Content uploaded by Gita Martohardjono
Author content
All content in this area was uploaded by Gita Martohardjono on May 21, 2020
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rbeb20
International Journal of Bilingual Education and
Bilingualism
ISSN: 1367-0050 (Print) 1747-7522 (Online) Journal homepage: https://www.tandfonline.com/loi/rbeb20
Impact of home language reading skills’ on L2
reading comprehension: a study of newcomer
emergent bilinguals
Leigh Garrison-Fletcher, Gita Martohardjono & Martin Chodorow
To cite this article: Leigh Garrison-Fletcher, Gita Martohardjono & Martin Chodorow (2019):
Impact of home language reading skills’ on L2 reading comprehension: a study of newcomer
emergent bilinguals, International Journal of Bilingual Education and Bilingualism, DOI:
10.1080/13670050.2019.1671307
To link to this article: https://doi.org/10.1080/13670050.2019.1671307
Published online: 20 Oct 2019.
Submit your article to this journal
Article views: 112
View related articles
View Crossmark data
Impact of home language reading skills’on L2 reading
comprehension: a study of newcomer emergent bilinguals
Leigh Garrison-Fletcher
a
, Gita Martohardjono
b
and Martin Chodorow
c
a
Department of Education and Language Acquisition, LaGuardia Community College of the City University of
New York, New York City, USA;
b
Linguistics Department, The Graduate Center of the City University of New York,
New York City, USA;
c
Department of Psychology, Hunter College of the City University of New York, New York City,
USA
ABSTRACT
Previous research on the development of second language (L2) reading
comprehension among adolescents and adults has looked at the relative
importance of students’second language (L2) linguistic skills and home
language (L1) reading comprehension skills in the development of L2
reading comprehension, with many studies finding a more important
role for L2 language proficiency. However, the previous research has
been done largely in foreign language settings with students who have
age-appropriate reading skills in the L1. We extend the question of the
relative importance of L1 reading skills and L2 language proficiency to
L2 reading comprehension to Spanish-speaking adolescent newcomer
emergent bilinguals in New York City with low levels of L1 reading, as
research on these students is lacking and little is known about their
development of L2 academic skills. We find that for this population L1
reading comprehension plays a significant role in L2 reading
comprehension equal to, if not greater than, L2 language proficiency.
These findings suggest the need to include continued support of L1
reading skills in order for reading in the L2 to develop, at least for
newcomer students with low literacy skills.
ARTICLE HISTORY
Received 8 August 2018
Accepted 8 September 2019
KEYWORDS
Language ability; biliteracy;
ELL; language proficiency
Introduction
Adolescent newcomers who enter school in the U.S. as English Language Learners (ELLs) must quickly
and efficiently develop academic skills in a second or additional language (L2), English. Studies that
focus on a comparison of L2 learners and their monolingual peers note a difference in performance
on higher-level skills that are essential to school success, such as reading comprehension (Melby-
Lervåg and Lervåg 2014). According to Lesaux et al. (2006)‘[e]xisting large-scale data sets on the
school achievement of language-minority students in the United States and abroad suggest that
comprehension is a significant area of difficulty for these learners’(100). In the 2006 report of the
National Literacy Panel, August and Shanahan note that language-minority students rarely approach
the same levels of proficiency in reading comprehension as their monolingual peers. The attainment
of adequate levels of academic literacy in the L2 clearly requires substantial efforts on the part of ELLs,
yet relatively little research has been conducted to better understand how greater gains can be made
in this critical skill area.
Age is another important factor in the successful attainment of academic literacy. In a paper sum-
marizing the avenues for future research in second language literacy acquisition, Snow notes the
© 2019 Informa UK Limited, trading as Taylor & Francis Group
CONTACT Leigh Garrison-Fletcher lgarrisonfletcher@lagcc.cuny.edu
INTERNATIONAL JOURNAL OF BILINGUAL EDUCATION AND BILINGUALISM
https://doi.org/10.1080/13670050.2019.1671307
need for ‘a much greater focus on postprimary English-language learners (ELLs).’She goes on to say
that ‘practitioners are desperate for information about how best to serve older immigrant students’
(Snow 2006, 642). This information is important because these students have traditionally been
underserved, have a disproportionately high attrition rate, and need research-based services to
help them succeed. Adolescent ELLs have less time to develop L2 academic skills than young chil-
dren, as they enter the school system in the later grades and must learn English while also learning
the academic content needed to graduate from high school (Klein and Martohardjono 2009). Perhaps
because of this, the graduation rates for ELLs in U.S. schools are very low; about 23% of ELLs age 16–
24 are either not enrolled in school or do not have a high school diploma or the equivalent (Morse
2005). Research on adolescent newcomers is lacking and educators are calling for more information
on how to meet the educational needs of these students (Freeman and Freeman 2002; Francis et al.
2006; Short and Fitzsimmons 2007). Research into how adolescent ELLs develop the critical academic
skills required for high school graduation, such as reading comprehension, is important in order to
better inform instruction and lead to higher levels of academic achievement for these students.
It may be a commonplace observation that the development of reading in a second language
requires adequate levels of language proficiency in that same language. However, research suggests
that existing reading levels in the ELL’s home language (L1) also contribute to the development of
reading in the second language (L2) (Cummins 1981,1991; Verhoeven 1990; Bossers 1991; Royer
and Carlo 1991; Lee and Schallert 1997; Payne, Kalibatseva, and Jungers 2009; Sparks et al. 2012;
Mirza, Gottardo, and Chen 2017). Thus, in his seminal 1984 article, Alderson asked, ‘Is reading in a
foreign language a reading problem or a language problem
1
?’Shaped by Alderson’s question,
much of the research on L2 reading comprehension has looked at the relative importance of L1
reading ability and L2 language proficiency to L2 reading comprehension. And among this research
on adolescents, the consensus has been that L2 language skills play the most important role in L2
reading comprehension, with L2 vocabulary knowledge generally thought to be the most important
factor (Bernhardt and Kamil 1995; Brisbois 1995; Lee and Schallert 1997; Yamashita 1999). It should be
noted, however, that in all of these studies, participants had age-appropriate levels of reading ability.
