ABSTRACT

Vocabulary tests administered on multilingual populations should take into account the unique linguistic and cultural makeup of the population by adopting test development methods that allow responses in several languages. Our aims were to develop a picture-naming test for multilingual Lebanese school-age children (L1: Lebanese, L2: French and/or English) using the dual-focus approach for test development in order to obtain test items in several languages simultaneously and examine sources of variability in performance across gender, age, and type of schooling. The test compiled pictures with high cultural familiarity, high name agreement, an age of acquisition between 3 and 9 years, and varying frequency of use based on the ratings of 8 indigenous Lebanese experts on the test content. The test was administered to 74 multilingual Lebanese children aged 3 to 9 years. Results showed that older age, studying at a private school and being female contributed to better test performance, although gender was not a significant variable when inserted in a regression model. Psychometric properties of the test revealed excellent levels of internal reliability, good item parameters and sources of validity. This study delineates a method of test construction that accommodates to multilingual individuals in a resource-limited setting.

摘要

对多语言人群进行的词汇测试应考虑到该人群独特的语言和文化构成，采用允许以多种语言进行回答的测试开发方法。我们的目标是使用双焦点测试开发方法，为多语言黎巴嫩学龄儿童（L1：黎巴嫩语，L2：法语和/或英语）开发图片命名测试，以便同时获得多种语言的测试项目，并且检查不同性别、年龄和学校教育类型之间表现差异的来源。该测试根据8位黎巴嫩原住民专家对测试内容的评分，编制了文化熟悉度高、名称一致度高、获取年龄在3至9岁之间、使用频率不同的图片。该测试针对 74 名 3 至 9 岁的多语言黎巴嫩儿童进行。结果显示，年龄较大、在私立学校学习和女性身份有助于提高考试成绩，尽管在回归模型中性别并不是一个显着变量。该测试的心理测量特性显示出出色的内部可靠性水平、良好的项目参数和有效性来源。这项研究描述了一种测试构建方法，可以在资源有限的环境中适应多语言个体。良好的项目参数和有效性来源。这项研究描述了一种测试构建方法，可以在资源有限的环境中适应多语言个体。良好的项目参数和有效性来源。这项研究描述了一种测试构建方法，可以在资源有限的环境中适应多语言个体。