{"title":"二语搭配知识评价的改进","authors":"Senyung Lee, Sun-young Shin","doi":"10.1080/15434303.2021.1908295","DOIUrl":null,"url":null,"abstract":"ABSTRACT Multiple test tasks are available for assessing L2 collocation knowledge. However, few studies have investigated the characteristics of a variety of recognition and recall tasks of collocation simultaneously, and most research on L2 collocations has focused on verb-noun and adjective-noun collocations. This study investigates (1) the relative informativeness of different tasks for assessing L2 collocation knowledge and (2) the effect of collocation type on learners’ scores on collocation tasks. Four tasks were developed based on an extensive review of research on L2 collocations: a sentence writing task, fill-in-the-blank task, multiple-choice task, and Yes/No acceptability judgment task. Each task targeted 64 English collocations, including verb-noun, adjective-noun, adverb-adjective, and adverb-verb collocations. Four groups of adult ESL learners representing different levels of academic English literacy (n = 205) completed the tasks. An item response theory analysis showed that the sentence writing and fill-in-the-blank-tasks had similar difficulty and discriminating power, the eight-option multiple-choice task had the highest discriminating power, and the Yes/No judgment task had the lowest difficulty and discriminating power. The type of collocation did not have a significant effect on learners’ scores when collocation frequency was held constant, regardless of task and learners’ level of academic English literacy.","PeriodicalId":46873,"journal":{"name":"Language Assessment Quarterly","volume":null,"pages":null},"PeriodicalIF":1.4000,"publicationDate":"2021-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/15434303.2021.1908295","citationCount":"3","resultStr":"{\"title\":\"Towards Improved Assessment of L2 Collocation Knowledge\",\"authors\":\"Senyung Lee, Sun-young Shin\",\"doi\":\"10.1080/15434303.2021.1908295\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"ABSTRACT Multiple test tasks are available for assessing L2 collocation knowledge. However, few studies have investigated the characteristics of a variety of recognition and recall tasks of collocation simultaneously, and most research on L2 collocations has focused on verb-noun and adjective-noun collocations. This study investigates (1) the relative informativeness of different tasks for assessing L2 collocation knowledge and (2) the effect of collocation type on learners’ scores on collocation tasks. Four tasks were developed based on an extensive review of research on L2 collocations: a sentence writing task, fill-in-the-blank task, multiple-choice task, and Yes/No acceptability judgment task. Each task targeted 64 English collocations, including verb-noun, adjective-noun, adverb-adjective, and adverb-verb collocations. Four groups of adult ESL learners representing different levels of academic English literacy (n = 205) completed the tasks. An item response theory analysis showed that the sentence writing and fill-in-the-blank-tasks had similar difficulty and discriminating power, the eight-option multiple-choice task had the highest discriminating power, and the Yes/No judgment task had the lowest difficulty and discriminating power. The type of collocation did not have a significant effect on learners’ scores when collocation frequency was held constant, regardless of task and learners’ level of academic English literacy.\",\"PeriodicalId\":46873,\"journal\":{\"name\":\"Language Assessment Quarterly\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":1.4000,\"publicationDate\":\"2021-04-25\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://sci-hub-pdf.com/10.1080/15434303.2021.1908295\",\"citationCount\":\"3\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Language Assessment Quarterly\",\"FirstCategoryId\":\"98\",\"ListUrlMain\":\"https://doi.org/10.1080/15434303.2021.1908295\",\"RegionNum\":2,\"RegionCategory\":\"文学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"0\",\"JCRName\":\"LANGUAGE & LINGUISTICS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Language Assessment Quarterly","FirstCategoryId":"98","ListUrlMain":"https://doi.org/10.1080/15434303.2021.1908295","RegionNum":2,"RegionCategory":"文学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"0","JCRName":"LANGUAGE & LINGUISTICS","Score":null,"Total":0}
Towards Improved Assessment of L2 Collocation Knowledge
ABSTRACT Multiple test tasks are available for assessing L2 collocation knowledge. However, few studies have investigated the characteristics of a variety of recognition and recall tasks of collocation simultaneously, and most research on L2 collocations has focused on verb-noun and adjective-noun collocations. This study investigates (1) the relative informativeness of different tasks for assessing L2 collocation knowledge and (2) the effect of collocation type on learners’ scores on collocation tasks. Four tasks were developed based on an extensive review of research on L2 collocations: a sentence writing task, fill-in-the-blank task, multiple-choice task, and Yes/No acceptability judgment task. Each task targeted 64 English collocations, including verb-noun, adjective-noun, adverb-adjective, and adverb-verb collocations. Four groups of adult ESL learners representing different levels of academic English literacy (n = 205) completed the tasks. An item response theory analysis showed that the sentence writing and fill-in-the-blank-tasks had similar difficulty and discriminating power, the eight-option multiple-choice task had the highest discriminating power, and the Yes/No judgment task had the lowest difficulty and discriminating power. The type of collocation did not have a significant effect on learners’ scores when collocation frequency was held constant, regardless of task and learners’ level of academic English literacy.