That is, these studies investigate the growth of reading ability in a second language when reading
ability in the first language is already developed. It could be argued, that since the participants in
these studies are already good readers in their L1, difficulty in L2 reading is unlikely to come from
L1 reading abilities. What the L2 reading literature has not addressed, however, is a related and
equally critical question: In a population of L2 learners whose existing literacy skills are low, are
delays in academic literacy in the language being learned primarily due to poor L2 language profi-
ciency or poor L1 reading ability? This is the question we will address in our study.
Background: interdependence and threshold hypotheses
Much of the research on L2 reading among adolescents has focused on the transfer of home
language (L1) reading skills to the L2, with the research clearly showing the cross-linguistic transfer
of reading skills (see, for example, Cummins 1981,1991; Verhoeven 1990; Royer and Carlo 1991; Lee
and Schallert 1997; Payne, Kalibatseva, and Jungers 2009; Sparks et al. 2012; Mirza, Gottardo, and
Chen 2017). This has been termed the Interdependence Hypothesis (IH), first coined in Cummins
(1979). Much of this work has been conducted on children aged 6–10, as reported in Melby-
Lervåg and Lervåg (2011), for example. This study is a meta-analysis looking at cross-linguistic transfer
in L2 reading, but as the authors note, there were not enough studies on older learners to include in
the meta-analysis of cross-linguistic transfer of L1–L2 reading.
Cummins’interdependence hypothesis (1979) proposes that L1 reading skills immediately transfer
to the L2, due to a common underlying proficiency shared by the two languages. This would predict a
strong relationship between L1 and L2 reading ability, at all levels of L2 language proficiency. Support
for the IH comes from a longitudinal study with eighth through tenth grade L1 Dutch speakers learn-
ing English (L2) by van Gelderen et al. (2007) which found that reading comprehension skills in Dutch
2L. GARRISON-FLETCHER ET AL.
were the stronger predictor of reading comprehension in English. They included assessments of both
L1 and L2 reading for three years, and found that L1 reading comprehension was a stronger predictor
of L2 reading comprehension than L2 vocabulary or grammar. In a study with 49 fifth and sixth grade
students enrolled in bilingual programs, Royer and Carlo (1991) found that the student’s L1 (Spanish)
reading ability in the fifth grade was the strongest predictor of the student’s L2 (English) reading
ability in the sixth grade, thus also providing support for Cummins’(1979,1981) interdependence
hypothesis which posits that students who have developed literacy and academic concepts in one
language can easily transfer these to the other language and that this transferability of cognitive abil-
ities is due to the common underlying proficiency shared by the two languages.
Another hypothesis arising from this literature is the threshold hypothesis
2
(Clarke 1980), also
referred to as the short-circuit hypothesis. The threshold hypothesis posits that for L1 literacy skills
to transfer to the L2, a learner must first reach a certain level of L2 language proficiency.
3
As Alderson
(1984) noted, under the threshold hypothesis, problems at the beginning stages of L2 literacy acqui-
sition would primarily be due to language, as the low level of L2 language proficiency blocks the use
of L1 reading ability. However, once a learner acquires enough L2 language knowledge and reaches a
threshold, existing L1 reading ability should become a strong predictor of L2 reading. Research by
Bossers (1991) and Pichette, Segalowitz, and Connors (2003) has supported the threshold hypothesis,
finding that at the beginning stages of L2 acquisition, L2 proficiency played a more important role in
determining L2 reading comprehension. But in students with more advanced L2 proficiency, the L1
reading ability played a more important role. In a year-long longitudinal study, Pichette, Segalowitz,
and Connors (2003) looked at 52 Bosnians learning French as an L2 and found that only L2 knowledge
significantly predicted L2 reading ability at the beginning of the study but once students developed
more L2 knowledge, both L2 knowledge and L1 reading ability emerged as significant factors. These
results suggest that L1 reading skills began to transfer to L2 reading as learners’L2 knowledge
improved. The authors claim that their results supported the threshold hypothesis as learners
were only able to transfer their reading skills once the L2 proficiency reached a certain threshold level.
Research contexts
The studies reviewed above test the interdependence and threshold hypotheses by focusing on stu-
dents with varying levels of L2 language proficiency; however, participants with varying levels of L1
reading ability have not been systematically included. In order to have a better understanding of the
interaction between L1 and L2 reading comprehension, we propose that it is necessary to include
participants with both low and high literacy in the L1. The need for including students with a
range of L1 reading ability in the research program on L2 reading has been noted by other research-
ers, but to our knowledge the one study that aimed to do this (Lee and Schallert 1997), was still
restricted to relatively good L1 readers. Restricting the investigation to students with age-appropriate
L1 reading ability, important as it may be, leaves out of the equation a sizeable but important student
population that has not reached expected literacy levels in their L1, due to various factors in their
educational history.
Another issue related to the context of participants in reading research is raised by Carrell (1991)
who finds conflicting results with respect to the relative importance of reading abilities in the first
language compared to proficiency in the second language. Looking at two groups of U.S. college stu-
dents, one speaking Spanish as their L1 and learning English as a second language; the other group
speaking English as their L1 language and learning Spanish as a foreign language, Carrell found that
among the L1 English speakers, Spanish language proficiency was the stronger predictor of Spanish
reading skills. Different results were found for the L1 Spanish-speaking students, however. For this
group level of reading in the L1/Spanish was the stronger predictor of L2/English reading than L2/
English proficiency. Carrell posited that the differing results between the two groups could be
explained due to the different learning situations for the groups (L2 vs. foreign language learning).
This is an important point to consider, and is of particular relevance to our study. Much of the
INTERNATIONAL JOURNAL OF BILINGUAL EDUCATION AND BILINGUALISM 3
previous research on L2 reading among adolescents has been conducted in foreign-language learn-
ing situations, with little done on second language learners in U.S. high schools. We have already
mentioned the relatively high levels of reading abilities in the first language, typically shown in
foreign language learning situations, compared to immigrant L2 environments. And there are
other, potentially important socioeconomic status (SES) and educational differences between
foreign language learners and immigrant second language learners that also need to be considered
in research on the acquisition of L2 reading comprehension.
