【作者】 罗杰； 陈维； 曾练平； 潘运； 赵守盈；

【Author】 Luo Jie;Chen Wei;Zeng Lianping;Pan Yun;Zhao Shouying;School of Educational Science,Guizhou Normal University;

【机构】 贵州师范大学教育科学学院；

【摘要】 潜在类别分析(Latent Class Analysis,LCA)是通过对类别型的外显变量和潜在变量之间的关系建立统计模型,根据模型参数得到各种潜在类别的具体外在表现的潜在特征分类技术。其基本假设为对各外显变量各种反应的概率分布可以由少数互斥的潜在类别变量来解释,每种类别对各外显变量的反应选择都有特定的倾向。LCA的最基本应用就是根据个体在外显测查题目上的作答模式来判断个体的潜在特征分类,并了解各种类别在整个群体中的人数比例。这种基于概率模型的分类方法不但能够保证划分出来的各类之间差异最大,类别内部差异最小,而且还能利用客观的统计指标去衡量分类的准确性与有效性。本文拟采用潜在类别分析以考察大学新生学习适应状况的潜类别结构或群体异质性,并对不同学习适应类型在性别和专业内的分布进行差异检验。采取方便取样,采用大学新生学习适应问卷对4146名大学新生进行调查。结果表明:大学新生学习适应分成4个潜在类别,分别是高适应型(35.8%)、混合型伴高效能(11.9%)、混合型伴高动机(28.4%)和低适应型(23.8%);进一步的分析发现,男生在高适应组(χ~2=5.696,p=0.017)、高效能组(χ~2=43.832,p<0.001)、高目标组(χ~2=12.832,p<0.001)和低适应组(χ~2=38.883,p<0.001)的比例显著高于女生;文史类学生在高目标组(χ~2=8.313,p=0.004)的比例显著高于理工类学生。更多还原

【Abstract】 Latent class analysis(LCA) is a statistical technique used to identify subtypes of related cases. In the method, statistic model is built according to the relationship between categorical observed and latent variables, and then manifest performances of each latent class are respectively described based on the related parameters in this model. The basic hypothesis of LCA was that, the reaction probability distribution of manifest variables can be explained by some incompatible latent variables, and each latent variable has certain tendency to the selection of manifest variables. The most application of LCA was to classify individuals into distinct groups or categories based on individual response patterns so that individuals within a group more similar than individuals between groups. By the way, LCA can provide more objective statistical indicators to evaluate the accuracy and effectiveness of classification. In order to investigate the latent structures or groups heterogeneity of freshmen learning adjustment, latent class analysis was used to explore types of freshmen learning adjustment, and percentage difference test was employed to investigate the difference of frequency in different age and specialty. A total of 4146 freshmen were selected by convenience sampling, they were surveyed with the freshmen learning adjustment inventory. The results showed that: According to LCA, the learning adjustment in freshmen were divided into four groups, namely, high-adjustment, mixed-adjustment with high efficacy, mixed-adjustment with high motivation, and low-adjustment, their proportions of all the samples were 35.8%, 11.9%, 28.4%, 23.8%, respectively. Further analysis indicated that, the prevalence of high-adjustment(χ~2=5.696, p=0.017), high-efficacy(χ~2=43.832, p<0.001), high-motivation(χ~2=12.832, p<0.001) and low-adjustment(χ~2=38.883, p<0.001) of boys was significantly higher than girls; the prevalence of high-motivation of learning liberal arts was significantly higher than learning science(χ~2=8.313, p=0.004).更多还原