Revisiting fixed- and random-effects models: some considerations for policy-relevant education research
Revisiting fixed- and random-effects models: some considerations for policy-relevant education research
复制标题
DOI：
10.1080/09645292.2013.855705
复制
发表时间：
2015-01-01
影响因子：
1.2
通讯作者：
Vignoles, Anna
中科院分区：
文献类型：
--
作者：
Clarke, Paul;Crawford, Claire;Vignoles, Anna
在教育研究的背景下，使用固定（FE）和随机效应（RE）在两个层次的线性回归进行了讨论。我们比较了有限元模型的鲁棒性与建模的灵活性和潜在的效率，从RE模型。我们认为，这两种方法应被视为互补的方法。然后，我们比较两种建模方法在我们的实证例子。结果表明，特殊教育需要（SEN）的地位对教育程度的负面影响，选择到SEN状态主要是由学生而不是学校层面的因素。
The use of fixed (FE) and random effects (RE) in two-level hierarchical linear regression is discussed in the context of education research. We compare the robustness of FE models with the modelling flexibility and potential efficiency of those from RE models. We argue that the two should be seen as complementary approaches. We then compare both modelling approaches in our empirical examples. Results suggest a negative effect of special educational needs (SEN) status on educational attainment, with selection into SEN status largely driven by pupil level rather than school-level factors.