Issues in the Assessment of

Reading Disabilities in L2
ChildrenBeliefs and
Research Evidence
Esther Geva*
Ontario Institute for Studies in Education of the University of Toronto, Canada

In bilingual and multilingual settings one is constantly

challenged by the difficulty of teasing apart phenomena
associated with normal second language (L2) reading acquisition
from authentic warning signs of reading failure. The bulk of this
paper focuses on a critical discussion of a cluster of beliefs that
pertain to the issues concerning the diagnosis of reading
disability in multilingual and bilingual settings among school
children. Findings from available research on reading acquisition
among bilingual children and research focusing specifically on
the assessment of English-as-a-second language (ESL) children
who might be at risk for reading disability are used to evaluate
the validity of these beliefs. While some beliefs are supported by
research, others are not. In particular, the research suggests that
reliable diagnosis of dyslexia among ESL children can be
achieved by examining within-language differences on various
indices of basic reading skills such as phonological processing,
and by noting a significant gap between oral and reading
comprehension. Copyright 2000 John Wiley & Sons, Ltd.
Keywords: reading disabilities; bilinguals; ESL; assessment issues; diagnosis

INTRODUCTION

estern, industrialized countries are becoming increasingly heterogeneous as a result of international migration, and most immigrants are attracted to metropolitan centres. For example, about
42% of the total non-white population in Canada live in Toronto. Relatedly,
* Correspondence to: Esther Geva, Human Development and Applied Psychology, Ontario
Institute for Studies in Education of the University of Toronto, 252 Bloor W., Toronto,
Ontario M5S 1V6, Canada. E-mail: egeva@oise.utoronto.ca

This paper is based on a keynote address presented at the First Conference on Multilingualism, Bilingualism and Dyslexia, Manchester, UK, June 1999.
Copyright 2000 John Wiley & Sons, Ltd.

DYSLEXIA 6: 13 28 (2000)

14

E. Geva

by 1996, while a mother tongue other than English or French was reported
by 17% of the general population in Canada (Statistics Canada Census, 1996),
in metropolitan areas such as Toronto, which is considered to be the most
ethnically diverse city in the world, it has reached 42%.
Similar statistics apply in other large metropolitan areas in other Western
countries. Given such demographic trends, it is no wonder that there is a
growing awareness and concern over the ability of educational, social and
mental welfare systems to meet the needs of an increasingly diverse clientele
(Cole, 1996; Snow, Brows and Griffin, 1998; Geva, Barsky and Westernoff,
2000). These statistics have far-reaching implications for educational planning. They also have far reaching implications for addressing the needs of
immigrant children who may have a reading disability. This paper deals
with issues concerning the latter point.
Not too long ago, immigrant children and children from certain ethnic
backgrounds were much more likely to be identified as learning-disabled
(LD) and placed in special education classes (Cummins, 1984). Theoreticians
such as Ogbu (1978) and Cummins (1989) cautioned that in Western countries psychological assessments were (mis)used to legitimize the educational
disabling of these children. Critics of these practices warned that children
with limited English proficiency were over-represented in programmes designed for learning-disabled or language-handicapped children (Benavides,
1989; Ortiz and Yates, 1989).
Researchers and clinicians pointed out that professionals often have little
understanding of issues related to limited second language (L2) proficiency,
leading to misinterpretation of data gathered as part of the referral and
evaluation process and to erroneous decisions made using these data. Social
critics argued that placement in special education was often related to
socioeconomic, linguistic and cultural factors rather than psychoeducational
factors. The appreciation that educational difficulties may reflect normal
linguistic and acculturation processes has percolated to educational and
clinical practice in North America. In response to such criticisms and a
genuine concern over the social risks associated with biased practices, there
has been an emergence of alternative, culturally sensitive assessment procedures (Feuerstein et al., 1981; Gavillan-Torres, 1983; Campione, 1989; Duran,
1989; Figueroa, 1989; Oller and Damico, 1990; Dao, 1991; Schiff-Myers, 1992).
These changes led also to changes in professional training (Lewis, 1989; Ortiz
and Yates, 1989; Oller and Damico, 1990; Cole, 1996) and to a set of
concomitant powerful beliefs that guide practice and professional training.
These beliefs pertain to how children learn to communicate in L2 and how
they acquire literacy skills in L2; they also pertain to beliefs about sources of
learning difficulties in L2 learners and to ways to minimize biased assessment of oral and written language skills in children.
The bulk of this paper focuses on a critical discussion of a cluster of beliefs
concerning the diagnosis of reading disability in multilingual and bilingual
settings among school children. In this paper the term reading disability
refers to difficulty in developing word recognition skills. It is argued that
while some of these beliefs are supported by research, others have not
been examined within a rigorous research framework. It should be noted
that there is a dearth of research on these issues. Where relevant, I will
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

