McManus I C
Br J Psychol. 1980 Nov;71(Pt 4):505-24. doi: 10.1111/j.2044-8295.1980.tb01763.x.
After reviewing the literature on experimental rectangle aesthetics ('the golden section') it was concluded that all of the effects demonstrated in the literature were dubious, either due to methodological limitations or inadvertent experimenter bias, these defects being compounded by most studies only considering population preferences and ignoring individual differences in preference. A series of experiments is described in which highly significant and temporally stable, but somewhat idiosyncratic, individual preferences were found. A taxonomy of these preferences, as well as of those for three separate types of triangle preference, is provided, based on factorial analysis. Two clear factors were demonstrated, one based on the square, and the other upon a proportion similar to that of the golden section.
在回顾了关于实验性矩形美学（“黄金分割”）的文献后得出结论，文献中所展示的所有效果都值得怀疑，要么是由于方法上的局限性，要么是由于实验者无意中的偏差，而大多数研究只考虑总体偏好并忽略偏好的个体差异，这些缺陷使得情况更加复杂。本文描述了一系列实验，其中发现了高度显著且随时间稳定但有些独特的个体偏好。基于因子分析，提供了这些偏好以及三种不同类型三角形偏好的分类法。结果表明有两个明显的因素，一个基于正方形，另一个基于与黄金分割比例相似的比例。