Current study
In order to get a clearer picture of the interaction between L1 and L2 reading, we re-examine Alder-
son’s question in a population of learners who have not been systematically included in existing
reading research: newly arrived immigrants in U.S. high schools, or adolescent newcomers. Specifi-
cally, we focus on Spanish-speaking adolescents who entered high school in New York City upon
first arrival in the United States. This population differs from those previously studied in one very
important way –they have a range of L1 literacy, with many being underschooled or having
missed years of formal schooling in their home countries. In New York City schools, there is a
subset of adolescent newcomers labeled as Students with Interrupted Formal Education (SIFE). By
definition, SIFE have missed at least two years of schooling prior to beginning school in the U.S. In
the 2013–14 school year, 9.2% of the English Language Learner (ELL) population in New York City
schools were SIFE (New York City DOE 2015). In a population of adolescent immigrant high school
students, we ask the following:
Research Question 1: Which of the three predictor variables –L1 reading comprehension, L2 vocabulary, or L2
syntax –plays the strongest role in L2 reading comprehension?
Research Question 2: Does the relationship of the predictor variables with L2 reading comprehension change
when using a separate measure of L1 reading comprehension?
Research Question 3: Do reading skills in the L1 or language skills in the L2 play a stronger role in L2 reading
comprehension?
Methodology
Participants
The participants in the study were 62 Spanish-speaking adolescent newcomers attending one
New York City public high school.
4
The majority (59%) of the students were enrolled in the ninth
grade at the time of the study while the remaining were enrolled in the tenth grade. The students
had been placed into grades based on the length of time in the United States upon enrollment;
the ninth graders had been in school in the United States for one to two months and the tenth
graders had been in school in the United States for fourteen months at the time of the study. The
students ranged in age from 15 to 20 with an average age of 17.5.
5
The majority (85%) of the students
were from the Dominican Republic; 7% were from Honduras, 5% were from Ecuador, and 1.5% each
were from Mexico and Colombia.
6
Since we wanted to have a group of L2 learners with a range of L1 literacy skills, we asked school
administrators to select students for the study from two different groups of newcomers: English
Language Learners (ELLs) and Students with Interrupted Formal Education (SIFE). Twenty of the par-
ticipants were labeled SIFE; the remaining forty-two were ELLs. The New York City Department of Edu-
cation (NYCDOE) designates students as SIFE if they had missed at least two years of formal schooling
prior to arrival in the U.S. Due to missed schooling, SIFE have low levels of L1 academic skills (Klein and
Martohardjono 2006). Newly arrived ELLs who were not labeled SIFE were also selected for the study;
as these students had consistently attended school in their home countries, they were expected to
4L. GARRISON-FLETCHER ET AL.
have higher levels of L1 literacy than SIFE. These were the initial selection criteria for participation in
the study, but all students were given two tests of L1 reading comprehension which served as inde-
pendent measures of L1 reading ability. This was an important step, as research on SIFE and ELLs has
suggested that the designation based on missed schooling may not be sufficient to predict age-
appropriate literacy skills. ELLs, who can also be under-schooled for their age, but who are not cate-
gorized as SIFE as they do not have interruptions in their schooling can also exhibit low literacy in the
home language when appropriately tested (Klein and Martohardjono 2006).
Measurements and testing procedures
Since we were interested in the development of critical reading skills in the second language, we
made our outcome or dependent variable a measure of academic reading comprehension in the
L2 English. Following findings in the literature (Brisbois 1995; Yamashita 1999; Shiotsu and Weir
2007), we reasoned that the potential L2 skills contributing to this dependent variable are L2 voca-
bulary knowledge and L2 syntactic comprehension. We juxtaposed these two L2 skills to L1 (Spanish)
reading comprehension, as a potentially contributing factor. Because reading comprehension is
necessarily a composite of several skills, measurable in different ways, we further distinguished
between two types of L1 reading comprehension measures: general and academic. The following
list contains all variables used in the study, and the corresponding assessments:
Dependent Variable
.L2 (English) academic reading comprehension –English Academic Language and Literacy Diagnos-
tic (ALLD) Reading Comprehension subtest
Predictor Variables
.L1 (Spanish) reading comprehension
(a) Academic reading comprehension –Spanish ALLD Reading Comprehension subtest
(b) General reading comprehension –Spanish Evaluación Diagnóstica de la Comprensión Lectora
(EDICOLE)
.L2 (English) academic vocabulary –English ALLD Academic Vocabulary subtest
.L2 (English) syntactic comprehension –RISLUS English Syntax Test
Control Variable
.L1 (Spanish) syntactic comprehension –RISLUS Spanish Syntax Test
In the following sections, we describe each of the assessments listed above in more detail. Table 1
shows the assessments used in the study to measure reading comprehension and those used to
measure language proficiency.
Measures of L1 (English) and L2 (Spanish) academic reading comprehension
We use the term ‘academic reading comprehension’to refer to the reading of texts typically encoun-
tered in an academic setting, containing some academic vocabulary and complex syntactic
Table 1. Assessments used in the study.
Assessments Testing Reading Comprehension Assessments Testing Language Skills
English ALLD Academic Reading Comprehension Subtest English ALLD Academic Vocabulary Subtest
Spanish ALLD Academic Reading Comprehension Subtest English RISLUS Syntax Test
Spanish EDICOLE Spanish RISLUS Syntax Test
INTERNATIONAL JOURNAL OF BILINGUAL EDUCATION AND BILINGUALISM 5
structures. The Reading Comprehension section of the Academic Language and Literacy Diagnostic
(ALLD) was the assessment used to measure this skill in both Spanish (L1) and English (L2). The
test was designed for high school immigrant students in New York City, and is a cumulative assess-
ment that includes passages and test items from the second through the eleventh grade. There are a
total of 38 items in both the English and Spanish versions, with three items from second grade level,
four items from third grade level, six items from fourth grade level and five items each in fifth through
seventh, ninth and eleventh grade levels. The test format follows that of typical academic standar-
dized reading comprehension tasks, with passages followed by multiple-choice questions. The ques-
tions assess basic understanding and higher-level thinking skills such as critical analysis, strategies,
and interpretation.
Measure of L1 (Spanish) general reading comprehension
Research has suggested that tests of reading comprehension have differing relationships between
one another and between various predictor variables (Cutting and Scarborough 2006;Melby-
Lervåg and Lervåg 2014). The Spanish and English ALLD described above are parallel test versions
targeting academic reading. In order to control for a potentially inflated relationship between L1
and L2 reading comprehension tasks of similar types, we also measured what we call ‘general
reading comprehension.’For this, we chose the Evaluación Diagnóstica de la Comprensión
Lectora (EDICOLE).TheEDICOLE is a novel reading comprehension test designed by the Center
for Applied Linguistics.
