Medical specialization students are both educated in the clinic and serve in the provision of services. For this reason, it is very difficult for them to go to another center to take their exams. This process brings both time consuming and economic burdens. These requirements have been a starting point for online surveys and assessments (10). Because the students are located in different provinces, the problem of time and space is solved easily with the central online examination. It is ensured that the trainers evaluate this process by the central examination of the students receiving education at different faculties and different geographic regions at the same time. As a matter of fact, it is possible for the students of different educational institutions to evaluate themselves. In addition, it is ensured that research assistants in all educational institutions can see the shortcomings of students and their programs according to their questions. If technology can be used well, more communication, interaction and cooperation can be achieved with online exams. There are studies showing that the use of online evaluation has contributed positively to the learning and teaching process. There are publications reporting that students feel more comfortable in their online surveys and evaluations, their negative attitudes towards learning and measurement-evaluation have disappeared and their academic achievement has increased (14,15).
When the seniority of the students taking the exam is examined, it was observed that the first year research assistants participated in the exam in the most intense way and it was noticed that the number of research assistants who took the exam in the following years decreased according to seniority. This may be due to different reasons. This may be due to the fact that the first year assistants are more willing to test themselves and are in the clinic. Last-year assistants in the period of rotation or thesis preparation may be less motivated.
The online evaluation systems have many positive contributions to the education process. When appropriate infrastructure is provided to the web environment, it is accessible and applicable from anywhere. Students' achievements in exams can be seen instantly in the system (2). Online exams allow the creation of a question bank. The students who prepare the exam keep their knowledge up-to-date and follow the subjects they will prepare for the exam carefully. Exams are done through the network, thus paper and publication expenditures are also eliminated (2,16). It takes a long time for classical examinations to be conducted and evaluated by academics. Thanks to online exams, trainers also save time. In addition, faculty members can look at the response rates of the questions and focus on subjects in that students' knowledge is insufficient, so that these subjects can be better understood. Individualization of the exam, instant feedback, quick scoring, and no limitations in terms of time and place are the advantages of the online exam (2,17). As a matter of fact, EMEK exams were conducted with one hour duration and many research assistants were evaluated online.
An exam is expected to be reliable, valid, useful and contributing to education. The fact that the margin of error in the exam is low means that it is a very reliable test (5). The measurement method should be valid and reliable (18,19). Cronbach's alpha reliability coefficient is mostly used to measure the reliability of a test. The higher the coefficient, the higher the internal consistency. For reliable tests, this coefficient is expected to be 0.80 or higher. It has also been reported that 0.70 or more can be accepted for formative exams (20). When EMEK examinations are considered, it is seen that the best value in reliability is taken in the first exam and the most negative measurement is taken in May 2017 exam. The high margin of error in this period is attributed to both the reduction of motivation in the examiners and the technical disadvantages in the preparation of the questions and in the implementation of the online exam. However, the reliability coefficient of the EMEK research assistant development examinations, which were made for formative purposes, was found to be quite satisfactory (Table 1).
The discrimination index is an index used to separate those who do not have the characteristic to be measured by the test. Questions with high levels of discriminative features increase the reliability of the test. The discriminative feature of a question with a discriminant index of 0.40 or higher is defined as very high. The discriminativity of a question with index 0.30-0.39 is not complete, so it needs to be reviewed. The question with a 0.20-0.29 index should be corrected and developed, and should not be reused without any changes. 0.19 index of a question is low, the question must be removed from the test and not transferred to the question bank (20). In an exam made for the evaluation of research assistants, the expected feature of the questions is to distinguish the student who does not know the subject matter (20). When we look at the most of the EMEK examinations, although the discrimination rate is sufficient in some questions, it is necessary to review many questions (Table 2). In the examinations, the question rate which can distinguish the student who does not know the subject with the relevant subject is found to be 16%. However, the percentage of questions that should not be used after reviewing is 43% in total. 41% of the exam questions should be reviewed and included in the question bank after making the necessary corrections. These data suggest that it is necessary to receive specific training for measurement and evaluation during question preparation.
In the analysis of exams, the investigation and interpretation of why the difficulty of the questions may have been caused is a gain of the evaluation (2,17,20). Table 4 presents the subjects in which the students have difficulty in answering correctly and have lack of knowledge. The remarkable result in this table is that less common or rare infections and basic medical science questions are answered less accurately. This will be the educational gain of the units with particular attention to these topics in the training process.
In the online applications, problems such as the emergence of unexpected technical problems, infrastructure and hardware dependence, security problems, students' use of different computers and especially requirement of having a computer are the limitations of online exams. No matter how good the network is, there may be some problems such as power failure or voltage change, lack of features of the computer being used. In addition, although online exams are strictly assigned to supervisors, cheating is much easier than classical exams. For this reason, the security of the system, (the user codes must not be created on the systems) communication security, difficulties in identification of the user and cheating affect the reliability of the online examination negatively (2,17).
It is the limitation of the EMEK examinations that the number of questions in categorized subject titles are not determined by taking into consideration the learning objectives of the IDCM.