A Contribution to the Study of Actuarial and Individual Methods of Prediction
A Contribution to the Study of Actuarial and Individual Methods of Prediction
复制标题
对精算和个人预测方法研究的贡献
DOI：
--
复制
发表时间：
1943
影响因子：
4.4
通讯作者：
T. Sarbin
中科院分区：
文献类型：
--
作者：
T. Sarbin
案例研究法对行为的预测是否比精算法更准确？这项研究是一个实验性的，而不是一个争论性的尝试，以确定两种预测模式的相对准确性。一种受限制的行为形式——学业上的成功——是根据临床（案例研究）材料和先前导出的回归方程来预测的。相关系数表明，案例研究法可能会考虑无数种变量，但它并不比只考虑两个变量的简单统计方法更准确。对预测的分析表明，案例研究方法采用已知预测权值的行为段，并应用其他效率较低的权值。案例研究预测——至少基于效率——不应该取代精算预测。作为精算预测的补充，临床预测没有增加任何东西。讨论了推广到其他类型判据预测的可能性。
Are predictions of conduct more accurate when made by case-study methods than by actuarial methods? This study is an experimental rather than a polemical attempt to determine the relative accuary of the two modes of prediction. A restricted form of behavior-academic success-was predicted from clinical (case-study) material and from a previously derived regression equation. The correlation coefficients demonstrate that the case-study method which presumably accounts for an innumerable asortment of variables is no more accurate than a simple statistical method which accounts for only two variables. Analysis of the predictions suggests that the case-study method takes behavior segments with known predictive weights and applies other weights which are less efficient. Case-study predictions-at least on the grounds of efficiency-should not be substituted for actuarial predictions. As a complement to the actuarial predictions, the clinical predictions add nothing. The posibility of generelizing to the prediction of other kinds of critera is discussed.