7
The passages in the EDICOLE use ‘simple, regular, high-frequency
words’and minimize ‘the impact of background knowledge …by limiting topics to very familiar
ones and by introducing nonce words for novel concepts’(Francis et al. 2006, 304). In spite of
its simplicity, the test assesses text inferencing and the ability to integrate background knowledge
with text information. The EDICOLE consists of three passages each followed by 18 questions, for a
total of 54 questions. Each passage is followed by two items in the format of a question (e.g.
‘Which animal has hair?’). The student is given three choices, one of which is ‘Idon’tknow.’The
remaining sixteen items following each passage are in the format of a true/false question; the
student reads a statement (e.g. ‘Ana’s animals like to play in the living room.’) and must choose
‘yes,’‘no,’or ‘Idon’tknow.’
Measure of L2 (English) academic vocabulary
The academic vocabulary assessment is a subtest of the ALLD and includes items from the third
through seventh, ninth and eleventh grade levels. This assessment is multiple-choice and is com-
posed of three sections: synonyms, multiple meaning words, and context clues. There are a total
of 39 vocabulary items, with grade levels three, four and six each having five items and grade
levels five, seven, nine and eleven each having six items. The synonyms section requires a student
to choose a synonym for a printed word. There are a total of twelve synonyms items. The multiple
meaning words section provides a target sentence with a target word underlined, and four sub-
sequent sentences all containing the target word used in four different ways. The student must
select the sentence in which the word is used in the same way as in the target sentence. There
are a total of thirteen multiple meaning words items. The context clues section provides a target sen-
tence with an underlined word; the target sentence provides a clue to the meaning of the underlined
word. The student must choose the meaning of the underlined word. There are a total of fourteen
context clues items.
Measure of L2 (English) syntactic comprehension
The assessment of syntactic comprehension (RISLUS Syntax Test) was developed by the Research Insti-
tute for the Study of Language in Urban Society (RISLUS) and tests the comprehension of complex
syntactic structures through listening comprehension. This assessment of syntactic comprehension
is designed to evaluate mastery of complex sentence structure independently from vocabulary.
Therefore, the sentences use very simple vocabulary. The test was administered in a group setting
6L. GARRISON-FLETCHER ET AL.
and is composed of 44 items. Each student had a test booklet with each item on a separate page. The
student saw three pictures for each item. The test administrator said a sentence for each item and
repeated the sentence once. The student was then asked to choose the picture that corresponds
to the sentence. The structures tested are coordination, relative (or adjective) clauses, temporal
adverbial clauses, and subjectless subordinate clauses. See examples of each structure below:
(a) Subject coordination: The bear and the dog chase the cat.
(b) Relative clause: The bear who touches the dog dances.
(c) Temporal adverbial clauses: After jumping, the dog kicks the cat.
(d) Subjectless subordinate clauses: The dog tells the bear to sit, and he does.
Measure of L1 (Spanish) syntactic comprehension
The RISLUS Syntax Test has an English and a Spanish version, both with 44 items administered in the
same way. The Spanish version was used in this study as a control variable, to ensure that all of the
participants had typical home language development. As the syntactic structures tested are mile-
stones in first language development, we would expect all students with typical language develop-
ment to score very well on this assessment in the home language.
Hypotheses and predictions
We hypothesized that if L2 reading comprehension is primarily guided by language proficiency in the
L2, then the L2/English measures we used (the ALLD academic vocabulary test and the RISLUS Syntax
Test in English) should show higher contributions to English (L2) reading comprehension. Conversely,
if reading comprehension abilities already established in the L1 are more important contributors to
reading comprehension in the L2, we should see higher contributions of the Spanish ALLD Reading
Comprehension subtest and the EDICOLE to English reading comprehension. Furthermore, as the par-
ticipants are in the beginning stages of English language acquisition, having been in the U.S. for less
than two years, if the threshold hypothesis holds, we would expect to find L2 language proficiency
the stronger predictor of L2 reading ability. If the IH holds, we would expect to find L1 reading com-
prehension to play a stronger role in L2 reading comprehension from the beginning stages of L2
development.
Results
Descriptives
Table 2 shows the results from the Spanish assessments and Table 3 shows the results from the
English assessments.
Recall that the ALLD in both Spanish and English is a cumulative assessment that includes test
items from the second grade level (in the case of academic reading comprehension) or the third
grade level (in the case of academic vocabulary) up to the eleventh grade level. Thus, the results
from the ALLD are reported not only as mean percent correct, but also as average grade level.
Note that the standard deviations for all variables except Spanish syntactic comprehension range
Table 2. Mean percent (SD) and grade level (SD) for the Spanish assessments (N= 62).
Spanish General RC (EDICOLE) Spanish Academic RC (ALLD) Spanish Syntactic Comprehension
Mean (SD) 67%
(13%)
52%
(15%)
90%
(6%)
Grade Level (SD) –5.8
(2.5)
–
INTERNATIONAL JOURNAL OF BILINGUAL EDUCATION AND BILINGUALISM 7
from 12% to 16%, confirming that the population tested here indeed has a range in both the L1 and
L2 skills tested. Furthermore, the scores on the Spanish syntactic comprehension test indicate that
these students have typical home language development.
Correlations
Pearson product-moment correlation coefficients were calculated in order to measure the relation-
ship between the variables. See Table 4 for the results.
The high correlations amongst all variables seen in this table were expected, as the same skills are
being assessed in two languages and they are all related to reading. However, the main research
questions will be addressed using regression analyses in order to look at the relative importance
of the predictor variables on L2 reading comprehension. We present these in the next section.
Regression results
A number of regression analyses were run in order to address the research questions dealing with the
most important components involved in L2 academic reading comprehension among adolescent
learners. Below we report on results from regressions using the proportion correct values for all vari-
ables, relating them to our research questions.
Research Question 1: Which of the three predictor variables –L1 reading comprehension, L2 vocabulary, or L2
syntax –plays the strongest role in L2 reading comprehension?
