Gender‐Based EFL Writing Error Analysis Using Human and Computer‐Aided Approaches

Abstract

Committing errors is expected in the development of language acquisition and learning; however, there is limited research that contributes to the literature on the effect of gender of English as a foreign language (EFL) writing. This study explored the gender differences in EFL students' writing using two approaches: human evaluation and computer-aided error analysis (CEA). A corpus of 90,753 words was compiled from written samples collected from 197 participants (98 males and 99 females [freshmen or sophomores]) who were studying on a full-time basis in a private institution in the Kingdom of Saudi Arabia (KSA). By utilizing descriptive analysis and factorial two-way ANOVA to detect counting errors, the findings reveal that the distinctions between male and female writings are nonsignificantly related to specific errors. Generally, the results exhibit noteworthy differences between the error types of the two male and female groups (more error frequency in the male-written productions). The findings support educators and decision-makers on implementing more learner-conscious teaching approaches, enhancing male and female students' EFL writing skills. This study further suggests diagnostic information about challenging EFL writing areas, which can benefit educators and course designers in establishing effective teaching strategies and feedback approaches that should be implemented to promote EFL writing skills.

An Analysis of Turkish EFL Students’ Writing Errors

“…Likewise, in a study done by Pouladian et al (2017) to examine writing errors made by Iranian EFL learners, it was asserted that the male students made more speaking and writing errors than the female ones. Contrarily, in another study (Almusharraf & Alotaibi, 2021), no statistically significant difference was found between the total number of writing errors detected for the male and female students.…”

“…Some studies investigating gender differences on writing errors have been carried out, but they have pointed out different results. For instance, Almusharraf and Alotaibi (2021) found no statistically significant difference between the total number of writing errors detected for the male and female students. Boroomand and Rostami Abusaeedi (2013); however, reported that female EFL students made more writing errors than the male students did.…”

The Analysis of Gramatical and Lexical Errors on the Students' Theses of Iain Takengon

“…The primary benefit of Error Analysis is that it successfully transforms an error from a negative to a positive norm (Almusharraf & Alotaibi, 2021). Errors are now seen as students' constructive contributions to their second language learning rather than as "undesirable forms" (Kearney & Ellis, 1995).…”

Machine Learning Literacy for Measurement Professionals: A Practical Tutorial

“…Pretrained NLP models can also be applied to the texts to create theoretically meaningful features. For example, in automated essay scoring, theoretical meaningful features such as the number of grammatical errors (Almusharraf & Alotaibi, 2020), text complexity (Sheehan, 2017), and text cohesion metrics (McNamara et al., 2010) can be first extracted using various pretrained/theoretically defined NLP models. The features can later be used in more specific ML models to score essays.…”