D-F .271*** E-F -.008 F-E a -.008 DIGITAL NEWSPAPER “ABC” Observer A ! Observer B ! Observer C ! Observer D ! Observer E ! Observer F ! A-B .041 B-A a .041 C-A a .188*** D-A a .243*** E-A a .331*** F-A a .385*** A-C .188*** B-C .039 C-B a .039 D-B a .139*** E-B a .088** F-B a .060* A-D .243*** B-D .139*** C-D .135*** D-C a .135*** E-C a .127*** F-C a .249*** A-E .331*** B-E .088** C-E .127*** D-E .233*** E-D a .233*** F-D a .171*** A-F .385*** B-F .060* C-F .249*** D-F .171*** E-F .152*** F-E a .152*** Table 2. Inter-observer concordance. Note: a The result is redundant. *p <05, **p<01, ***p<001. Source: Self-made Observer ! A1-A2 .151** B1-B2 -.007 C1-C2 .020 D1-D2 .079 E1-E2 .008 F1-F2 .080* Table 4. Intra-observer concordance El País versus ABC. Note: *p <05, **p<01, ***p<001. Source: Self-made Kappa Coefficients Strength of Agreement 0.00 Poor 0.01 to 0.20 Slight 0.21 to 0.40 Fair 0.41 to 0.60 Moderate 0.61 to 0.80 Substantial 0.81 to 1.00 Almost perfect Table 3. ValuaGon of the kappa coefficient Source: Landis & Koch, 1977. The cri0cal discourse analysis, as exposed by Van Dijk (2003), has been our star0ng point in this brief analysis. Since the concep0on of this work and some of the specific objec0ves require a quan0ta0ve approach, it has been chosen to carry out an analysis of Kappa of Cohen (Du Toit, Swingler & Iloni, 2002) with the correspondent con0ngency table and a sta0s0cal waste analysis. It was used a textual corpus of 142 daily online news drawn from Spanish newspapers. • Du Toit, G., Swingler, G. & Iloni, K. (2002). “Observer varia0on in detec0ng lynphadenophacy on chest radiography. The Interna*onal Journal of Tuberculosis and Lung Disease. 6 (9). pp. 814-817. • Landis, J., & Koch, G. (1977). The measurement of observer agreement for categorical data. Biometrics. 33 (1).pp 74-159. • Lobato, R. M., Moya, M., & Trujillo, H. (2015). The image of “conflict”: The content of the stereotype towards Pales0nian and Israelis is influenced by the media. Interna*onal Humani*es Studies, 2, 69–90. • Lobato, R.M., Moldes-Anaya, S., Trujillo, H., & Moya, M. (2016) “Approach to the representa0on of Pales0nians and Israelis in the Spanish online newspapers”, Third ISA Forum, Vienna, Austria. • Van Dijk, T. A. (2003). Ideología y discurso. Barcelona: Editorial Ariel. Newspaper Fr. of news (%) Political Ideology Date El País 71 (49%) Right-Wing From April 2015 to September 2017 Abc 72 (50%) Left Wing / Center Total 142 (100%) Table 1. News in the analysis Observer & Themes O,A O,B O,C O,D O,E O,F Total N (%) N (%) N (%) N (%) N (%) N (%) N (%) Migrations 20 (27.8) 23 (31.9)* 6 (8.3) 23 (31.9)* 14 (19.4) + 8 (11.1) 94 (22.8)* ositive Action 0 6 (8.3) 7 (9.7) 6 (8.3) 12 (16.7) 4 (5.6) 35 (8.5) ocial Problems 16 (22.2) 6 (8.3) 2 (2.8) 5 (6.9) 7 (9.7) 9 (12.5) 45 (10.9) egal process 2 (2.8) 2 (2.8) 3 (2.4) 2 (2.8) 0 8 (11.1) 17 (4.1) sts and demands 8 (11.1) 10 (13.9) 9 (12.5) 10 (13.9) 7 (9.7) 11 (15.3) 55 (13.3) nflict resolution 1 (1.4) 0 1 (1.4) 0 14 (19.4) + 5 (6.9) 21 (5.1) Violence 3 (4.2) 2 (2.8) 24 (33.4)* 2 (2.8) 5 (7) 10 (13.9) 46 (11.1) Politics 22 (30.6)* 19 (20.9) 20 (27.8) 19 (26.4) 2 (2.8) 17 (23.6) 82 (19.9) Others 0 0 0 5 (6.9) 11 (15.3) 0 16 (3.8) Total observations 411 (100%) Table 5. Relevance of topics according to the observers (El País) Note: * Most relevant topic. Observer & Themes O,A O,B O,C O,D O,E O,F Total N (%) N (%) N (%) N (%) N (%) N (%) N (%) Migrations 13 (18.3) 6 (8.5) 2 (2.8) 10 (14.1) 20 (28.2)* 7 (9.9) 58 (13.8) ositive Action 6 (8.5) 6 (8.5) 7 (9.9) 5 (7) 9 (12.7) 6 (8.5) 39 (9.2) ocial Problems 7 (9.9) 14 (19.7) 3 (4.2) 5 (7) 1 (1.4) 0 30 (7.1) egal process 0 5 (7) 0 2 (2.8) 1 (1.4) 7 (9.9) 15 (3.5) sts and demands 4 (5.6) 5 (7) 11 (8.7) 17 (23.9) 9 (12.7) 9 (12.7) 55 (13.1) nflict resolution 0 5 (7) 0 2 (2.8) 6 (8.5) 0 13 (3.1) Violence 17 (23.9) 5 (7) 26 (36.5)* 7 (9.8) 11 (15.5) 19 (26.8)* 85 (20.2) Politics 24 (33.8)* 19 (26.8)* 22 (31) 20 (28.2)* 7 (9.9) 17 (23.9) 109 (25.9)* Others 0 6 (8.5) 0 3 (4.2) 7 (9.9) 0 16 (3.8) Total observations 420 (100%) Table 5. Relevance of topics according to the observers (ABC) Note: * Most relevant topic. As shown in Tables 2 and 3, the representa0on of the main themes within each newspaper varies from one medium to another. On the one hand, with respect to the newspaper "El Pais" it has been detected that the most prevalent discursive theme is "migra0ons". Although discursive themes such as "poli0cs" or "protests and demands" also have a significant representa0on. On the other hand, regarding the newspaper "ABC" the most prevalent discursive themewould be the one that has to do with "poli0cs". Although, as in the previous case, there are other discursive themes that may be significant, such as those related to "violence", "migra0ons" or "protests and demands". Likewise, in rela0on to the levels of inter-observer agreement in the newspaper analysis "El Pais", it is observed in table 4 that in general there is a reasonable agreement, especially between B and D and vice versa. Regarding the levels of inter-observer agreement in the analysis of the newspaper "ABC", it is observed that the conconcordance is somewhat weaker when compared with the newspaper "El Pais" which is "reasonable" towards a "slight" concordance in the ABC. Finally, in terms of the intraobserver agreement, it has been detected that the agreement between the analyzed digital media is hardly agreed, that is, the agreement is poor.