In order to address this question, a hierarchical multiple regression analysis was performed, in which
L2 academic reading comprehension (English ALLD RC subtest) was the dependent variable and the
predictor variables included L1 academic reading comprehension (Spanish ALLD RC subtest), L2 aca-
demic vocabulary (English ALLD Vocabulary subtest) and syntax (English RISLUS Syntax Test). See
Table 5 for the results.
From these results, it appears that L1 academic reading comprehension is the strongest contribu-
tor to L2 academic reading comprehension, above L2 academic vocabulary and L2 syntax, although
L2 academic vocabulary is also a significant contributor.
Research Question 2: Does the relationship of the predictor variables with L2 reading comprehension change
when using a separate measure of L1 reading comprehension?
A similar regression was run, this time including the EDICOLE, the measure of L1 general reading com-
prehension. As mentioned above, looking at another test of L1 reading comprehension will give us an
Table 3. Mean percent (SD) and grade level (SD) for the English assessments (N= 62).
English Academic RC (ALLD) English Vocabulary (ALLD) English Syntactic Comprehension
Mean (SD) 35%
(12%)
38%
(16%)
77%
(15%)
Grade Level (SD) 3.1
(1.9)
4
(2.2)
–
Table 4. Correlation matrix (N= 62).
1. 2. 3. 5. 6.
1. English Academic Reading Comprehension (ALLD)–
2. Spanish General Reading Comprehension (EDICOLE) .450*** –
3. Spanish Academic Reading Comprehension (ALLD) .571*** .483*** –
5. English Academic Vocabulary (ALLD) .435*** .305* .370** –
6. English Syntax .397** .472*** .414** .386** –
***p< .001; **p< .01; *p< .05.
8L. GARRISON-FLETCHER ET AL.
independent measure and therefore more information as to the relative importance of reading skills
developed in the L1 and language skills in the L2 to the development of L2 reading. See Table 6 for
the results from this analysis.
Again, we see the strong role of L1 reading comprehension in the development of L2 reading
comprehension. Thus, with two different measures of L1 reading comprehension, we find that the
reading skills already gained in the L1 significantly contribute to the L2 reading comprehension in
the L2.
Research Question 3: Do reading skills in the L1 or language skills in the L2 play a stronger role in L2 reading
comprehension?
In order to address the central question of the study, we ran two other regression analyses in which
the outcome variable was L2 reading comprehension and the predictor variables of L1 reading com-
prehension and both the L2 language variables were entered in different orders. Tables 7 and 8show
the results. L1 academic reading comprehension was a stronger predictor than the combination of
the English language variables. It uniquely accounted for 14.6% of the variance in English reading
comprehension compared to only 7.1% accounted for by English vocabulary and syntax.
8
Results based on demographic data
The participants in the study differed in terms of many demographic factors, including: country of
origin, length of time enrolled in school in the United States, age, and whether or not the student
is labeled SIFE due to interrupted schooling in the home country. In order to determine if any of
these factors played a role in a model of L2 reading comprehension, regression analyses were con-
ducted with L2 academic reading comprehension as the dependent variable and the demographic
variables as the predictors. None of the predictors was significant. This finding is not especially
Table 5. Regression analysis with L2 academic reading comprehension as dependent variable, including predictor variables of L1
academic reading comprehension, L2 academic vocabulary, L2 syntax (N= 62).
BSE Bβ
Step 1
L1 Academic Reading Comprehension .47 .09 .57***
Step 2
L1 Academic Reading Comprehension .39 .09 .48***
L2 Academic Vocabulary .20 .08 .26*
Step 3
L1 Academic Reading Comprehension .35 .09 .43***
L2 Academic Vocabulary .17 .09 .22*
L2 Syntax .11 .09 .13
***p< .001; *p< .05.
Note: R
2
= .33 for Step 1, ΔR
2
= .06 ( p< .05) for Step 2, ΔR
2
= .01 (ns; p= .26) for Step 3.
Table 6. Regression analysis with L2 academic reading comprehension as dependent variable, including predictor variables of L1
general reading comprehension (EDICOLE), L2 academic vocabulary, L2 syntax (N= 62).
BSE Bβ
Step 1
L1 General Reading Comprehension .44 .11 .45***
Step 2
L1 General Reading Comprehension .34 .11 .35**
L2 Academic Vocabulary .25 .09 .33**
Step 3
L1 General Reading Comprehension .28 .12 .29*
L2 Academic Vocabulary .22 .09 .29*
L2 Syntax .001 .001 .15
***p< .001; **p< .01; *p< .05.
Note: R
2
= .20 for Step 1, ΔR
2
= .10 ( p< .01) for Step 2.
INTERNATIONAL JOURNAL OF BILINGUAL EDUCATION AND BILINGUALISM 9
surprising. The majority of the students are from the Dominican Republic (85%), and all other
countries of origin are in South America. The limited number of students from any country other
than the Dominican Republic makes it difficult to look at the effect of country of origin. And as all
the students had been enrolled in U.S. schools for at most one year, the length of time in U.S.
schools also did not play a role in L2 reading comprehension. The age of the students also was
not a factor in how well they read in the L2.
Finally, the SIFE status of the students was not relevant to the L2 reading comprehension scores of
the students. Twenty of the participants were identified as SIFE by the school, while the others had
not met the criteria of having missed two years of schooling. This result, although perhaps unex-
pected, given the differential school labeling (SIFE/non-SIFE) is actually reflected in the results
reported in Table xx above, showing that the group as a whole showed comparable levels of L1
reading ability (as measured by the Spanish ALLD RC subtest). Thus, non-SIFE ELLs also had lower
than expected grade-level reading, similar to SIFE. These results point to the need to use objective
measures of L1 literacy, such as the Spanish ALLD, to gauge students’academic skills, rather than
using interrupted schooling as the only measure of classification. At the time of the study SIFE des-
ignation was based solely on an interview in the home language, and not an objective assessment of
L1 language or reading skills or educational records.
Discussion
The study presented here aimed to address the question of whether existing reading skills in the L1 or
linguistic knowledge of the L2 played the most important role in L2 reading comprehension among a
population of newcomer adolescent ELLs, with and without interrupted schooling, who speak
Spanish as their home language and have a range of academic skills in their L1. Based on one of
the measures we used, the Spanish ALLD, the students evidenced a range of L1 reading abilities,
with some students scoring at or below the second grade level and some students scoring at the ele-
venth grade level. One important point to note is that among our participants, the individual scores
Table 7. Regression analysis with L2 academic reading comprehension as dependent variable, and predictor variables of L2
academic vocabulary and L2 syntax entered in step 1, then predictor variable of L1 academic reading comprehension entered
in step 2 (N= 62).
