SUBJECTIVITY IN THE EVALUATION OF FOREIGN LANGUAGE WRITING

Abstract

Evaluation is one of the most important components of the educational process. Despite the tendency to use various kinds of tests in teaching practice, essays and other written works in a foreign language are still being checked manually by teachers in the vast majority of cases. The issue of subjectivity in such an evaluation is solved by defining explicit assessment criteria, however the author questions the objectivity even of an evaluation based on comprehensive benchmarks. The article presents results of a study determining the subjectivity degree of the expert writing evaluation in German. A questionnaire survey of school and university teachers from different cities in Russia and Germany reveals that even in cases where an incorrect choice of lexeme strongly distorts the meaning of the sentence, expert evaluations differ as much as possible in determining the severity degree of the error. Based on the data analysis a conclusion about the unreliability of manual expert evaluations is made and the idea of partial automation of writing evaluations based on developing artificial intelligence technology is put forward.