West Virginia University.
J Appl Behav Anal. 1973 Winter;6(4):713-6. doi: 10.1901/jaba.1973.6-713.
Issues related to assessing change and retention of change were discussed. An alternative analysis was suggested for the data of a recent study by Walker and Buckley (1972). These authors had found that peer reprogramming, equating stimulus conditions, teacher training, and control groups maintained 77, 74, 69, and 67%, respectively, of appropriate behavior produced in a token economy. Their analysis made no use of baseline levels. Two analyses incorporating baseline scores were suggested. One involved change scores; the other, analysis of covariance using baselines as the covariate. Problems with the data made a clear preference difficult, but it was concluded that either analysis would have resulted in conclusions different from those of Walker and Buckley.
讨论了与评估变化和保留变化相关的问题。为 Walker 和 Buckley（1972）最近的一项研究的数据提出了另一种分析方法。这些作者发现，同伴重新编程、等价刺激条件、教师培训和对照组分别保持了代币经济中产生的适当行为的 77%、74%、69%和 67%。他们的分析没有利用基线水平。建议进行两种分析，纳入基线分数。一种涉及变化分数；另一种是使用基线作为协变量的协方差分析。数据的问题使得很难做出明确的偏好，但得出的结论是，任何一种分析都将得出与 Walker 和 Buckley 不同的结论。