BSE Bβ
Step 1
L2 Academic Vocabulary .25 .09 .33**
L2 Syntax .22 .1 .27*
Step 2
L2 Academic Vocabulary .17 .09 .22*
L2 Syntax .11 .09 .13
L1 Academic Reading Comprehension .35 .09 .43***
***p< .001; **p< .01; *p< .05.
Note: R
2
= .25 for Step 1, ΔR
2
= .146 ( p< .001) for Step 2.
Table 8. Regression analysis with L2 academic reading comprehension as dependent variable, and predictor variable of L1
academic reading comprehension entered in step 1, then predictor variables of L2 academic vocabulary and L2 syntax entered
in step 2 (N= 62).
BSE Bβ
Step 1
L1 Academic Reading Comprehension .47 .09 .57***
Step 2
L1 Academic Reading Comprehension .35 .09 .43***
L2 Academic Vocabulary .17 .09 .22*
L2 Syntax .11 .09 .13
***p< .001; **p< .01; *p< .05.
Note: R
2
= .33 for Step 1, ΔR
2
= .07 ( p< .05) for Step 2.
10 L. GARRISON-FLETCHER ET AL.
are clustered near the low end of L1 reading ability; on average the students, who were in the ninth
and tenth grades according to age-based school placement, scored between the fifth and sixth grade
levels of reading comprehension in the home language. Thus, while the participants do have a range
of L1 literacy skills, there are fewer students who are scoring at grade level with most students
reading below the expected grade level in the L1.
Another important characteristic of the students in the study that is necessary to consider is that
they have low levels of L2 proficiency. As the students were all newly arrived immigrant students,
having been in the United States for less than two years at the time of the study, they had not
had the time to become advanced L2 learners. Thus, although the students have a range of L2 profi-
ciency (scoring between the third and eleventh grade level on L2 academic vocabulary and between
39% and 97% correct on L2 syntactic comprehension), they are again concentrated on the lower end
of the spectrum due to their limited time learning English, scoring on average at the fourth grade
level on L2 academic vocabulary and 77% correct on L2 syntactic comprehension.
Since the students in our study had low L2 proficiency, the threshold hypothesis would predict that
L2 language proficiency would have a stronger relationship with L2 reading comprehension than L1
reading comprehension. Our results, in fact, do not support this hypothesis, but rather evidence cross-
linguistic transfer of reading skills from the beginning stages of acquiring a second language. As men-
tioned earlier, previous research supporting the threshold hypothesis has only included students with
age-appropriate L1 reading ability. The findings here are notable in that for the group tested in our
study, we see the level of L1 reading comprehension being the strongest contributing factor to
level of L2 reading comprehension. Given that all our participants were at beginning stages of
second language acquisition, but varied in L1 reading ability, our results indicate that the threshold
hypothesis is modulated not only by L2 proficiency, but also by existing L1 reading ability.
The finding that L1 reading comprehension was a strong, significant predictor of L2 reading com-
prehension was consistent when L1 reading comprehension was measured with different assess-
ments. When looking at two separate measures, academic reading, as measured by the Spanish
ALLD, and general reading, as measured by the EDICOLE, we found that L1 reading comprehension
played a significant role in determining L2 reading comprehension. These results give strong support
for the idea that for students with lower levels of literacy in the L1, established L1 reading skills play a
critical role in learning to read in the L2. This is not to say that gaining L2 linguistic skills is not impor-
tant, but that existing reading abilities in the L1 cannot be neglected. Past research has not had
similar findings, most likely due to the fact that the students all had age-appropriate L1 reading
skills, and potentially due to the different sociolinguistic characteristics of the populations.
Our main research question as to the stronger predictors of L2 reading comprehension found that
for the particular group of students we tested, L1 reading comprehension made a unique contri-
bution to accounting for the variance in L2 English reading comprehension, a contribution that is
independent of L2 language skills. Existing reading skills appear to be as important as L2 linguistic
knowledge in the development of L2 reading comprehension among adolescents who are in the
beginning stages of L2 development. Thus, our study supports the interdependence hypothesis
and the notion that a learner’s L1 reading ability transfers to the L2: reading skills developed in
the L1 are available to the learner when he or she is learning to read in a second language.
Importantly, our study included participants with a broader range of L1 reading comprehension
than has been included before. It appears that for this population, existing reading ability in the
L1 is indeed a strong contributor to L2 reading comprehension, independent of L2 vocabulary.
Conclusion
Based on the study reported here we have a better picture of the components involved in L2 aca-
demic reading comprehension among adolescent newcomers who speak Spanish as their L1 and
are in the beginning stages of learning English. The results suggest that the existing reading skills
a student has in his or her home language are strong predictors of the student’s reading skills in
INTERNATIONAL JOURNAL OF BILINGUAL EDUCATION AND BILINGUALISM 11
the second language, providing evidence that the reading skills a student has developed in the L1
contribute significantly to the development of reading skills in a second language. In line with pre-
vious research, L2 language proficiency and L1 reading comprehension both contributed to L2
reading. However, unlike in previous studies, L2 vocabulary did not contribute to L2 reading as
much as reading comprehension in the L1. This is our most notable finding.
The population of students included in the present study have not been included in the research
program on L2 reading comprehension which has primarily focused on students in foreign language
learning situations with expected grade-level reading skills in the L1. Thus, our results are particularly
important and speak to the education of newcomer adolescents in U.S. schools. We suggest that pre-
vious research has found L2 language skills to be the strongest predictor of L2 reading comprehen-
sion because those studies only included participants with age-appropriate levels of L1 reading
comprehension. The potential role of L1 reading comprehension was therefore not adequately
addressed. Other student characteristics may also explain some of the differential findings. For
example, many studies were conducted with students learning a foreign language at the college
level, a very different learning context than that of our population. Clearly then, learning context
and sociolinguistic factors are also at play in determining L2 reading abilities, as already suggested
by Carrell’s(1991)findings, reported above, where L2 proficiency was the strongest predictor of L2
reading comprehension for foreign language learners, but where for ESL learners, L1 reading compre-
hension was the strongest predictor. However, Carrell did not compare the L1 reading levels of the
two groups and L2 proficiency was measured according to the course students were enrolled in;
these two methodological issues make it difficult to conclude that the different learning situations
were the cause of the different findings.
