Lester G, Kelman M
UCLA School of Law 90095, USA.
J Learn Disabil. 1997 Nov-Dec;30(6):599-607. doi: 10.1177/002221949703000603.
We investigated the hypothesis that interstate disparities in the diagnosis of pupils with learning disabilities (LD) are more strongly correlated with demographic and sociopolitical factors than with the biological prevalence of the disability. We also investigated the relationship of these factors to placement practices. Thirteen independent variables representing state characteristics were simultaneously regressed against each of seven static dependent variables, measuring diagnostic and placement practices in 1989, and two dynamic dependent variables, measuring changes in practices between 1976 and 1989. Results of the regression indicated that although demographic and sociopolitical factors explained none of the total prevalence of the four most common physical disabilities (adjusted R2[R2] = 0), they did explain to a moderate degree the state prevalence of LD (R2 ranged from .15 to .28), and were more predictive still depending on measure of LD prevalence. Moreover, these same factors strongly predicted the extent to which states mainstreamed their pupils (R2 = .59) and the size of the nonmainstreamed cognitively disabled (LD and educable mentally retarded) population (R2 = .56).
我们研究了这样一种假设，即学习障碍（LD）学生诊断方面的州际差异与人口统计学和社会政治因素的关联，要强于与该障碍生物学患病率的关联。我们还研究了这些因素与安置方式之间的关系。代表州特征的13个自变量，同时针对七个静态因变量中的每一个进行回归分析，这些静态因变量衡量1989年的诊断和安置方式，以及两个动态因变量，衡量1976年至1989年期间这些方式的变化。回归结果表明，虽然人口统计学和社会政治因素对四种最常见身体残疾的总体患病率没有解释力（调整后的R2[R2]=0），但它们对LD的州患病率有一定程度的解释力（R2范围从0.15到0.28），并且根据LD患病率的衡量标准，其预测性更强。此外，这些相同因素强烈预测了各州将学生纳入主流教育的程度（R2 = 0.59）以及未被纳入主流教育的认知障碍（LD和可教育的智力迟钝）人群的规模（R2 = 0.56）。
