In this study, gain score, and categorical growth models were used to examine the role of student (gender and socioeconomic level) and school characteristics (school size and school resources) in the student growth on comprehension skills in language. The participants of this study were 2,416 sixth-grade students in 2011 who became seventh-grade students in 2012. The data was collected through two achievement tests, student and school questionnaires. Two achievement tests were calibrated using the Rasch Model and were scaled using the concurrent estimation method. Moreover, the cut-off scores of these tests were determined by using the bookmark method. Students’ growth was modelled with the gain score and categorical growth models. All data was analyzed using multilevel models. Results showed that some students did not achieve sufficient gains to advance to higher performance levels. Although some schools’ average gains were higher, their performance was still not significant enough in terms of tests’ standards. Moreover, the analyses demonstrated that the student gain scores and growth categories varied significantly among the schools. In addition, the study was able to determine student and school characteristics that have an impact on the students' gain scores and categorical growth. Given the different aspects gained about students’ performance with these models, it is recommended to utilize different growth models in schools.

中文翻译：

---

使用收益得分和分类增长模型检验学生的增长

在这项研究中，使用增益分数和分类增长模型来检查学生对语言理解能力的作用（性别和社会经济水平）和学校特征（学校规模和学校资源）在学生生长中的作用。这项研究的参与者为2011年的2,416名六年级学生，并在2012年成为了七年级学生。数据是通过两个成绩测试（学生和学校问卷）收集的。使用Rasch模型对两个成就测试进行了校准，并使用并发估计方法对其进行了缩放。此外，使用书签方法确定这些测试的分数。学生的成长是通过获得分数和分类成长模型来建模的。使用多级模型分析所有数据。结果表明，有些学生没有获得足够的收获，无法晋升到更高的成绩水平。尽管一些学校的平均成绩更高，但就考试标准而言，他们的表现仍然不够显着。此外，分析表明，各学校的学生获得分​​数和成长类别差异很大。此外，该研究还可以确定对学生的学习成绩和分类成长有影响的学生和学校特征。考虑到使用这些模型获得的学生表现的不同方面，建议在学校中使用不同的增长模型。分析表明，各学校的学生学习成绩得分和成长类别差异很大。此外，该研究能够确定对学生的学习成绩和分类成长有影响的学生和学校特征。考虑到使用这些模型获得的学生表现的不同方面，建议在学校中使用不同的增长模型。分析表明，各学校的学生学习成绩得分和成长类别差异很大。此外，该研究还可以确定对学生的学习成绩和分类成长有影响的学生和学校特征。考虑到使用这些模型获得的学生表现的不同方面，建议在学校中使用不同的增长模型。