Glenn Jerry, Eslinger Paul, Chinchilli Vernon, Eitington Norma Jo, Martel Julie, Salisbury Jill, Karwacki Marianne, Deegan Dorothy
The Pennsylvania State University College of Medicine, Hershey, PA Fax: 717-531-4151; E-Mail:
Adv Health Sci Educ Theory Pract. 1997;2(3):213-220. doi: 10.1023/A:1009799811690.
Learning disabilities (LD) may cause academic failures among university students, but minor accommodations may in some instances allow students to progress satisfactorily. The current study validates a questionnaire for detecting students likely to have an LD. University students, 57 with a documented LD and 134 without, completed a Likert-scale questionnaire to assess the degree of self-perceived difficulty for 50 learning or testing situations. Wilcoxon rank sum test, separate logistic regression models, and determination of the best joint set of predictors of LD were carried out. The mean per-item questionnaire score was higher in the LD group (p < 0.0001); 47 individual items were significant predictors (i.e., p < 0.05) of LD status. The best joint set of predictors consisted of items dealing with word pronunciation, comprehending mathematical word problems, formulating summaries, spelling, slow reading rate, finishing tests in the allotted time, note-taking in lectures, and falling asleep while reading. Incorporating responses to these eight items, a formula was derived to determine the probability of LD. Taking 0.26 as the threshold for predicting the presence of an LD, the sensitivity of the questionnaire was 88%, the specificity was 87% for this population. The questionnaire could be an effective screening tool to detect university or professional students who should consider comprehensive testing for LD.
学习障碍（LD）可能导致大学生学业失败，但在某些情况下，微小的调整可能会使学生取得令人满意的进步。本研究验证了一份用于检测可能患有学习障碍的学生的问卷。57名有学习障碍记录的大学生和134名没有学习障碍记录的大学生完成了一份李克特量表问卷，以评估在50种学习或测试情况下自我感知的困难程度。进行了威尔科克森秩和检验、单独的逻辑回归模型以及确定学习障碍的最佳联合预测指标集。学习障碍组的问卷平均每题得分更高（p < 0.0001）；47个单项是学习障碍状态的显著预测指标（即p < 0.05）。最佳联合预测指标集包括与单词发音、理解数学应用题、撰写总结、拼写、阅读速度慢、在规定时间内完成测试、课堂笔记以及阅读时入睡相关的项目。结合对这八个项目的回答，得出了一个公式来确定学习障碍的概率。以0.26作为预测学习障碍存在的阈值，该问卷的敏感性为88%，对该人群的特异性为87%。该问卷可能是一种有效的筛查工具，用于检测那些应考虑进行学习障碍综合测试的大学生或专业学生。
