The Detected and Undetected Errors in Automated Writing Evaluation Program’s Result

Abstract

The investigation of error is quite challenging to be conducted at school and university. Interestingly, with the technology development, detecting error can be conducted by using an automated writing evaluation program. This study aimed to analyze the errors in writing by applying an automated writing evaluation program. This study applied a mixed methods research with exploratory design. The total of 48 undergraduate students participated in this study and each student submitted one essay which was then re-submitted to the automated writing evaluation program. As the result, there are 484 errors detected by the program which came from 23 types of errors. However, the program still left some errors undetected with the total number of 158 errors which came from 24 types of errors; 13 types of errors have been identified and 11 types of errors have not been identified by the program. From the result of the program, the use of automated writing evaluation program in detecting error seems giving some benefits for the user. However, the application of this program still needs the teacher and lecturer’s supervision to reduce the weaknesses of the program in detecting the errors.