Automated Evaluation to Explore Writing Skills

Abstract

Writing is a complex skill to develop; therefore, the integration of technological software should be considered to obtain its maximum benefits. The purpose of this research is to analyze the impact of automated evaluation in terms of developing students’ writing performance. For instance, this investigative work was based on a quantitative approach and pre- experimental design. Hence, 36 undergraduate students from the third semester belonging to the major Pedagogía de los Idiomas Nacionales y Extranjeros were encouraged to use Virtual Writing Tutor during six interventions that were destined to teach three genres of email, article, and story and generate one composition per day to obtain feedback from the AWE system. Additionally, a pre-test and post-test based on the B1 Preliminary Writing Section were implemented to corroborate students’ improvement, they comprised 3 parts email, article, and story writing and were graded according to the criteria provided by Cambridge content, communicative, achievement, organization and language which were scored over five individually. To examine the results of the pre and post-test statistics were implemented; thus, the Wilcoxon rank test was implemented to accept or reject the alternative hypothesis. The results revealed that automated evaluation positively impacts students’ writing skills since students have a significant improvement in each aspect evaluated. Specially, students exhibited a notable improvement in email writing where most of them got a perfect grade. Overall, students showed a great improvement since there was a difference according to the pre and post-test of each genre. For instance, regarding email writing there was a difference of 5, 12 points, article writing presented a difference of 3,66, and story writing exhibited a difference of 3.66 points.