REGRESSION AND MATCHING FALLACY IN QUASI-EXPERIMENTAL RESEARCH
REGRESSION AND MATCHING FALLACY IN QUASI-EXPERIMENTAL RESEARCH
复制标题
DOI：
10.1177/002246696900300403
复制
发表时间：
1969-01-01
影响因子：
2
通讯作者：
HOPKINS, KD
中科院分区：
文献类型：
--
作者：
HOPKINS, KD
因此，我们可以准确地预测具有给定测试前分数的人在后测试中高于（或低于）任何其他智商分数的比例（前提是满足两个变量之间的线性和均匀性的共同假设）。在第一次测试中得分为70分的人，在第二次测试中平均得分为82分，智商标准差为12分。使用正常曲线表，很明显，大约84%的人会回归，因此即使没有任何练习效果，也会在后测中获得更高的智商分数！一半的遗嘱；gain&dquo;智商12分以上；六分之一的人的智商将达到；in-creased&dquo;提高24分或更多（即智商达到96分或更多）。此外，在智商为70的人当中，大约有10%的人会获得一个
s yx, we can accurately predict the proportion of those with a given pretest score who will fall above (or be-low) any other IQ score on the posttest (provided the common as-sumptions of linearity and homosce-dasticity between the two variables are met). Those scoring 70 on the first test will have a mean of 82 on the second test, with a standard deviation of 12 IQ points. Using a normal curve table, it is readily apparent that about 84 percent will regress and hence re-ceive higher IQ scores on the posttest even without any practice effect! One-half will &dquo; gain&dquo; 12 or more IQ points; one-sixth will have lQs that &dquo; in-creased&dquo; by 24 or more points (ie, obtain IQ scores of 96 or more). Further, about 10 percent of those with an initial IQ of 70 will obtain an