Bu çalışmada, Arapça konuşma becerisi özyeterlik algısını ölçmeye yönelik bir ölçek geliştirmek amaçlanmıştır. Ölçeğin geçerlik güvenirlik çalışması, 2014-2015 eğitim öğretim yılı bahar döneminde Gazi Üniversitesi Gazi Eğitim Fakültesi'nde Arapça öğretmenliği bölümünde ve Kırıkkale Üniversitesi Fen-Edebiyat Fakültesi Arapça Mütercim-Tercümanlık bölümünde öğrenim gören toplam 99 katılımcı ile gerçekleştirilmiştir. Ölçek maddelerinin, Arapça konuşma özyeterlik algısı ile ilişkili olup olmadığını incelemek amacıyla düzeltilmiş maddetoplam puan korelasyonları hesaplanmıştır. Kaiser Mayer Olkin (KMO) katsayısı ve Barlett testi yapılmış ve KMO katsayısının .808 olduğu görülmüştür. Elde edilen verilerin faktör analizi için uygun olduğu sonucuna ulaşılmıştır, bundan dolayı düzeltilmiş madde-toplam puan korelasyon değeri .40 tan düşük olan maddeler ölçekten çıkarıldıktan sonra, ölçeğin yapı geçerliğine ilişkin kanıtlar elde etmek amacıyla açımlayıcı faktör analizi yapılmıştır. Yapılan faktör analizi sonucunda birinci faktörün açıkladığı varyans oranı, %27.591 çıkmıştır. Maddelerin ölçekte yer alabilmeleri için faktör yükünün en az 0.40 olması ve birinci faktör dışında yüksek yük değeri almıyor olması ölçüt alınarak faktör azaltma yoluna gidilmiştir. Bu maddelerin ölçekten çıkarılmasından amaç yapı geçerliğinin azalmasına engel olmaktır. Geriye kalan 24 madde için tekrar Kaiser Mayer Olkin (KMO) katsayısı ve Barlett testi yapılmış, KMO katsayısı .829 çıkmıştır. 24 maddeye tekrar faktör analizi uygulanmış ve bu maddelerin ilk faktörde yüksek yük değeri aldıkları görülmüştür (%30,909). Toplam 24 maddeden oluşan 'Arapça Konuşma Becerisi Özyeterlik Algısı Ölçeği' 424 öğrenciye uygulanmıştır. Ancak doğru ve eksiksiz olarak doldurulmamış olduğu tespit edilen 28 adet form iptal edilmiş ve toplam 396 öğrencinin verileri kullanılmıştır. Bilgisayar ortamına aktarılan veriler SPSS 15,0 istatistik programı vasıtasıyla analiz edilmiştir. Araştırmanın sonucunda, katılımcıların Arapça konuşma becerileri özyeterlik algılarının bazı değişkenler açısından incelenmiş ve istatistiksel bilgiler sunulmuştur. Buna göre kız öğrenciler ile erkek öğrenciler arasında, Arapça konuşma becerisi özyeterlik algıları açısından istatistiksel olarak anlamlı bir fark bulunmazken, katılımcıların Arapça konuşma becerisine ilişkin özyeterlik algıları; yaş ve sınıf düzeyi değişkenlerine göre istatistiksel olarak anlamlı derecede farklılık gösterdiği sonucuna ulaşılmıştır.