Assessment of Reading Disabilities in L2 Children

15

summarize findings from my research programme concerned with the identification of English-as-a-second-language (ESL) at-risk children. The discussion of these beliefs is preceded by a brief description of the methodology
and objectives of the research on ESL at-risk children.
THE ESL AT-RISK RESEARCH PROGRAMME METHODOLOGY AND
DESIGN
This research programme is concerned with the development of betterinformed approaches and benchmarks for the early identification of ESL
learners who might be at risk of being reading-disabled. In order to gain a
better understanding of the correlates and precursors of normal and delayed
reading ESL development in the primary grades, the development of the
following skill clusters is being tracked: (a) components of ESL literacy skills
such as word recognition, pseudoword decoding, spelling, reading comprehension and reading efficiency; (b) ESL oral proficiency indices such as
listening comprehension, receptive and expressive vocabulary and grammatical knowledge; and (c) development of prerequisite skills such as phonological processing and awareness, verbal memory and speed of naming.
Non-verbal intelligence is also assessed periodically.
A cross-sequential, longitudinal design is used in the ESL at-risk project,
involving three cohorts (total n =331) in four low-to-middle SES areas of
metropolitan Toronto. Mindful of the potential effects of first language (L1)
on ESL development, primary-level children from specific language groups
were targeted (e.g. Punjabi, Cantonese, Tamil, Portuguese) because they
reflect current demographic trends and represent different languages. This
allows us to examine universal and language-specific trends in the development of reading and language skills in primary-level ESL children. Only
children who speak their L1 at home and who at the onset of the study have
been in Canada for at least 6 months are included. By including children
from two to four schools in each of the areas, school effects are minimized.
Data collection is not complete yet, and results referred to here pertain to
the first cohort of 35 English-as-L1 (EL1) children and 69 Punjabi-as-L1
children who come from the same four schools, and 35 children with
Cantonese as L1, who come from four schools in a different area. The
Cantonese children come from middle-class backgrounds. The EL1 and
Punjabi children come from a low-to-lower middle-class background. All the
linguistic and reading tasks are presented in English. Children are being
tested twice a year, in the fall and spring, and all testing is done on an
individual basis.
Children of the first cohort, who are now in Grade 3, have been followed
since Grade 1 (mean age 6 years 3 months at Time 1, i.e. fall of Grade 1). It
should be pointed out that all instruction in the schools takes place in
English, though some children may attend heritage language classes, where
they are exposed to literacy instruction in their L1 on weekends or after
school. Information about childrens language status (ESL/L1) is noted when
children first register at a school. Subsequently it was confirmed by interviews with classroom teachers and by our testing. At the time of testing,
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

16

E. Geva

many of the ESL children were attending, on a withdrawal basis, special ESL
classes offered by the respective schools. These 3040 minute classes are
offered a number of times each week. Where standardized measures are
involved, published test norms are not employed, since the norms were not
developed for ESL children. In data analyses raw scores or standard scores
and percentiles, calculated within language groups, are used. This procedure
is essential to minimize bias and confounding with oral language proficiency
in defining at-risk status among ESL children.

BELIEFS ABOUT BILINGUALISM AND READING DISABILITY IN

YOUNG CHILDREN
Belief: Children Experience Difficulties in Reading in Their L2 Because Their
Oral Language Proficiency Is Not Adequate Yet

According to a global, L2 proficiency based explanatory framework, it

would be reasonable to expect to find a gap between childrens performance
on parallel L1 and L2 reading tasks. Poorer text reading in L2 may be the
result of the cumulative effects of slow and inefficient decoding, of slow and
inefficient activation of the existing lexicon, as well as of incomplete, slow
and inefficient activation of morphosyntactic knowledge. A rather widespread belief among educators and some L2 researchers concerns the notion
that in order to learn to read it is necessary to ensure adequate levels of
language proficiency in L2. Language proficiency and reading skill develop
gradually (Chall, 1996). For L2 school children, language proficiency does
not precede reading development the way it does when children learn to
read in their L1 (Chall, 1989). Instead, for the child learning to read in L2,
language and reading are developing in tandem. It is not surprising therefore that educators and researchers who often focus on the development of
reading comprehension in ESL learners tend to attribute reading difficulties
among L2 learners to poor language proficiency and to downplay the role of
word recognition processes.
Oral language proficiency in school children has been examined in terms
of global conversational and academic linguistic indices (Cummins, 1984;
Dickinson and Snow, 1987; Snow et al., 1989; Ricard and Snow, 1990). Others
discuss L2 oral proficiency as involving phonological, lexical, morphological
and syntactic skills (Geva and Ryan, 1993; Johnson and Newport, 1989).
Only a limited amount of direct research exists, however, on the relationship
between language proficiency and L2 childrens developing reading skills.
To address the research pertaining to the belief that L2 reading is enhanced with better developed linguistic proficiency, it is useful to consider
three bodies of research, one pertaining to reading comprehension processes,
another pertaining to word recognition processes, and the third relating to
the relationships between word recognition and reading comprehension.
L1-based research supports the existence of a positive relationship between language and reading comprehension in elementary school children
(e.g. Chall, 1989; Dale, Crain-Thoreson and Robinson, 1995; Carver, 1997;
Rupley and Willson, 1998). The available research on L2 also supports the
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

