To the top

Page Manager: Webmaster
Last update: 9/11/2012 3:13 PM

Tell a friend about this page
Print version

Quantitative measures on … - University of Gothenburg, Sweden Till startsida
Sitemap
To content Read more about how we use cookies on gu.se

Quantitative measures on student texts

Conference paper
Authors Ulrika Magnusson
Sofie Johansson Kokkinakis
Published in Papers from the ASLA Symposium in Stockholm, 7-8 November, 2008, Association suédoise de linguistique appliquée (ASLA)
Volume Language and Learning
Issue ASLA:s skriftserie nr 22
Pages 43-56
ISBN 978-91-87884-22-1
ISSN 1100-5629
Publication year 2009
Published at Department of Swedish
Pages 43-56
Language en
Links demo.spraakdata.gu.se/svesj/publica...
Keywords elevtext, kvantitativ analys, datorbaserad analys, statistiska mått, svenska som andraspråk, språkteknologi
Subject categories Mathematical statistics, Language Technology (Computational Linguistics), History and Archaeology, Bilingualism

Abstract

Quantitative measures have been widely used in Swedish research on students texts (Hultman & Westman 1977, Nyström 2000, Östlund-Stjärnegårdh 2002), often related to grades and to spoken and written style. However, there are few quantitative descriptions of mono- and multilingual students’ texts. In our study, we compare 220 national tests in Swedish written by first- and second language secondary students living in urban areas where the students have various linguistic background, gathered within the project Language and language use among students in multilingual urban settings. The purpose of this study is to investigate whether lexical quantitative and qualitative statistical measures can be used to describe and compare different aspects of this text corpus. We found four measures that suited this purpose: lexical density – comparing the number of content words vs. total number of words; nominal ratio – measuring number of nominal phrases vs. verbal phrases in a text; word variation index – measuring number of different words or lexical items found in a text; and word length – describing the length of a token. The three dependencies investigated were: linguistic background vs. grades, student variables vs. quantitative measures, and correlation between quantitative measures.

Page Manager: Webmaster|Last update: 9/11/2012
Share:

The University of Gothenburg uses cookies to provide you with the best possible user experience. By continuing on this website, you approve of our use of cookies.  What are cookies?