We suggest that the findings from previous research on L2 reading comprehension cannot be
directly applied to adolescent newcomers in U.S. schools, and that the assumption of L2 proficiency
as the strongest predictor of L2 reading comprehension is weakened when looking at this particular
population of students. Our results clearly show that for newcomer adolescent ELLs, existing reading
skills in the home language are critical in the development of reading comprehension in the
language they are learning. We are not the first to make this observation, and our results are sup-
ported by other recent studies that have underscored the important role of L1 literacy in the devel-
opment of a second language. Sparks et al. (2012) in a longitudinal study of fifth through tenth
graders in the U.S. found that L1 (English) reading comprehension in the tenth grade significantly
predicted L2 (either Spanish, French, or German) reading comprehension in the same grade, over
and above the variance of L1 reading ability in first grade. Furthermore, tenth grade L1 reading
achievement made significant and unique contributions not only to L2 reading comprehension,
but also to total L2 proficiency, L2 word decoding, and L2 listening/speaking. Payne, Kalibatseva,
and Jungers (2009) found that in college-age students in the U.S., L1 (English) reading comprehen-
sion was strongly related to L2 (Spanish) reading comprehension, mediated by working memory,
which itself did not uniquely contribute to L2 reading. L1 reading ability also contributed to L2
reading ability more than domain specific knowledge about the reading. More research must be
done to explore the impact of L1 reading skills in the development of a second language. Much
suggests that the two are importantly linked, though perhaps in different ways for different
learner populations. What we conclude from our results, however, is that for newcomer adolescents
with lower-than-expected literacy in their home language, the skills already gained in the L1 are
transferable to literacy development in the L2. This argues for a program of support and maintenance
of the L1 as a pedagogical best practice for immigrant students who must gain advanced skills in the
L2 in a short period of time.
Notes
1. Note that when we use the term ‘language problem’we do not mean language impairment. We follow the usage
in Alderson’s paper to refer to low language proficiency.
12 L. GARRISON-FLETCHER ET AL.
2. This is a different threshold hypothesis from that of Cummins (1981) which is prevalent in the bilingualism litera-
ture. The reading literature’s threshold hypothesis referred to here was formulated by Clarke (1980) and posits a
threshold for L2 proficiency in the acquisition of L2 reading. Cummins’threshold hypothesis posits a threshold of
L1 proficiency in order to acquire an L2 and become bilingual.
3. As noted by Yamashita (1999), studies have not focused on defining this L2 proficiency level and what it involves.
4. A total of 105 students were tested. Of these, 64 completed all of the tests. Two outliers were identified, thus
bringing the total number of students included in the study to 62.
5. The school was missing the date of birth for one of the students; thus, the age is reported for N= 61.
6. The school was missing country of origin for one of the students; thus, this data is reported for N= 61.
7. There is an English version of this assessment, the Diagnostic Assessment of Reading Comprehension (DARC). The
description of the EDICOLE is taken from the description of the DARC, as there is more information describing the
English version of the assessment.
8. In order to test if the effects of L1 reading comprehension and L2 language proficiency were significantly
different, we ran an additional test which previous studies have not. We found that the difference in effect
size was not statistically significant (Hotelling’st(59) = 0.664, p= .51). Future studies should make this comparison
in order to see if there is a statistically significant difference between the effects of the predictor variables (Die-
denhofen and Musch 2015). Further research with larger numbers of participants may well find significant differ-
ences in effect size.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Leigh Garrison-Fletcher is an Associate Professor of Linguistics and ESL in the Department of Education and Language
Acquisition at LaGuardia Community College of the City University of New York. Her research focuses on the develop-
ment of literacy in emergent bilinguals. She is currently engaged in professional development to help educators best
meet the needs of emergent bilingual students in higher education.
Gita Martohardjono is professor of Linguistics at Queens College and the Graduate Center of the City University of New
York. Her area of specialization is Second Language Acquisition, and her research interests include the development of
syntax and literacy in bilingual immigrant populations across the age span.
Martin Chodorow is a professor of Psychology and Linguistics at Hunter College and the Graduate Center of the City
University of New York. His research interests include automated assessment of writing quality and computational
methods for detecting grammar and word usage errors in nonnative writing.
ORCID
Leigh Garrison-Fletcher http://orcid.org/0000-0003-4252-5043
References
Alderson, J. C. 1984.“Reading in a Foreign Language: A Reading Problem or a Language Problem?”In Reading in a Foreign
Language, edited by J. C. Anderson and A. H. Urquhart, 1–27. London: Longman.
Bernhardt, E. B., and M. L. Kamil. 1995.“Interpreting Relationships between L1 and L2 Reading: Consolidating the
Linguistic Threshold and the Linguistic Interdependence Hypotheses.”Applied Linguistics 16: 15–34.
Brisbois, J. E. 1995.“Connections between First- and Second-Language Reading.”Journal of Reading Behavior 27: 565–584.
Bossers, B. 1991.“On Thresholds, Ceilings and Short-Circuits: The Relation between L1 Reading, L2 Reading and L2
Knowledge.”AILA Review 8: 45–60.
Carrell, P. L. 1991.“Second Language Reading: Reading Ability or Language Proficiency?”Applied Linguistics 12: 159–179.
Clarke, M. A. 1980.“The Short Circuit Hypothesis of ESL Reading –Or When Language Competence Interferes with
Reading Performance.”The Modern Language Journal 64: 203–209.
Cummins, J. 1979.“Linguistic Interdependence and the Educational Development of Bilingual Children.”Review of
Educational Research 49 (2): 222–251.
Cummins, J. 1981.“The Role of Primary Language Development in Promoting Educational Success for Language-Minority
Students.”In Schooling and Language-Minority Students: A Theoretical Framework, edited by M. Ortiz, D. Parker, and F.
Tempes, 3–49. Los Angeles: California State University, Evaluation, Dissemination, and Assessment Center.
INTERNATIONAL JOURNAL OF BILINGUAL EDUCATION AND BILINGUALISM 13
Cummins, J. 1991.“Interdependence of First- and Second-Language Proficiency in Bilingual Children.”In Language
Processing in Bilingual Children, edited by E. Bialystok, 70–89. Cambridge: Cambridge University Press.