Assessment of Reading Disabilities in L2 Children

17

belief that comprehension-based aspects of reading, including reading fluency, are facilitated by increasing levels of L2 oral proficiency. Several
studies have shown that various aspects of L2 language proficiency play a
significant role in reading comprehension. This appears to be true when
children come from middle-class backgrounds and are studying to read in
their L2 in frameworks which value bilingual literacy development (e.g.
Chitiri et al., 1992; Geva and Clifton, 1993; Geva and Ryan, 1993; Geva,
Wade-Woolley and Shany, 1997). It also appears to be true when children
come from less privileged, immigrant backgrounds (e.g. Verhoeven, 1990;
Durgunoglu, Nagy and Hancin-Bhatt, 1993; Schuster and Geva, 1999). For
example, in a recent paper based on the ESL at-risk project, Geva and
Petrulius-Wright (2000) used vocabulary and syntactic judgement as indices
of oral language proficiency. They report that the L1 children outperformed
the Punjabi-as-L1 children on these English oral language proficiency indices
and on reading comprehension. In both groups vocabulary and syntactic
knowledge were good predictors of reading comprehension. In another
study based on the ESL at-risk project, Schuster and Geva (1999) report no
L1L2 differences on indices of reading efficiency; by the time they were in
Grade 2, both L1 and ESL children did not differ on accurate and fast
reading of simple narratives.
When one examines the research pertaining to word-based processes in
children learning to read in their L2, evidence for the importance of oral
language proficiency is somewhat less compelling. In recent years a growing
body of research has accumulated which shows that word-based reading
processes in children are less closely related to oral language proficiency. For
example, Durgunoglu, Nagy and Hancin-Bhatt (1993) found that oral language proficiency in L2 (English) did not predict basic reading skills (i.e.
word recognition, pseudoword reading) in Latino children, while phonological processing skills did. Other studies involving young school children and
various writing systems being learned as L2 (e.g. French, Hebrew, Farsi)
have found only weak relationships between L2 proficiency and performance on measures of word recognition and pseudoword decoding skills in
L2 (Geva and Clifton, 1993; Geva, Wade-Woolley and Shany, 1997; Geva and
Siegel, 2000; Gholamain and Geva, 1999). In fact, Geva and colleagues found
in a series of studies that children were able to decode words with more
accuracy in Hebrew, L2, than in English, their L1. These counterintuitive
results were attributed to the fact that Hebrew has a less complex (shallow)
orthography than English (see Geva, Wade-Woolley and Shany, 1993; Geva
and Wade-Woolley, 1998; Geva and Siegel, 2000). In these studies oral
language proficiency contributed only 3%7% to the explained variance.
Verhoeven (1990), however, reports a positive significant relationship between Dutch oral language proficiency and Dutch word recognition skills for
Turkish-as-L1 primary-level children being instructed in Dutch (L2).
In the ESL at-risk project, Geva, Yaghoubzadeh and Schuster (1999) found
that ESL childrens performance did not differ from that of the EL1 children
on basic reading skills such as word recognition and decoding skills. At the
same time, statistical analysis indicated that the EL1 children were more
proficient in English and performed significantly better than the two L2
groups on all the language proficiency measures (though the children with
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

18

E. Geva

Cantonese as L1 performed somewhat better than the Punjabi children). In

particular, L1 children performed significantly better than the ESL children
on listening comprehension. Contrary to the pattern noted on the oral
language measures, the Punjabi and EL1 groups did not differ from each
other on reading measures (though the Cantonese group means were higher
than those noted in the Punjabi group).
Conclusion

The belief that individual differences in L2 reading comprehension are

related to L2 oral language proficiency is supported by research. Nonetheless, the research does not support a strong version of this belief when it
comes to word-based reading skills. Instead, it appears that, provided that
children have been exposed to appropriate literacy instruction, there is no
reason why they should not be able to decode words even when their L2
language proficiency continues to develop. In fact, depending on the complexity of the L2 writing system (i.e. whether it is deep or shallow)
children may sometimes be able to decode words in L2 with more accuracy
than in their L1. Oral language proficiency plays only a marginal role in
explaining why some young L2 learners continue to experience difficulties in
reading words and pseudowords in spite of adequate instruction. For some
children the difficulties reside in word-based component processes and not
in limited vocabulary or incomplete command of the grammar.
Belief: The Relationship Between Word Processes and Reading Comprehension
is Different in L1 and L2

Hoover and Gough (1990) argued that there is an affinity between word
recognition and reading comprehension skills. Carver (1993, 1997) and Rupley and Willson (1998) have shown with L1 learners that slower rates of
word recognition affect comprehension and that word recognition continues
to play an important role in reading comprehension through Grade 6. To
what extent can findings pertaining to the relationships between word and
text processes in children learning to read in their L1 apply to children
learning to read in L2? In a study of young EnglishHebrew bilingual
children, Geva, Wade-Woolley and Shany (1997) found that despite differences in language familiarity, accuracy and speed rates of isolated word
reading in English, L1, and Hebrew, L2, were highly similar. At the same
time, in L1, children were able to rely on their linguistic knowledge and
were therefore more efficient in reading narratives than isolated words. This
pattern was not observed in Hebrew, L2, presumably because childrens
command of L2 was below some unspecified L2 proficiency threshold.
One of the outcomes of the ESL at-risk research project is the finding that
efficient word recognition plays a significant role in facilitating the comprehension of simple narratives by young ESL learners. Schuster and Geva
(1999) found that while vocabulary knowledge, a measure of oral language
proficiency, did not predict significantly text reading efficiency of L1 or ESL
children, accurate word recognition did for both L1 and ESL children.
Merbaum and Geva (1998) report that listening comprehension, a discoursebased measure of oral language proficiency, explained 22% of the variance
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

19

Assessment of Reading Disabilities in L2 Children

on the performance of ESL children on a reading comprehension task

involving simple narratives. However, with the addition of word-based
measures (pseudoword decoding, phonological processing and rapid naming), 82% of the variance in L2 reading comprehension was accounted for. In
other words ESL children who have a better command of English are more
likely to comprehend simple narratives. However, by considering also word
based measures, it was possible to understand with much more accuracy
individual differences in reading comprehension.
Conclusion

