Investigating minimum text lengths for lexical diversity indices

By F. Zenker, K. Kyle. 2021

132 citations recorded by the source corpus.

Abstract Lexical diversity (LD) is an important feature of a second language (L2) writer’s lexical knowledge, and indices of LD have been widely used in the field of writing assessment (e.g., Cumming et al., 2006 ; Engber, 1995 ). Research with longer native speaker (L1) texts has indicated, however, that many commonly used LD indices are sensitive to text length and may conflate lexical breadth and fluency. Because of the importance of measuring LD in L2 writing assessment research, it is essential to know the degree to which particular LD indices are resistant to text length effects and the minimum text lengths at which these indices produce stable values. In this study, we investigate text length effects for nine indices of LD in a corpus of 4542 L2 argumentative essays. The results indicate that MATTR ( Covington & McFall, 2010 ) and two versions of MTLD ( McCarthy, 2005 ; McCarthy & Jarvis, 2010) are the most stable of the indices included in the study. MATTR performs particularly well, maintaining a high degree of stability across all text lengths. Comparisons based on essay prompt and proficiency level are also discussed.

Subfields: Artificial Intelligence, Developmental and Educational Psychology

View Paper: https://www.semanticscholar.org/paper/69712305d9ff719605c8e7657495ad78c3cba141

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

  1. Scott Andrew Crossley
  2. K. Kyle
  3. Scott Jarvis
  4. Danielle S. McNamara
  5. Tom Salsbury
  6. P. M. McCarthy
  7. Judit Kormos
  8. Andrea Révész

In Developmental and Educational Psychology

  1. Scott Andrew Crossley
  2. K. Kyle
  3. Scott Jarvis
  4. Danielle S. McNamara
  5. T. Odlin
  6. Tom Salsbury
  7. P. M. McCarthy
  8. Jeanine Treffers-Daller

Citations in the corpus