A meta-analysis of 53 papers with 16,522 participants is conducted in the menu-design literature. To conceptualize different design elements, we extend a four-dimensional model by including two new dimensions—menu card label and menu item characteristics. We find that effect sizes vary among these six dimensions. Specifically, menu card characteristics that concern non-descriptive design elements have the largest effect size. The six dimensions are divided into two broad categories that relate to the entire menu or individual menu items. Our meta-analytical results show that the former category yields larger effect than the latter. Concerning the dependent variables, menu design has a large effect on physiological measures, a moderate effect on intention and attitude, and a marginal effect on actual purchase. Furthermore, laboratory studies yield larger effects than field studies do, and mock menus have stronger effect than real menus. Lastly, publication bias seems to be moderate.

在菜单设计文献中对 16,522 名参与者的 53 篇论文进行了荟萃分析。为了概念化不同的设计元素，我们扩展了一个四维模型，包括两个新维度——菜单卡标签和菜单项特征。我们发现效果大小在这六个维度之间有所不同。具体来说，涉及非描述性设计元素的菜单卡特征具有最大的影响大小。这六个维度分为与整个菜单或单个菜单项相关的两大类。我们的元分析结果表明，前一类产生的影响大于后者。关于因变量，菜单设计对生理测量有很大影响，对意图和态度有中等影响，对实际购买有边际影响。此外，实验室研究比实地研究产生更大的影响，模拟菜单比真实菜单具有更强的效果。最后，发表偏倚似乎是适度的。