A better picture of what drives reading comprehension in young ESL (and

L1) learners can be achieved by considering, along with oral language
indices, factors associated with efficient word recognition skills. These results suggest that, as is the case in L1, some ESL children may not be able to
comprehend text not only because their oral language is not fully developed,
or because they do not use comprehension strategies efficiently, but also
because their word recognition skills are poor. Such results suggest that it is
beneficial to develop and track concurrently oral language proficiency as
well as word recognition and reading comprehension skills. This is especially important because of the mutually enhancing role that exists between
reading and vocabulary growth.
Belief: Poor Phonological Processing Skills May Contribute to Poor Word
Recognition in L1 Children, but Are Less Informative When Oral Language
Proficiency is Still Developing

In L1-based research, phonological processing has been linked to the successful development of word recognition skills (Share et al., 1984; Wagner
and Torgeson, 1987; Schneider and Naslund, 1993; Vandervelden and Siegel,
1995). Phonological processing is assessed by measuring childrens ability to
conceptualize and manipulate sublexical elements. This is often done by
asking children to delete, count or segment phonemes. The research indicates that the development of phonological processing skills follows a
developmental path, with awareness of syllables developing before awareness of onset and rhyme, which in turn develop before awareness of
phonemes. The ability to reflect on and manipulate phonemes increases the
chances that young children will develop good decoding skills, leading to
success at reading from the early grades. Research has also shown that
children who experience reading difficulty are those who have limited
ability to perform sublexical manipulation tasks (Adams, 1990; Rack, Snowling and Olson, 1992; Stanovich, 1992; Siegel, 1993), and that these deficits
persist until adulthood (Bruck, 1990; Pennington et al., 1990).
The relation between reading difficulty and poor phonological processing
skills is not restricted to English. Phonological awareness has been shown to
play an important role in the development of reading skills in other Latinbased and non-Latin based orthographies, including Danish, French, German, Hebrew, Italian, Portuguese, Spanish, Turkish and Swedish (Alegria,
Pignot and Morais, 1982; Cossu et al., 1988; Wimmer and Hummer, 1990;
Naslund and Schneider, 1991; Caravolas and Bruck, 1993; Ben-Dror, Frost
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

20

E. Geva

and Bentin, 1995; Oney and Durgunoglu, 1997; Elbro, Borstrom and Petersen, 1998). Phonological processing skills appear to play a role in logographic languages such as Chinese as well (Hu and Catts, 1998).
Of particular relevance to this paper is the rather consistent finding in a
handful of cross-linguistic studies that phonological processing skills in one
language are predictive of word recognition skills within and cross-linguistically. Again, this finding seems to be universal, as it occurs across different
language combinations such as English French, EnglishHebrew, English
Spanish and English Farsi (e.g., Durgunoglu, Nagy and Hancin-Bhatt, 1993;
Cisero and Royer, 1995; Geva and Siegel, 2000; Gholamain and Geva, 1999).
The notion that aspects of phonological processing (as well as related
processes such as naming speed and phonological memory) are linguistically
interdependent appears to be substantiated by these studies. Wade-Woolley
and Geva (2000) found evidence for this L1L2 linguistic interdependence in
a study focusing on childrens performance on an English (L1) phoneme
deletion task. They report that childrens performance on this task was
correlated with a number of Hebrew (L2) tasks, including word reading,
sentence comprehension, phoneme identification and skills in recognizing
orthographic patterns. In a longitudinal study of young school children
learning to read in French within a French immersion setting, Comeau et al.
(1999) also found evidence for such transfer. Owing to this linguistic interdependence, performance on phonological awareness skills predicted word
recognition cross-linguistically.
Conclusion

The ability to reflect on and conceptualize the sounds of ones native

language is clearly related to the ability to read words and understand the
L2 oral language. The cross-linguistic research reviewed here provides
support for a universalist transfer notion. According to this notion, the same
set of linguistic and cognitive predictors underlies the development of
reading skills in L1 and L2 (even when different alphabetic writing systems
are involved). Across different writing systems individual differences in
phonological processing skills are related to individual differences in word
recognition skills. That phonological processing skills in one language, be it
the learners L1 or L2, can explain significant amounts of variance on word
recognition skills in another language (L1 or L2) is especially important in
the context of assessing word recognition skills in multilingual contexts such
as that of the ESL at-risk children. This emerging pattern suggests that it is
possible to assess reading disability cross-linguistically even when linguistic
proficiency is not fully developed yet. This point is the focus of the next two
subsections.
Belief: Processes Driving Reading Development in L1 and L2 Differ

The ESL at-risk research programme was designed to develop a better

understanding of what should be considered normal ESL literacy development, what are some of the early linguistic and cognitive warning signs of
reading disability among ESL learners, and whether these warning signs are
common to L1 and ESL children. Geva et al. (1999) examined whether
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

