The complex process of foreign language learning cannot beadequately explained solely by cognitive or non-affective factors. It isobvious that in a holistic understanding of language learning process,affective variables need to be examined so that learners' diverse needsand interests can be better understood and addressed.Since the 1970s, research interest in foreign language learninghas shifted from teachers to learners and increasing numbers of studieshave been undertaken from the learners' perspectives. These studieshave focused on learners’ affective variables such as self-esteem, risktaking, inhibition, empathy, tolerance of ambiguity, beliefs, attitudes,motivation and especially anxiety. Anxiety is one of the key factors thathas considerable influence on language learning. Spielberger (1966),defines anxiety as subjective, consciously perceived feelings ofapprehension and tension, accompanied by and associated withactivation or arousal of the autonomic nervous system. Test anxietyrefers to phenomenological, physiological, and behavioral responsesthat accompany concern possible negative consequences and failure onan exam or similar evaluative situation (Zeidner, 1998). According toHorwitz, Horwitz and Cope (1986), foreign language anxiety is separatecomplicated phenomenon of self-perception, beliefs, feelings andbehavior related to classroom language learning process. Horwitz et al.(1986) defines three components of Foreign Language Anxiety (FLA).They associate FLA with the fear of negative evaluation, test anxiety andcommunication apprehension. Fear of negative evaluation consists ofthe feelings of apprehension related to other people’s evaluations,refraining from the situations that need evaluation and false perceptionthat they will always be evaluated in a negative way. Foreign languagelearners often experience this fear as they are evaluated both by theirteachers and by their peers. Test anxiety arises from a fear of failureand includes worry because of tested. Communication apprehension isa type of shyness characterized by fear or anxiety sbout communicatingwith people. It derives from the personal knowledge that one will almostcertainly have difficuly understanding others.Liebert and Morris (1967) described two major components of testanxiety: a cognitive component they labeled "worry," that referred toconcerns about being evaluated and about the consequences of failure,and an affective component they labeled "emotionality," that referred tothe perception of autonomic reactions evoked by the test situation.It can be easily observed that there is an ever-growing body ofresearch on Foreign Language Anxiety, research in test anxiety levels intwo subscales receives much less interest. In order to fill this gap, thisstudy aims to determine test anxiety levels of university preparatorystudents learning English as a foreign language. Another purpose of thestudy is to find out whether there is a significant relationship betweenlearners’ gender, age, the department which they will study, compulsoryor voluntary status of the students and two subscales of test anxiety.Yabancı Dil Öğrenenlerin Sınav Kaygı Düzeyleri 245Turkish StudiesInternational Periodical For the Languages, Literature and History of Turkish or TurkicVolume 10/7 Spring 2015Method: The participants of the study were 301 (211 female and90 male) students aged between 17-22 years old attending a one-yearEFL preparatory school at Ondokuz Mayıs University, School of ForeignLanguages. Students, who will study at Foreign Language Departmentand Faculty of Medicine, are obliged to enroll in the preparatory school(n= 187). The students of other departments attend voluntarily (n= 114).The TAI (Test Anxiety Inventory) developed by Spielberger (1980) wasused to collect the data. The TAI is a self-report scale that was designedto measure individuals’ test anxiety as a situation-specific trait, itconsists of 20 items which are to be responded by using a four-pointrating-scale format. In addition to a total score, separate scores forworry (8 items) and emotionality (12 items) can be obtained. Turkishversion of TAI adapted by Öner (1990) was used in the present study.Cronbach alpha of worry subscale was to be 0.89, for emotionalitysubscale was found to be 0.93 by the researcher. The analysis of thedata was carried out using SPSS (Version 17.0). Descriptive statisticssuch as means, standard deviations, an independent-t test and ANOVAwere used.Results: This study aims to determine worry and emotionalityscores in test anxiety of university students learning English as aforeign language and also investigate the effects of variables such asgender, age, compulsory or voluntary status of the students anddepartments they will study on the students’ worry and emotionalitysubscales. The results showed that the participants in the currentstudy exhibited relatively low level anxiety in both worry andemotionality subscales of test anxiety. The second research questionexamined if there is a significant relationship among gender and testanxiety subscales. The results indicated that female students’ meanscores were higher than those of male students in both worry andemotionality subscales. Females reported higher levels of test anxietythan males. This suggests that females are significantly more anxiousthan males in two subscales of test anxiety. According to results, agewasnot significant statistically. Another research question investigatedwhether there is a relationship between compulsory and voluntarystatus of the students and subscales of test anxiety. The results showedthat there was no meaningful relationship between compulsory andvoluntary status of the students and emotionality subscale of testanxiety. On the other hand, it was clearly evident that compulsorypreparatory st