Rubric proposal for evaluation of speaking in teaching Turkish as a foreign language: A1-A2 levels

Assessment and evaluation are one of the important stages in the foreign language teaching process. It appears at all stages from the beginning of the language learning process. Learners demonstrate oral performance in the exams for speaking skills, which is a productive skill. In such exams, rubrics can be used to measure the learner's performance and facilitate evaluation. Rating scales are generally created in two ways: holistic and analytical. While holistic rating scales evaluate people's performance as a whole, analytical rating scales enable people's performances to be measured in more detail according to determined dimensions and criteria. It is important to carry out the foreign language learning process by observing a certain standard. It is important to use rating scales that will make it easier for evaluators to act objectively and create an inter-institutional standard. In the field of Turkish as a foreign language, studies on beginner levels A1 and A2 are limited. Therefore, this study is an important step in the preparation and use of valid and reliable assessment tools at A1 and A2 levels. The aim of this study is to create analytical rubrics for measuring and evaluating speaking skills at A1 and A2 levels. Mixed method was used in the study. For this reason, two analytical rubrics were created at A1 and A2 levels. In order to create rubrics, a literature review was first conducted. In addition, the rating scales currently used in language centers and international foreign language exams were examined and as a result of the data obtained, the dimensions and the criteria under the dimensions were determined. Then, expert opinion was sought and content validity studies were conducted and scoring keys were prepared by taking the opinions of field experts with at least ten years of experience. During the application phase, twenty evaluators teaching Turkish in language centers evaluated the speaking performances of six A1 and six A2 level international students with prepared analytical rating scales and the results were analyzed with SPSS. For reliability studies, inter-rater agreement was checked and Kendall's Coefficient of Concordance was calculated. According to the results, high agreement was observed between the evaluators in both A1 and A2 level rating scales. It is concluded that the evaluators' experiences did not have a statistical effect on their ratings. As a result of the study, valid and reliable rating scales for A1 and A2 levels were obtained and are thought to be ready to be used in the relevant field.