Optimizing Vocabulary Learning Through Text Covering Efficiency and Word Tier Analysis

Dr. Tatsuhiko Matsushita is a research professor at National Institute of Japanese Language and Linguistics (, or NINJAL) in Japan. 

It is generally true that the more words in a text that are known, the better the comprehension will be. L2 vocabulary learning, however, is a significant burden for learners, and exploring efficient vocabulary learning order is one of the applications where quantitative linguistics can potentially contribute to second language education. This study demonstrates the usefulness of an index titled Text Covering Efficiency (TCE) to perform Word Tier Analysis (WTA) proposed by Matsushita (2012). TCE is a simple and robust index which is the expected per-million-word frequency (text coverage) of a word in a group of words. TCEs of various grouped words in Japanese sub corpora were calculated to exemplify the order in which various groups of words can be learned most efficiently. It will also be shown that lexical differences among various text genres can be clarified by WTA using TCE and there are differences in the relative importance of different groups of words according to the purpose of learning.

This event is generously co-sponsored by Department of Linguistics at University of Pittsburgh,  SLA program of Modern Languages at Carnegie Mellon University, Department of East Asian Languages and Literatures at University of Pittsburgh, Language Technology Institute at Carnegie Mellon University, Learning and Research & Development Center at University of Pittsburgh, and the Department of Philosophy, Linguistics Program, at Carnegie Mellon University.