Reading and Writing: An Interdisciplinary Journal 17: 517–536, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
Assessing reading skills by means of paper-and-pencil lexical
decision: Issues of reliability, repetition, and word-pseudoword
MARTINE A.R. GIJSEL, WIM H.J. VAN BON and ANNA M.T. BOSMAN
University of Nijmegen, The Netherlands
Abstract. This study focused on the feasibility of a group-administered paper-and-pencil
lexical-decision test as a plausible alternative or supplementary tool for the assessment of
reading skills. Lexical-decision tests and oral-reading tests were administered to 130 Dutch
students from primary grades 1, 2, 3, 5, and 6. Correlations were moderate to high in low
grades, but declined in the high grades. The reliability of the lexical-decision test assessed
by means of a test–retest procedure was generally good. A second presentation of the
lexical-decision test caused repetition effects (i.e., better performance on the second test), but
generally remained within reasonable limits. The presence of different numbers of pseudo-
words (25% vs. 75%) in both lexical decision and oral reading, indicated that a large number
of pseudowords made oral reading harder, but lexical decision easier. Educational and clinical
implications are discussed.
Key words: Lexical decision, Literacy assessment, Oral reading, Test repetition, Word–
In the Netherlands, reading skills of young students are usually assessed by
means of standardized oral-reading tests. The most widely used tests require
that students read aloud as quickly and accurately as possible, within a given
time-span, a list of isolated words. All oral reading tests have an important
requirement in common, namely, the production of an overt, oral response,
a type of behavior that is usually not involved in ﬂuent, competent reading.
Kusters (1987) investigated this issue and showed that Dutch students who
had been diagnosed as poor readers were better at meaning identiﬁcation
than at word pronunciation, suggesting that these poor readers often knew
more about the meaning of words than was apparent from their oral-reading
performance. A similar ﬁnding has been presented in English by Carlisle,
Stone, and Katz (2001). They showed that both poor and good readers were
better at lexical decision than oral reading, but the difference in perfor-
mance between the two tasks was more pronounced in poor than in good
readers. Lexical decision is probably the most widely used task in reading