Assessment of Reading Disabilities in L2 Children

21

cognitive and linguistic correlates used to identify at-risk L1 learners are

equally reliable for identifying at-risk ESL learners. Using a longitudinal
perspective, stepwise regression analyses were performed with Grade 1
measures used to predict performance on various reading indices measured in Grade 2.
In the first analysis, pseudoword readingthe Word Attack subtest of
the Woodcock Reading Mastery Test (Woodcock, 1987)was the dependent variable. In these analyses the oral language proficiency measures,
(consisting of various oral language proficiencies) explained 12% of the
variance in the L1 group and 7% in the L2 group. The phonological
processing factor (consisting of three phonological processing measures)
explained a substantial proportion of the variance in both language
groups: 50% in the L1, and 41% in the L2 group. Rapid automatized
naming (RAN) was another factor that was significant in both language
groups. It explained an additional 8% in the L1, and 5% in the ESL group.
In other words, oral language proficiency was a much less important
predictor of pseudoword decoding than phonological processing and
rapid automatized naming.
This pattern repeated itself when word recognitionthe Reading subtest of the Wide Range Achievement TestRevised, (WRAT) (Jastak and
Jastak, 1984)was the dependent measure. Once again, the oral language
factor, entered first, explained 12% of the variance in the L1 group and
8% in the L2 group. An additional 38% of the variance in the L1 group
was explained by the phonological processing factor, and 44% in the L2
group.
As might be expected, when reading comprehension was the dependent
variable the pattern of results was different. The oral language factor explained 28% of the variance in L1 and 13% in L2. The phonological factor,
which was considered next, did not explain additional variance in either
group. However, word recognition skills (measured with the WRAT) explained an additional 13% of the variance in L1 and 13% in L2. That is
both oral language proficiency and accurate with word recognition are
important in understanding individual differences in reading at the end of
Grade 2 comprehension regardless of language background.
Conclusion

Among L1 and ESL children alike, better-developed phonological processing skills in the spring of Grade 1 were associated with more accurate
decoding and word recognition skills in English a year later. In turn,
word recognition skills predicted reading comprehension of narratives in
both groups. Regardless of oral language proficiency, individual differences in phonological awareness and rapid letter naming exert their influence on word recognition among children learning to read in their L1 or
in their L2. Individual differences on such prerequisite skills can be indicative of smooth or problematic acquisition of ESL reading skills in later
years. Again, we can conclude that developing good reading comprehension among ESL children entails a combination of good word recognition
skills and L2 proficiency
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

22

E. Geva

Belief: It Is Difficult to Diagnose Accurately Reading Disability in L2 Until

Oral Language Proficiency is Well Established

One of the outcomes of the growing awareness that children from ethnic and
linguistic minorities were over-represented in programmes for slow learners
or for the learning-disabled has been the emergence of a trend whereby
professionals and school officials avoid diagnosing ESL learners as learning
disabled for a number of years in order to minimize practices believed to be
insensitive or biased. Cummins (1984) suggested that it may take 57 years
to achieve such oral language proficiency. The logic underlying this practice
has been a concern that the validity of psychoeducational assessments may
be confounded with oral language proficiency. According to this position, as
long as the learner has not had a chance to develop full oral language
proficiency in L2, reliable and valid L2-based assessment of learning problems cannot be achieved.
The common practice of delaying assessment and remediation for several
years is well intentioned. Yet, it may lead to under-identification and
cumulative deprivation. This is a phenomenon I and my colleagues experience often in Toronto. One outcome of the ESL at-risk research is the finding
that elementary classroom teachers tend to be more hesitant and less reliable
in nominating as potentially at-risk ESL than L1 children (Limbos and Geva,
1999). This is related to the fact that they attribute poor reading skills to L2
status and assume that until childrens English is well developed, deficits
related to reading disabilities cannot be identified.
To identify at-risk children, scores were converted to standard scores.
These calculations were based on the means and standard deviations within
each language group. Children were defined as at risk in comparison to their
reference group. This was done to distinguish children who are reading
disabled from garden variety poor readers. A child was classified as at
risk if at the end of Grade 2 they performed at least one standard deviation
below the mean on the WRAT and, in addition, their oral proficiency indices
were at least in the average range.
Figures 1 and 2 present the profiles of two children showing a typical
profile of reading disability, the first with English as L1 and the second with
Punjabi as L1, respectively. The profiles are based on within-group standard
scores and should be examined in relation to their peer group. Three clusters
of selected measures are included. Moving from left to right, each profile
shows first cognitive-linguistic skills such as non-verbal ability (Matrix
Analogies Test, Naglieri, 1989), measured at the end of Grade 1 (Time 2) and
then again at the end of Grade 2 (Time 4); rapid automatized naming of
letters (RAN), measured in the fall of Grade 1 (Time 1) and at Time 4; and
phonological processing skills (phoneme segmentationdeletion) measured
at Time 1 and Time 4. The oral language profile appearing in the centre
includes Time 1 and Time 4 measures of receptive vocabulary (PPVT-R;
Dunn and Dunn, 1981), grammatical knowledge, measured with a sentence
repetition task, and listening comprehension of simple narratives measured
at Time 2 and Time 4. The reading profile cluster, appearing on the right,
includes reading comprehension of simple narratives measured at Time 2
and Time 4. (Note that reading comprehension was not measured at the
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

Assessment of Reading Disabilities in L2 Children

23

Figure 1. Cognitive, linguistic and reading profile of an at-risk English-as-L1 child

(group mean =0.00).

beginning of Grade 1 as this would be too challenging to many.) The profile

also includes, at the far right, Time 1 and Time 4 word recognition (WRAT)
and pseudoword decoding (Woodcock, 1987).
As can be seen, the EL1 child and the Punjabi child whose performance is
described in the two profiles were slower than their respective peer groups
in naming letters rapidly and had difficulties in performing the phonemic
segmentationdeletion task. In comparison to their peers, their performance

Figure 2. Cognitive, linguistic and reading profile of an at-risk Punjabi-as-L1 child

