Issues in the Evaluation and the Measurement of Learner Language

• Kondo Yusuke

Abstract

To construct learner corpus is a project that brings together different fields of study and knowledge. Although second language acquisition, language testing, automated scoring, and task-based language teaching/learning benefit from learner corpus research, learner corpus research does not fully refer to the research results in its adjacent fields. The purpose of this study is to identify common issues in automatic scoring of second language (L2) and learner corpus research and to point out the need for knowledge from their adjacent disciplines to build a more useful learner corpus. The issues presented are found in the evaluation and the measurement of learner language. After reviewing the basic concept of automated scoring, issues of predictor variables and human rating are presented in automated scoring, and then, issues of features and data collection time period are mentioned in a longitudinal learner corpus research. One of the ways to solve the issues presented in this paper is the collaboration among researchers in automated scoring and in learner corpus, expertise in language testing and in task design. This collaboration gives useful insights to the researchers and the expertise in these fields

Journal

• Learner Corpus Studies in Asia and the World

  Learner Corpus Studies in Asia and the World 5 95-104, 2020-12-21

  神戸大学国際コミュニケーションセンター

Keywords

• Automated scoring
• Learner corpus research
• Human rating
• Linguistic features

Details

• CRID: 1390009224956603904
• NII Article ID: 120006942793
• DOI: 10.24546/81012492
• ISSN: 24352632
• ISSN: 21876746
• HANDLE: 20.500.14094/81012492
• Text Lang: en
• Data Source:
  • JaLC
  • IRDB
  • CiNii Articles
• Abstract License Flag: Allowed