This study aims to develop a Self-Efficacy Scale for Arabic Speaking Skills. The scale was tested for validity and reliability with overall 99 participants majoring in the Department of Arabic Language Teaching at Gazi Faculty of Education at Gazi University and in the Department of Arabic Translation and Interpreting at the Faculty of Letters at Kırıkkale University in the spring semester of 2014-2015. Adjusted item-total values correlations were calculated for the scale items in order to see whether the scale items related to the perception of self-efficacy for Arabic speaking. Kaiser-Mayer-Olkin (KMO) coefficient and Bartlett’s test were applied; and KMO coefficient was found to be .808. Obtained data was concluded to be suitable for factor analysis; for that reason, factor analysis was carried out in order to collect evidence on the construct validity of scale after excluding the items with an adjusted item-total values correlation values lower than .40 from the scale. The variance rate explained by the first factor was found to be 27.591% as a result of the factor analysis carried out. Factor reduction was adopted by ruling that items should have a minimum factor loading of 0.40 in order to be included in the scale and didn’t take any high loading value other than the first factor. The reason why those items were excluded from the scale was to prevent decrease in the construct validity. The remaining 24 items were applied Kaiser-Mayer-Olkin (KMO) coefficient and Bartlett’s test again; and KMO coefficient was found to be .829. The 24 items were applied factor analysis again; and those items were observed to take high loading value in the first factor (30.909%). Consisting of 24 items in total, ‘Scale for Perception of SelfEfficacy for Arabic Speaking Skills’ was carried out on 424 students. However, 28 forms were eliminated, being detected to be filled inaccurately and incompletely; and the data from 396 students in total was used. Computerized data was analyzed using SPSS 15.0 statistical software. Conclusion part of the study investigates perceptions of selfefficacy for Arabic speaking skills of participants with regards to certain variables; and presents statistical information. Accordingly, we came to the conclusion that no statistically significant difference was found with regards to perceptions of self-efficacy for Arabic speaking skills between female and male students while perceptions of self-efficacy regarding Arabic speaking skills for the participants statistically varied to a significant degree in respect to the variables of age and class level. STRUCTURED ABSTRACT Aim of the Research Study The basic principle of foreign language teaching is to teach for communication. Therefore, speaking skill is the most essential one among the language skills. Bygate (1987, p. 4) states that foreign language learners need to improve their speaking skill because people mostly use speaking skill in order to interact with each other from the beginning of the humanity. Self-efficacy is the basic concept of Bandura's social learning theory. Bandura (1977, p. 193) describes self-efficacy as ‘‘the conviction that one can successfully execute the behavior required to produce the outcomes’. When international literature is examined, it is observed that there are many research studies on self-efficacy in foreign language education (Davoudi, M. & Chavosh M, 2016; Redmond, 2015; Raoofi, Tan, & Chan, 2012; Stanger, Jhangiani and Tarry, 2011; Zulkosky, 2009). In view of national (Turkish) setting, it is observed that research studies on language teaching and self-efficacy were generally conducted in the field of English Language Teaching. (Genç, Kuluşaklı, and Aydın 2016; Kaşık, 2014; Ülkümen, 2013; Hancı Yanar and Bümen, 2012; Açıkel, 2011; Semiz, 2011). As regards the research studies in the field of Arabic Language Teaching, it may be concluded that there is no study on the concept of self-efficacy. In this context, the main purpose of this research study is to develop a valid and reliable instrument for measuring the students’ selfefficacy perceptions of Arabic speaking skill. The sub-questions of this study aim to determine whether there is a significant difference in the levels of the students’ self-efficacy perception of Arabic speaking skill in in the Departments of Arabic Language Teaching, Arabic Translation and Interpreting, Arabic language and literature according to variables of gender, age, and class level. Research Methodology In the research, descriptive research design is used to determine the students’ self-efficacy perceptions of speaking skill in the Arabic Language Teaching, Arabic Language and Literature and Arabic Translation and Interpreting Çepni (2009, p. 34) explains that descriptive research studies are used "to enlighten a given situation in general, to carry out evaluation in line with the standards, and to shed light on the interrelationships between the