(group mean =0.00).
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

24

E. Geva

was worse at Time 4 than at Time 1. In other words, while their peers
performance improved, their performance did not. In fact, the oral language
skills of both children was higher than the respective mean in their peer
groups. In each case their ability to comprehend stories read to them was
also not impaired, and indeed both performed above their group means.
However, both children performed below their comparison group on all
reading measures. There was a gap of at least two standard deviations
between their listening and reading comprehension scores, and their word
recognition and pseudoword reading scores were likewise very low.
These are typical profiles of children with dyslexia. Notably, 6% of the
children in the EL1 group, 7% of the children in the Punjabi group and 5.5%
of the children in the Cantonese group showed such profiles. In other words,
the prevalence of reading disability in each of the language groups is very
similar. This pattern suggests that the procedure used minimized overidentification and under-identification in each group.
CONCLUSIONS
It appears that it is not necessary to wait until oral language proficiency is
fully developed to note and assess reading disability. Conclusions drawn
from this and related research challenge simplistic notions concerning the
emergence of normal and problematic ESL reading skills. Performance on
oral language proficiency indices is not a good predictor of basic reading
skills in L2. Regardless of L1 L2 status it is erroneous to rely heavily on oral
language proficiency when adequate reading skills development is at issue.
Instead, it appears that two sets of complementary procedures, which are
useful in L1, are useful in identifying reading disability in bilingual or ESL
children as well. The first involves the assessment of phonological processing skills and rapid basic reading skills. The second procedure involves
checking whether there is a gap between reading and listening comprehension. For normally achieving children, who receive consistent instruction at
school this gap should be rather small. A striking gap between listening and
reading comprehension, with listening comprehension being much higher
than reading comprehension, is an important clue. It indicates that the
difficulty the child is experiencing is not with processing and comprehending verbal information, but with processing print.
In bilingual and multilingual settings one is constantly challenged by the
difficulty of teasing apart phenomena associated with normal L2 reading
acquisition from authentic warning signs of reading failure. The research
suggests that assessment can and should proceed separately for oral and
written language indices and that the assessment of reading disability may
proceed in spite of inadequate oral language proficiency in L2. In fact,
relying heavily on oral language proficiency as a clue may be misleading
and costly to the child. Refraining from an assessment until a child who has
serious difficulty in developing word recognition skills has developed oral
language skills could be damaging. In addition to the need to develop oral
language communication skills, this child requires intensive instruction
focusing on the development of word recognition skills and of compensatory
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

25

Assessment of Reading Disabilities in L2 Children

strategies. Our research suggests that rather than waiting until oral language
proficiency is fully developed, it is possible to employ by Grade 2 the same
set of tasks to chart, within EL1 and ESL groups alike, profiles of children
who fail to learn to read in spite of instruction and adequate oral language
proficiency

ACKNOWLEDGEMENTS

The ESL at-risk research was partially funded by research grants from
the Social Sciences and Humanities Research Council of Canada and the
Ontario Ministry of Education to the author.

References
Adams, M.J. (1990) Beginning to Read: Thinking and Learning about Print. Cambridge, Mass.:
MIT Press.
Alegria, J., Pignot, E. and Morais, J. (1982) Phonetic analysis of speech and memory codes in
beginning readers. Memory and Cognition, 10, 451456.
Benavides, A. (1989) High risk predictors and prereferral screening for language minority
students. In A.A. Ortiz and B.A. Ramirez (Eds), School and the Culturally Diverse Exceptional
Student: Promising Practices and Future Directions, pp. 6 19. Reston, VA: ERIC Clearinghouse
on Handicapped and Gifted Children.
Ben-Dror, I., Frost, R. and Bentin, S. (1995) Orthographic representation and phonemic
segmentation in skilled readers: a cross-language comparison. Psychological Science, 6,
176181.
Bruck, M. (1990) Word recognition skills of adults with childhood diagnoses of dyslexia.
Developmental Psychology, 26, 439454.
Campione, J.C. (1989) Assisted assessment: a taxonomy of approaches and an outline of
strengths and weaknesses. Journal of Learning Disabilities, 22, 151 165.
Caravolas, M. and Bruck, M. (1993) The effect of oral and written language input on
childrens phonological awareness: a cross-linguistic study. Journal of Experimental Child
Psychology, 55, 130.
Carver, R.P. (1993) Merging the simple view of reading with rauding theory. Journal of
Reading Behavior, 4, 439454.
Carver, R.P., (1997) Reading for one second, one minute, or one year from the perspective
of rauding theory. Scientific Studies of Reading, 1, 345.
Chall, J.S. (1989) From language to reading and reading to language. In B. Ackerman
Stewart (Ed.), Partnerships in Education: toward a Literate America (Asha Report No 17), pp.
28 33. Rockville, MD: American Speech-Language Hearing Association.
Chall, J.S. (1996) Stages of Reading Development, 2nd edn. Orlando, FL: Harcourt Brace
College Publishers.
Chitiri, H.F., Sun, Y., Willows, D.M. and Taylor, I. (1992) Word recognition in second
language reading. In R.J. Harris (Ed.), Cognitive Processing in Bilinguals. Amsterdam:
North-Holland/Elsevier Science.
Cisero, C.A. and Royer, J.M. (1995) The development and cross-language transfer of
Phonological awareness. Contemporary Educational Psychology, 20, 275 303.
Cole, E. (1996) Immigrant and refugee children and families: supporting a new road
traveled. In M. Luther (Ed.), Dynamic Assessment for Instruction: from Theory to Application,
pp. 3542. Toronto: Captus University Publications.
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

