Higher numeracy has been associated with decision biases in some numerical judgment‐and‐decision problems. According to fuzzy‐trace theory, understanding such paradoxes involves broadening the concept of numeracy to include processing the gist of numbers—their categorical and ordinal relations—in addition to objective (verbatim) knowledge about numbers. We assess multiple representations of gist, as well as numeracy, and use them to better understand and predict systematic paradoxes in judgment and decision‐making. In two samples (Ns = 978 and 957), we assessed categorical (some vs. none) and ordinal gist representations of numbers (higher vs. lower, as in relative magnitude judgment, estimation, approximation, and simple ratio comparison), objective numeracy, and a nonverbal, nonnumeric measure of fluid intelligence in predicting: (a) decision preferences exhibiting the Allais paradox and (b) attractiveness ratings of bets with and without a small loss in which the loss bet is rated higher than the objectively superior no‐loss bet. Categorical and ordinal gist tasks predicted unique variance in paradoxical decisions and judgments, beyond objective numeracy and intelligence. Whereas objective numeracy predicted choosing or rating according to literal numerical superiority, appreciating the categorical and ordinal gist of numbers was pivotal in predicting paradoxes. These results bring important paradoxes under the same explanatory umbrella, which assumes three types of representations of numbers—categorical gist, ordinal gist, and objective (verbatim)—that vary in their strength across individuals.

中文翻译：

---

数字和算术表示法如何预测决策悖论：模糊轨迹理论方法

在一些数字判断和决策问题中，较高的计算能力与决策偏差有关。根据模糊痕迹理论，理解此类悖论不仅涉及数字的客观（普通）知识，还包括扩展计算的概念，以包括处理数字的要点（其类别和序数关系）。我们评估要点和算术的多种表示形式，并使用它们来更好地理解和预测判断和决策中的系统悖论。在两个样品（Ñs = 978和957），我们评估了数字的分类（一些与没有）和有序要点表示（较高相对较低，如相对强度判断，估计，近似和简单比率比较），客观计算和非语言表达，用于预测的流体智力的非数字量度：（a）表现出阿拉斯悖论的决策偏好，以及（b）有或没有小额损失的投注的吸引力等级，其中损失投注的评分高于客观上优于无损失投注的投注。分类和顺序要点任务预测了悖论性决策和判断中的独特差异，超出了客观计算和智力。客观计算能力是根据字面数字优势预测选择或评级的，而理解数字的分类和序数依据是预测悖论的关键。