events". Development of Measurement Instrument The pilot scale was tested for validity and reliability with overall 99 participants majoring in the Department of Arabic Language Teaching at Gazi Faculty of Education at Gazi University and in the Department of Arabic Translation and Interpreting at the Faculty of Letters at Kırıkkale University in the spring semester of 2014-2015. Adjusted item-total values correlations were calculated for the scale items in order to see whether the scale items related to the perception of self-efficacy for Arabic speaking. Kaiser-Mayer-Olkin (KMO) coefficient and Bartlett’s test were applied; and KMO coefficient was found to be .808. Obtained data was concluded to be suitable for factor analysis; for that reason, factor analysis was carried out in order to collect evidence on the construct validity of scale after excluding the items with an adjusted item-total values correlation values lower than .40 from the scale. The variance rate explained by the first factor was found to be 27.591% as a result of the factor analysis carried out. Factor reduction was adopted by ruling that items should have a minimum factor loading of 0.40 in order to be included in the scale and didn’t take any high loading value other than the first factor. The reason why those items were excluded from the scale was to prevent decrease in the construct validity. The remaining 24 items were applied Kaiser-Mayer-Olkin (KMO) coefficient and Bartlett’s test again; and KMO coefficient was found to be .829. The 24 items were applied factor analysis again; and those items were observed to take high loading value in the first factor (30.909%). The Cronbach alpha reliability coefficient was measured at 899. Findings The first sub-question of the study is formulated as "Do selfefficacy perceptions of Arabic speaking skill differentiate significantly in view of gender?" It can be concluded that there is no significant difference in the self-efficacy perceptions of Arabic speaking skill between male students and female students [t (394) = -1,265 p> 0.01]. When the analysis results are examined, it is observed that the female and male students’ self-efficacy perceptions of Arabic speaking skills are very close. The second sub-question of the research study is formulated as "Do self-efficacy perceptions of Arabic speaking skill differentiate significantly in view of age?" There is a statistically significant difference between ‘18-19 and 22-23 age groups’ and ‘20-21 and 22-23 age groups’ in terms of selfefficacy perceptions of Arabic speaking skill [p < .05)]. There is no statistically significant difference between 18-19 and 20-21; 18-19 and 24-25; 18-19 and 25+; 20-21 and 24-25; 20-21 and 25+; 22-23 and 24-25; 22-23 and 25+; 24-25 / 25+ age groups in terms of self-efficacy perceptions of Arabic speaking skill [p> .05]. The third sub-question of the research study is formulated as "Do self-efficacy perceptions of Arabic speaking skill differentiate significantly in view of class level?" There is a statistically significant difference in the self-efficacy perceptions of Arabic speaking skill between the 2nd and 3rd grades and between the 2nd and 4th grades [p < .05])Conclusion and Discussion In this study, a valid and reliable scale, which could be used in determining the students’ self-efficacy perceptions of Arabic speaking skills in the Departments of Arabic Language Teaching, Arabic Translation and Interpreting, Arabic language and literature, was developed in view of the content and construct. As a result, the internal consistency and reliability of Scale of SelfEfficacy Perceptions of Arabic Speaking Skill are quite high. There is no statistically significant difference between male and female participants in terms of self-efficacy perceptions of Arabic speaking skills. This finding is consistent with the research studies carried out by Kıvılcım (2014), Bümen and Özaydın (2013) and Aykaç Duman (2007). On the other hand, there is a significant difference in the self-efficacy perceptions of the participants according to the variable of gender in the research studies conducted by Çalışkan, Selçuk, and Özcan (2010) and Akpınar-Dellal and Akın (2016). According to the variable of age, there is a statistically significant difference among the students in terms of self-efficacy perceptions of Arabic speaking skills. There is no difference in some studies in view of the variables of age and self-efficacy (Kıvılcım, 2014). There is a statistically significant difference in self-efficacy perceptions of Arabic speaking ability of students according to class levels. In accordance with the research data, there is a gradual increase in self-efficacy perceptions in line with the class level. That is to say, self-efficacy is promoted as the level of learning increases. The research studies conducted by Aykaç Duman (2007) and Çalışkan, Selçuk and Özcan (2010) show the same findings related to the level of class.