26

E. Geva

Comeau, L., Cormier, P., Grandmaison, E. and Lacroix, D. (1999) A longitudinal study of
phonological processing skills in children learning to read in a second language. Journal of
Educational Psychology, 91, 2943.
Cossu, G., Shankweiler, D., Liberman, I.Y., Katz, L. and Tola, T. (1988) Awareness of
phonological segments and reading ability in Italian children. Applied Psycholinguistics, 9,
1 16.
Cummins, J. (1984) Bilingualism and Special Education: Issues in Assessment and Pedagogy.
Cleveland, England: Multilingual Matters.
Cummins, J. (1989) A theoretical framework for bilingual special education. Exceptional
Children, 56, 111119.
Dale, P.S., Crain-Thoreson, C. and Robinson, N.M. (1995) Linguistic precocity and the
development of reading: the role of extralinguistic factors. Applied Psycholinguistics, 16,
173187.
Dao, M. (1991) Designing assessment procedures for educationally at risk Southeast
Asian-American students. Journal of Learning Disabilities, 24, 594 601.
Dickinson, D.K. and Snow, C.E. (1987) Interrelationships among prereading and oral
language skills in kindergartners from two social classes. Early Childhood Research Quarterly,
2, 125.
Dunn, L.M. and Dunn, L.M. (1981) Peabody Picture Vocabulary Test Revised. Circle Pines,
MN: American Guidance Center.
Duran, R.P. (1989) Assessment and instruction of at-risk Hispanic students. Exceptional
Children, 56, 154158.
Durgunoglu, A.Y., Nagy, W.E. and Hancin-Bhatt, B.J. (1993) Cross-language transfer of
phonological awareness. Journal of Educational Psychology, 85, 453 465.
Elbro, C., Borstrom, I. and Petersen, D.K. (1998) Predicting dyslexia from kindergarten: the
importance of distinctness of phonological representations of lexical items. Reading Research
Quarterly, 33, 3660.
Feuerstein, R., Miller, R., Rand, R. and Jensen, M.R. (1981) Can evolving techniques better
measure cognitive change? The Journal of Special Education, 15, 201 219.
Figueroa, R.A. (1989) Psychological testing of linguistic-minority students: knowledge gaps
and regulations. Exceptional Children, 56, 145152.
Gavillan-Torres, E. (1983) Issues of assessment of limited English proficiency students and
of truly disabled in the United States. In N. Miller (Ed.), Bilingualism and Language Disability,
pp. 131153. San Diego, CA: College Hill Press, Inc.
Geva, E. Barsky, A. and Westernoff, F. (2000) Developing a framework for interprofessional
and diversity-informed practice. In E. Geva, A. Barsky and F. Westernoff (Eds),
Interprofessional Practice with Diverse Populations: Cases in Point. Westport, CT: Greenwood, in
press.
Geva, E. and Clifton, S. (1993). The development of first and second language reading skills
in early French immersion. Canadian Modern Language Review, 50, 646 667.
Geva, E. and Petrulius-Wright, J. (2000) The role of oral language proficiency in the reading
development of L1 and L2 primary level children. Under review.
Geva, E. and Ryan, E.B. (1993) Linguistic and memory correlates of academic skills in first
and second languages. Language Learning, 43, 542.
Geva, E. and Siegel, L.S. (2000) Orthographic and cognitive factors in the concurrent
development of basic reading skills in two languages. Reading and Writing, 1 31.
Geva, E. and Wade-Woolley, L. (1998) Component processes in becoming English Hebrew
biliterate. In A. Durgunoglu and L. Verhoeven (Eds), Acquisition of Literacy in a Multilingual
Context: a Cross Cultural Perspective, pp. 85110. Hillsdale, NJ: Erlbaum.
Geva, E., Wade-Woolley, L. and Shany, M. (1993) The concurrent development of spelling
and decoding in different orthographies. Journal of Reading Behavior, 25, 383 406.
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