Cutting, L. E., and H. S. Scarborough. 2006.“Prediction of Reading Comprehension: Relative Contributions of Word
Recognition, Language Proficiency, and Other Cognitive Skills can Depend on how Comprehension is Measured.”
Scientific Studies of Reading 10: 277–299.
Diedenhofen, B., and J. Musch. 2015.“Cocor: A Comprehensive Solution for the Statistical Comparison of Correlations.”
PLoS ONE 10 (4): e0121945. doi:10.1371/journal.pone.o121945.
Francis, D. J., M. Rivera, N. Lesaux, M. Kieffer, and H. Rivera. 2006.Practical Guidelines for the Education of English Language
Learners: Research-Based Recommendations for Serving Adolescent Newcomers. Portsmouth, NH: RMC Research
Corporation, Center on Instruction. http://www.centeroninstruction.org/files/ELL2-Newcomers.pdf.
Freeman, Y. S., and D. E. Freeman. 2002.Closing the Achievement Gap: How to Reach Limited-Formal-Schooling and Long-
Term English Learners. Portsmouth, NH: Heinemann.
Francis, D. J., C. E. Snow, D. August, C. D. Carlson, J. Miller, and A. Iglesias. 2006.“Measures of Reading Comprehension: A
Latent Variable Analysis of the Diagnostic Assessment of Reading Comprehension.”Scientific Studies of Reading 10:
301–322.
Klein, E., and G. Martohardjono. 2006.Understanding the Student with Interrupted Formal Education (SIFE): A Study of SIFE
Skills, Needs and Achievement (Phase I). Report by the Research Institute for the Study of Language in Urban Society,
CUNY Graduate Center to the Office of English Language Learners. New York: New York City Department of Education.
Klein, E., and G. Martohardjono. 2009.Understanding the Student with Interrupted Formal Education (SIFE): A Study of SIFE
Skills, Needs and Achievement (Phase II). Report to the NYC DOE Office of English Language Learners. New York:
New York City Department of Education.
Lee, J. W., and D. Schallert. 1997.“The Relative Contribution of L2 Language Proficiency and L1 Reading Ability to L2
Reading Performance: A Test of the Threshold Hypothesis in an EFL Context.”TESOL Quarterly 31: 713–739.
Lesaux, N. K., K. Koda, L. S. Siegel, and T. Shanahan. 2006.“Development of Literacy.”In Developing Literacy in Second-
Language Learners: Report of the National Literacy Panel on Language-Minority Children and Youth, edited by D.
August and T. Shanahan, 75–122. Mahwah, NJ: Lawrence Erlbaum Associates.
Melby-Lervåg, M., and A. Lervåg. 2011.“Cross-linguistic Transfer of Oral Language, Decoding, Phonological Awareness
and Reading Comprehension: A Meta-Analysis of the Correlational Evidence.”Journal of Research in Reading 34:
114–135.
Melby-Lervåg, M., and A. Lervåg. 2014.“Reading Comprehension and Its Underlying Components in Second-Language
Learners: A Meta-Analysis of Studies Comparing First- and Second-Language Learners.”Psychological Bulletin 140:
409–433.
Mirza, A., A. Gottardo, and X. Chen. 2017.“Reading in Multilingual Learners of Urdu (L1), English (L2) and Arabic (L3).”
Reading and Writing: An Interdisciplinary Journal 30 (1): 187–207.
Morse, A. 2005. A Look at Immigrant Youth: Prospects and Promising Practices. National Conference of State Legislatures:
Children’s Policy Initiative.
New York City Department of Education. 2015.Department of English Language Learners and Student Support. School Year
2013–2014 Demographic Report.http://schools.nyc.gov/NR/rdonlyres/FC0B4035-00DF-4318-A1F7-6EF23C15B7F6/0/
20132014DemographicReportFinalWINTER2015.pdf.
Payne, T., Z. Kalibatseva, and M. Jungers. 2009.“Does Domain Experience Compensate for Working Memory Capacity in
Second Language Reading Comprehension?”Learning and Individual Differences 19: 119–123.
Pichette, F., N. Segalowitz, and K. Connors. 2003.“Impact of Maintaining L1 Reading Skills on L2 Reading Skill Development
in Adults: Evidence From Speakers of Serbo-Croatian Learning French.”The Modern Language Journal 87: 391–403.
Royer, J. M., and M. S. Carlo. 1991.“Transfer of Comprehension Skills from Native to Second Language.”Journal of Reading
34: 450–455.
Shiotsu, T., and C. J. Weir. 2007.“The Relative Significance of Syntactic Knowledge and Vocabulary Breadth in the
Prediction of Reading Comprehension Test Performance.”Language Testing 24: 99–128.
Short, D., and S. Fitzsimmons. 2007.Double the Work: Challenges and Solutions to Acquiring Language and Academic
Literacy for Adolescent English Language Learners –A Report to Carnegie Corporation of New York. Washington, DC:
Alliance for Excellent Education.
Snow, C. 2006.“Cross-cutting Themes and Future Directions.”In Developing Literacy in Second-Language Learners: Report
of the National Literacy Panel on Language- Minority Children and Youth, edited by D. August and T. Shanahan, 275–
300. Mahwah, NJ: Lawrence Erlbaum Associates.
Sparks, R. L., J. Patton, L. Ganschow, and N. Humbach. 2012.“Do L1 Reading Achievement and L1 Print Exposure
Contribute to the Prediction of L2 Proficiency?”Language Learning 62 (2): 473–505.
van Gelderen, A., R. Schoonen, R. D. Stoel, K. De Glopper, and J. Hulstijn. 2007.“Development of Adolescent Reading
Comprehension in Language 1 and Language 2: A Longitudinal Analysis of Constituent Components.”Journal of
Educational Psychology 99: 477–491.
Verhoeven, L. T. 1990.“Acquisition of Reading in a Second Language.”Reading Research Quarterly 25: 90–114.
Yamashita, J. 1999.“Reading in a First and a Foreign Language: A Study of Reading Comprehension in Japanese (the L1)
and English (the L2).”Unpublished PhD thesis. Lancaster University.
14 L. GARRISON-FLETCHER ET AL.
