Estimating psychological constructs from natural language has the potential to expand the reach and applicability of personality science. Research on the Big Five has produced methods to reliably assess personality traits from text, but the development of comparable tools for personal values is still in the early stages. Based on the Schwartz theory of basic human values, we developed a dictionary for the automatic assessment of references to personal values in text. To refine and validate the dictionary, we used Facebook updates, blog posts, essays, and book chapters authored by over 180 000 individuals. The results show high reliability for the dictionary and a pattern of correlations between the value types in line with the circumplex structure. We found small to moderate (r_s = .1–.4) but consistent correlations between dictionary scores and self‐reported scores for 7 out of 10 values. Correlations between the dictionary scores and age, gender, and political orientation of the author and scores for other established dictionaries mostly followed theoretical predictions. The Personal Values Dictionary can be used to assess references to value orientations in textual data, such as tweets, blog posts, or status updates, and will stimulate further research in methods to assess human basic values from text.

中文翻译：

---

个人价值词典的开发和验证：一种理论驱动的工具，用于调查文本中基本人类价值观的引用

从自然语言估计心理构造具有扩大人格科学的范围和适用性的潜力。对五巨头的研究已经产生了从文本中可靠地评估人格特质的方法，但是用于个人价值的可比较工具的开发仍处于早期阶段。基于Schwartz的基本人类价值理论，我们开发了一种字典，用于自动评估文本中对个人价值的引用。为了完善和验证字典，我们使用了由18万多人撰写的Facebook更新，博客文章，文章和书籍章节。结果表明该字典具有很高的可靠性，并且值类型之间的相关模式符合外接词结构。我们发现中小（r _s = .1–.4），但字典得分与自我报告得分之间的一致性（对于10个值中的7个）。词典分数与作者的年龄，性别和政治倾向之间的相关性以及其他已建立词典的分数之间的相关性基本上都遵循理论预测。个人价值字典可用于评估对文本数据中的价值取向的引用，例如推文，博客文章或状态更新，并将激发人们进一步研究从文本中评估人类基本价值的方法。