27

Assessment of Reading Disabilities in L2 Children

Geva, E., Wade-Woolley, L. and Shany, M. (1997) The development of reading efficiency in
first and second language. Scientific Studies of Reading, 1, 119 144.
Geva, E., Yaghoubzadeh, Z. and Schuster, B. (1999) A not so simple view of ESL reading.
Paper presented at the Annual Meeting of the Society for the Scientific Study of Reading,
Montreal, April.
Gholamain, M. and Geva, E. (1999) The concurrent development of word recognition skills
in English and Farsi. Language Learning, 49, 183218.
Hoover, W.A. and Gough, P.B. (1990) The simple view of reading. Reading and Writing, 2,
127160.
Hu, C. and Catts, H.W. (1998) The role of phonological processing in early reading ability:
what we can learn from Chinese. Scientific Studies of Reading, 2, 55 80.
Jastak, J.R. and Jastak, S.R. (1984) Wide Range Achievement Test Revised. Wilmington, DE:
Guidance Associates.
Johnson, J.S. and Newport, E.L. (1989) Critical period effects in second language learning:
the influence of maturational state on the acquisition of English-as-a-second-language.
Cognitive Psychology, 21, 6069.
Lewis, J. (1989) Innovative approaches in assessment. In R.J. Samuda and S.L. Kong (Eds),
Assessment and Placement of Minority Students. Toronto: C.J. Hogrefe.
Limbos, M. and Geva, E. (1999) Accuracy of teacher assessments of ESL children at-risk for
reading disability: a two year follow-up study. Paper presented at the Annual Meeting of the
Canadian Psychological Association, Halifax, May.
Merbaum, C. and Geva, E. (1998) The relationship between listening and reading
comprehension in L1 and L2 Grade one children. Presented at the Annual Meeting of the
National Reading Conference, Austin Texas, December.
Naglieri, J. (1989) Matrix Analogies Test. New York: The Psychological Corporation for
Harcourt Brace Jovanovich.
Naslund, J.C. and Schneider, W. (1991) Longitudinal effects of verbal ability, memory
capacity and Phonological awareness on reading performance. The European Journal of
Psychology of Education, 6, 375392.
Ogbu, J.U. (1978) Minority Education and Caste. New York: Academic Press.
Oller, J.W. and Damico, J.S.C. (1990) Theoretical considerations in the assessment of LEP
students. In E.V. Hamayan and J.S. Damico (Eds), Limiting Bias in the Assessment of Bilingual
Students, pp. 77110. Austin, TX: Pro-Ed.
Oney, B. and Durgunoglu, A. (1997) Beginning to read in Turkish: a phonologically
transparent orthography. Applied Psycholinguistics, 18, 1 15.
Ortiz, A.A. and Yates, J.R. (1989) Characteristics of learning disabled, mentally retarded, and
speech-language handicapped Hispanic students at initial evaluation and reevaluation. In
A.A. Ortiz and B.A. Ramierez (Eds), School and the Culturally Di6erse Exceptional Student. pp.
51 62. Reston, VA: ERIC Clearinghouse on Handicapped and Gifted Children.
Pennington, B.F., Van Orden, G.C., Smith, S.D., Green, P.A. and Haith, M.M. (1990)
Phonological processing skills and deficits in adult dyslexics. Child Development, 61,
17531778.
Rack, J.P., Snowling, M.J. and Olson, R.K. (1992) The nonword reading deficit in
developmental dyslexia: a review. Reading Research Quarterly, 27, 28 53.
Ricard, R.J. and Snow, C.E. (1990) Language use in and out of context. Journal of Applied
Developmental Psychology, 11, 251266.
Rupley, W.H. and Willson, V.L. (1998) Exploration of the developmental components
contributing to elementary school childrens reading comprehension. Scientific Studies of
Reading, 2, 143158.
Schiff-Myers, N.B. (1992) Considering arrested language development and language loss in
the assessment of second language learners. Language, Speech and Hearing Services in Schools,
23, 2833.
Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)

28

E. Geva

Schneider, W. and Naslund, J.C. (1993) The impact of early metalinguistic competencies and
memory capacity on reading and spelling in elementary school: results of the Munich
longitudinal study of the genesis of individual competencies (LOGIC). European Journal of
Psychology of Education, 8, 273287.
Schuster, B. and Geva, E. (1999) Reading efficiency in L1 and ESL children: the role of oral
proficiency, word recognition, and orthographic knowledge. Presented at the Annual Meeting
of the Society for the Scientific Study of Reading, Montreal, April.
Share, D., Jorm, A., Maclean, R. and Matthews, R. (1984) Sources of individual differences in
reading acquisition. Journal of Educational Psychology, 76, 1309 1324.
Siegel, L.S. (1993) The development of reading. In H.W. Reese (Ed.), Advances in Child
Behaviour and Development, Vol. 24, pp. 6397. San Diego, CA: Academic Press.
Snow, C.E., Brows, M.S. and Griffin, P. (Eds) (1998) Preventing Reading Difficulties in Young
Children. Washington, DC: National Academy Press.
Snow, C.E., Cancini, H., Gonzalez, P. and Shriberg, E. (1989) Giving formal definitions: an
oral language correlate of school literacy. In D. Bloom (Ed.), Classrooms and Literacy, pp.
233249. Norwood, NJ: Ablex.
Stanovich, K.E. (1992) Speculations on the causes and consequences of individual differences
in early reading acquisition. In P.B. Gough, L.C. Ehri and R. Treiman (Eds), Reading
acquisition. Hillsdale, NJ: Erlbaum.
Statistics Canada Census (1996) Canadian Statistics: Immigrant Population. Available:
http://www.statscan.ca [1996?].
Vandervelden, M.C. and Siegel, L.S. (1995) Phonological recoding and phoneme awareness
in early literacy: a developmental approach. Reading Research Quarterly, 30, 854 875.
Verhoeven, L.T. (1990) Acquisition of reading in a second language. Reading Research
Quarterly, 25, 90114.
Wade-Woolley, L. and Geva, E. (2000) Processing novel phonemic contrasts in the
acquisition of L2 word reading. Scientific Studies of Reading, in press.
Wagner, R. and Torgeson, J. (1987). The nature of phonological processing and its causal
role in the acquisition of reading skills. Psychological Bulletin, 101, 192 212.
Wimmer, H. and Hummer, P. (1990) How German-speaking first-graders read and spell:
doubts on the importance of the logographic stage. Applied Psycholinguistics, 11, 349 368.
Woodcock, R.W. (1987) Woodcock Reading Mastery Test. Circle Pines, MN: American
Guidance Service.

Copyright 2000 John Wiley & Sons, Ltd.

Dyslexia 6: 13 28 (2000)