Transcript Comparing TOEIC and two vocabulary tests

A Search for Discipline-Specific
Vocabulary
Masaya Kanzaki
Kanda University of International Studies
JALT 2013, October 27
6:45-7:10, Room 304
Acknowledgements
Scott Thornbury
Ken Hyland
Purpose
To suggest a way to create a
discipline-specific vocabulary list
Materials
• Research article abstracts
• WordSmith Tools 6.0
Why abstracts? (1)
Because they …
• contain the most important information of the
article; the essential vocabulary to discuss the
main topic is included in them.
• are concise; unnecessary words are omitted.
• condense many features of academic writing
into short space.
Why abstract? (2)
They are manageable because they …
• are freely available on the internet.
• can easily be copied and pasted digitally.
• are short.
• have no charts and tables.
About WordSmith Tools
Scott, M. (2013). WordSmith Tools version 6,
Liverpool: Lexical Analysis Software.
There are three tools:
WordList creates a list of words or word clusters.
Concord shows a word or a phrase in context.
KeyWord finds the key words in a text.
How to find “keywords”
By comparing the frequency of each word in a
corpus with the frequency of the same word in a
reference corpus
→In this study, the frequency of each word in
the RAA corpus was compared with that in the
British National Corpus.
Keyness
“Keyness” is determined by Dunning’s
long likelihood test.
Dunning, T. (1993). Accurate methods
for the statistics of surprised and
coincidence. Computational Linguistics,
19, 61-74.
Procedures
• Made a corpus of 1.5 million words with 9,565
abstracts from 33 applied linguistics journals
published between 2001 and 2012 (Table 1)
• Analyzed the corpus with WordSmith Tools
and created lists of most frequent words, key
words and 3-, 4- and 5-word clusters (Tables 2
and 3)
Results
See Tables 2 and 3
Implications
This method of creating a
discipline-specific vocabulary list
can be used any field of study.