Mechanical Design Engineering Department, National Formosa University, No. 64, Wunhua Rd., Huwei Township, Yunlin County 632, Taiwan.
Appl Ergon. 2012 Jan;43(1):141-50. doi: 10.1016/j.apergo.2011.04.006. Epub 2011 Apr 29.
This study explored the degree of rating consistency for different icon-background color combinations, and identified the color combinations with the highest rating for consistency. Rating consistency is the degree of agreement among subjective aesthetic preferences for a specific color combination. In total, 3306 color combinations were rated in this study. The standard deviation of each color combination was calculated as the performance measure for assessing rating consistency among subjects for each color combination. Analysis of variance (ANOVA) results demonstrated that the gender effect was insignificant and rating consistency was very low for both males and females. Furthermore, the experimental results showed that the color combinations affected rating consistency. Hence, cluster analysis with non-hierarchical procedures was then used to categorize color combinations via the clustering variable, the standard deviation of aesthetic preference ratings. Few color combinations in clusters had an extremely high rating consistency. The results indicated that the degree of rating consistency among subjects varied with color combinations. Therefore, using a single average of user aesthetic preference scores may not be appropriate for application to color combinations. Finally, the 30 color combinations determined experimentally can be used as default color combinations as the color scheme for an interface that can be changed because they have consistently high preference scores.
本研究探讨了不同图标背景颜色组合的评分一致性程度,并确定了具有最高评分一致性的颜色组合。评分一致性是指对特定颜色组合的主观审美偏好的一致性程度。本研究共对 3306 种颜色组合进行了评分。每个颜色组合的标准差被计算为评估每个颜色组合的主体之间评分一致性的性能指标。方差分析(ANOVA)的结果表明,性别效应不显著,男性和女性的评分一致性都非常低。此外,实验结果表明,颜色组合会影响评分一致性。因此,然后使用非层次程序的聚类分析通过聚类变量(审美偏好评分的标准差)对颜色组合进行分类。聚类中很少有颜色组合具有极高的评分一致性。结果表明,主体之间的评分一致性程度因颜色组合而异。因此,使用单个用户审美偏好得分的平均值可能不适用于颜色组合的应用。最后,实验确定的 30 种颜色组合可以作为默认颜色组合,用作界面的颜色方案,因为它们具有一致的高分,颜色方案可以更改。