Unknown Vocabulary Density and Reading Comprehension: Replicating Hu and Nation (2000)
Author: Benjamin Kremmel, Bimali Indrarathne, Judit Kormos, Shungo Suzuki · Type: academic · Status: draft · URL: https://onlinelibrary.wiley.com/doi/10.1111/lang.12622
Licence: Language Learning (Wiley), 2023; open access PDF available via Wiley pdfdirect link; quotable with attribution
Summary
A registered replication of the study behind the 98% coverage threshold. The authors note the influential 98% figure rests on a regression over just 66 university students in New Zealand. Replicating with 104 Sri Lankan adult L2 learners in a nonacademic context, using five versions of two texts at different unknown-word densities, the original results could not be fully replicated. The practical warning for Immersive: treat lexical coverage thresholds as useful engineering heuristics for grading texts, not as hard scientific laws to quote in marketing, and expect comprehension to vary with reader background and text type.
Insights
- The famous 98% coverage threshold rests on a single regression with only 66 university students, a fact rarely mentioned when the figure is cited. (high, draft)
coverage thresholdvocabularyreplicationthe 98% critical threshold figure is based on findings from a research project in which a regression analysis was conducted with only 66 university students in New Zealand — Abstract
- A 2023 replication with 104 adult L2 learners in Sri Lanka could not fully reproduce the original coverage-comprehension results, so coverage targets should be treated as heuristics rather than laws. (medium, draft)
coverage thresholdreplicationvocabularyThe results of the original study could not be fully replicated. — Abstract