Wu Wei, West Stephen G, Hughes Jan N
Department of Psychology, Arizona State University, Tempe, AZ 85287-1104, USA.
J Sch Psychol. 2008 Feb;46(1):85-105. doi: 10.1016/j.jsp.2007.01.003. Epub 2007 Mar 23.
We investigated the effects of grade retention in first grade on the growth of the Woodcock-Johnson broad mathematics and reading scores over three years using linear growth curve modeling on an academically at-risk sample. A large sample (n=784) of first grade children who were at risk for retention was initially identified based on low literacy scores. Scores representing propensity for retention were constructed based on 72 variables collected in comprehensive baseline testing in first grade. We closely matched 97 pairs of retained and promoted children based on their propensity scores using optimal matching procedures. This procedure adjusted for baseline differences between the retained and promoted children. We found that grade retention decreased the growth rate of mathematical skills but had no significant effect on reading skills. In addition, several potential moderators of the effect of retention on growth of mathematical and reading skills were identified including limited English language proficiency and children's conduct problems.
我们使用线性增长曲线模型，对一个学业上有风险的样本进行研究，探讨一年级留级对伍德科克-约翰逊综合数学和阅读分数在三年间增长情况的影响。最初，根据低识字分数确定了一个由784名有留级风险的一年级儿童组成的大样本。基于一年级综合基线测试中收集的72个变量构建了代表留级倾向的分数。我们使用最优匹配程序，根据倾向得分将97对留级和升级儿童进行了紧密匹配。该程序对留级和升级儿童之间的基线差异进行了调整。我们发现，留级降低了数学技能的增长率，但对阅读技能没有显著影响。此外，还确定了几个留级对数学和阅读技能增长影响的潜在调节因素，包括有限的英语水平和儿童的行为问题。
