The English Crowdsourcing Project (L2 speakers)

Authors: Marc BrysbaertEmmanuel KeuleersPawel Mandera
Updated: Mon 18 May 2020
Source: http://crr.ugent.be/programs-data/lexicon-projects
Type: digital lexicon
Languages: English
Keywords: word-prevalenceword-frequencyword-knowledgecrowdsourcinglanguage-acquisitionvocabulary
Open Access: yes
License:
Documentation: https://doi.org/10.31234/osf.io/fwhdy
Publications: Brysbaert et al. (2020)
Citation: Brysbaert, M., Keuleers, E., Mandera, P. (2020). Which words do English non-native speakers know? New supernational levels based on yes/no decision. PsyArXiv. https://doi.org/10.31234/osf.io/fwhdy
Summary:

To have more information about the English words known by L2 speakers, we ran a large-scale crowdsourcing vocabulary test, which yielded 17 million useful responses. It provided us with a list of 445 words known to nearly all participants. The list was compared to various existing lists of words advised to include in the first stages of English L2 teaching. The data also provided us with a ranking of 61 thousand words in terms of degree and speed of word recognition in English L2 speakers, which correlated r = .85 with a similar ranking based on native English speakers. The L2 speakers in our study were relatively better at academic words (which are often cognates in their mother tongue) and words related to experiences English L2 students are likely to have. They were worse at words related to childhood and family life. Finally, a new list of 20 levels of 1000 word families is presented, which will be of use to English L2 teachers, as the levels represent the order in which English vocabulary seems to be acquired by L2 learners across the world.