{"title":"Using grammar checkers in the ESL classroom: the adequacy of automatic corrective feedback","authors":"P. John, Nina Woll","doi":"10.14705/rpnet.2018.26.823","DOIUrl":null,"url":null,"abstract":"Our study assessed the performance of two Grammar Checkers (GCs), Grammarly and Virtual Writing Tutor, and the grammar checking function in Microsoft Word on a broad range of grammatical errors. The errors occurred in both authentic English as a Second Language (ESL) compositions and simple sentences we generated ourselves. We verified the performance in terms of (1) coverage (rates of error detection), (2) accuracy of proposed replacement forms, and (3) ‘false alarms’ (forms mistakenly flagged as incorrect). To the extent GCs provide accurate and comprehensive corrective feedback, they could relieve teachers of the time-consuming task of providing written feedback themselves. While inaccurate replacement forms and false alarms are relatively rare, we found GCs to have poor overall coverage (total error detection rates under 50%). Grammarly and Virtual Writing Tutor, however, outperform Microsoft Word. Coverage is also higher both for certain categories of error and for the sentences rather than the authentic compositions. Finally, although GCs do not provide comprehensive feedback, we suggest designing special activities that target select error types.","PeriodicalId":138095,"journal":{"name":"Future-proof CALL: language learning as exploration and encounters – short papers from EUROCALL 2018","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Future-proof CALL: language learning as exploration and encounters – short papers from EUROCALL 2018","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.14705/rpnet.2018.26.823","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
Our study assessed the performance of two Grammar Checkers (GCs), Grammarly and Virtual Writing Tutor, and the grammar checking function in Microsoft Word on a broad range of grammatical errors. The errors occurred in both authentic English as a Second Language (ESL) compositions and simple sentences we generated ourselves. We verified the performance in terms of (1) coverage (rates of error detection), (2) accuracy of proposed replacement forms, and (3) ‘false alarms’ (forms mistakenly flagged as incorrect). To the extent GCs provide accurate and comprehensive corrective feedback, they could relieve teachers of the time-consuming task of providing written feedback themselves. While inaccurate replacement forms and false alarms are relatively rare, we found GCs to have poor overall coverage (total error detection rates under 50%). Grammarly and Virtual Writing Tutor, however, outperform Microsoft Word. Coverage is also higher both for certain categories of error and for the sentences rather than the authentic compositions. Finally, although GCs do not provide comprehensive feedback, we suggest designing special activities that target select error types.