LEXSTATS: A program for the statistical analysis of word frequency distributions

Harald Baayen
University of Nijmegan
Max Planck Institute for Psycholinguistics
Wundtlaan 1
PB 310, 6500 AH NIJMEGEN
The Netherlands
baayen@mpi.nl

Fiona J. Tweedie
Department of Statistics
Mathematics Building
University Gardens
GLASGOW, Scotland G12 8QW
fiona@stats.gla.ac.uk

Various computationally intensive statistical models are available for the analysis of word frequency distributions (e.g., Carroll, 1967; Sichel 1975, and Chitashvili and Baayen, 1993). These models provide linguists and lexicographers with elegant means for obtaining sample-size invariant characteristic textual measures, for extrapolating the development of the vocabulary beyond sample sizes larger than the observed text size, and for estimating the population vocabulary size.

Thusfar, these models have not been used widely, which is not surprising given the absence of software implementing these models. At the conference, we will present the beta version of LEXSTATS, a user-friendly GUI interface to a series of C programs that implement a wide range of word frequency analyses. LEXSTATS and the underlying C code will become available as freeware under the GNU software license.

We will illustrate LEXSTATS by applying it to word frequency distributions of various kinds of texts as well as to word frequency distributions of a range of morphological categories.

References

Carroll, J. B.: 1967, "On Sampling from a Lognormal Model of Word Frequency Distribution," In: H. Kucera and W. N. Francis (Eds.), Computational Analysis of Present-Day American English, Brown University Press, Providence, pp. 406-424.

Chitashvili, R. J. and Baayen, R. H.: 1993, "Word Frequency Distributions," In: G. Altmann and L. Hrebicek, L. (Eds.), Quantitative Text Analysis, Wissenschaftlicher Verlag Trier, Trier, pp. 54-135.

Sichel, H. S.: 1975, "On a Distibution Law for Word Frequencies," Journal of the American Statistical Association, 70, 542-547.