Personality Trait Descriptors: 2,818 Trait Descriptive Adjectives characterized by familiarity, frequency of use, and prior use in psycholexical research
David M Condon, Joshua Coughlin, Sara J Weston
*Journal of Open Psychology Data*
Why we built it
Personality researchers have long relied on the lexical hypothesis, the idea that the differences between people that matter are the ones our language has words for. The trouble is that English has many thousands of personality-related adjectives, far more than anyone can rate. The most widely used shortlist, Goldberg’s 1,710 terms, was trimmed partly by judgment, and the familiarity data behind earlier lists came from college undergraduates in the 1960s. We wanted a current, open record of which of these words people actually know.
What we did
First, we assembled 2,818 trait descriptive adjectives: all of Goldberg’s terms, terms from the earlier Norman and Allport and Odbert lists that are common in modern English, and 57 new ones. Second, we took a short definition for each word from Oxford Dictionaries and wrote two multiple-choice vocabulary questions per word. Each question gave the definition, five randomly chosen wrong words, and the options “I don’t know” and “None of these”. Third, we gave the 5,636 questions to 1,572 U.S. adults on Prolific and MTurk in the summer of 2020. Each person answered 75 randomly drawn questions from each of two forms. The survey was shown only to people who had not reported a college degree, so that it would reflect typical literacy more than an undergraduate sample would.
What is in the dataset
The data give the share of respondents who chose the correct word for each term, along with how often each word appears in Google Books and whether it belongs to other well-known lists. The sample was still more White than the U.S. population (73%), and it only covers American English. We did not publish the questions themselves, so that future participants cannot look up the answers. Researchers can email the first author for access.
How to use it
The data are released under a CC0 license, and the code and a searchable database of the words are on the project site. We hope others will use them to choose words for new scales and to test earlier structural findings.