development and cross-language transfer of oral … · increasingly, oral reading fluency is...
TRANSCRIPT
DEVELOPMENT AND CROSS-LANGUAGE TRANSFER OF ORAL
READING FLUENCY USING LONGITUDINAL AND
CONCURRENT PREDICTORS AMONG CANADIAN FRENCH
IMMERSION PRIMARY-LEVEL CHILDREN
by
Kathleen Lee
A thesis submitted in conformity with the requirements
for the degree of Masters of Art
Applied Psychology & Human Development
Ontario Institute for Studies in Education
University of Toronto
© Copyright by Kathleen Lee 2014
ii
Development and Cross-Language Transfer of Oral Reading Fluency
using Longitudinal and Concurrent Predictors among Canadian French
Immersion Primary-Level Children
Kathleen Lee
Master of Arts
Applied Psychology and Human Development
Ontario Institute for Studies in Education
University of Toronto
2014
Abstract This thesis investigates development and transfer of oral reading fluency among early French
immersion students. Using a longitudinal design, students were assessed on phonological
awareness, rapid naming, word-level fluency and text-level fluency in English and in French in
Grade 2 and Grade 3. In three related studies, this thesis examines transfer both within levels of
fluency individually (word-level and text-level) and between levels of fluency (from word-level
to text-level). The results indicated that word-level fluency significantly improved over the one-
year period in both English and in French. Language status comparing English-as-first-language
students (EL1) and English-language-learners (ELLs) did not influence fluency performance in
either language. Further, results showed bidirectional transfer of fluency at the word-level and
the text-level independently, and unidirectional transfer from word to text fluency from French to
English only. These findings provide evidence supporting cross-language transfer of oral
reading fluency both within and between levels of the construct.
iii
Acknowledgements
There are few achievements in life that are ever accomplished without assistance from
others, and I am very appreciative of the people in my life who have helped me achieve mine.
First and foremost, I would like to express my sincere gratitude to my supervisor, Dr. Xi Chen,
who has patiently and consistently supported me throughout these past few years. Her expertise
in the field of second language learning has been invaluable to my learning and I am very
appreciative for all of the guidance she has given me. I feel extremely fortunate to be part of such
a wonderful and welcoming lab. I would also like to thank my second reader, Dr. Esther Geva,
for the insight and feedback she has given me as well.
In addition, I would like to thank my classmates and my friends for their immeasurable
support and encouragement throughout this process. These past two years would not have been
nearly as enjoyable and meaningful without all of you there with me. Lastly, I want to express
my gratitude to my parents whose unwavering love and support has allowed me to pursue every
dream I have ever reached for. Nothing I have ever achieved would have been possible without
the two of you and I will forever be appreciative of everything you have given me.
iv
Table of Contents
Abstract............................................................................................................................................ii
Acknowledgements ........................................................................................................................iii
List of Appendices..........................................................................................................................vi
1 Introduction.................................................................................................................................1
2 Literature Review .......................................................................................................................2
2.1 French Immersion Education in Canada .............................................................................2
2.2 Research on the Effects of Immersion Education ...............................................................4
2.3 Theory of Transfer...............................................................................................................6
2.4 Empirical Evidence of Cross-Language Transfer .............................................................10
2.5 Oral Reading Fluency........................................................................................................13
2.6 Cross-Language Transfer of Oral Reading Fluency..........................................................19
2.7 Rationale for Present Study ...............................................................................................21
3 Study 1: Word-Level to Word-Level Fluency..........................................................................22
3.1 Participants ........................................................................................................................22
3.2 Measures............................................................................................................................24
3.3 Procedure...........................................................................................................................26
3.4 Results ...............................................................................................................................26
3.4.1 Preliminary Analyses.............................................................................................26
3.4.2 Main Analyses .......................................................................................................27
3.5 Discussion..........................................................................................................................30
4 Study 2: Word-Level to Text-Level Fluency............................................................................31
4.1 Participants ........................................................................................................................32
4.2 Measures............................................................................................................................32
4.3 Procedure...........................................................................................................................33
v
4.4 Results ...............................................................................................................................34
4.4.1 Preliminary Analyses.............................................................................................34
4.4.2 Main Analyses .......................................................................................................34
4.5 Discussion..........................................................................................................................36
5 Study 3: Text-Level to Text-Level Fluency .............................................................................38
5.1 Participants ........................................................................................................................38
5.2 Measures............................................................................................................................38
5.3 Procedure...........................................................................................................................38
5.4 Results ...............................................................................................................................39
5.4.1 Preliminary Analyses.............................................................................................39
5.4.2 Main Analyses .......................................................................................................39
5.5 Discussion..........................................................................................................................42
6 General Discussion ...................................................................................................................43
References .....................................................................................................................................52
vi
List of Appendices
Appendix A Table 1. Descriptive Statistics for All Measures at T1 and T2.........................59
Appendix B Table 2. Pearson Correlations between All Variables in Study 1.....................60
Appendix C Table 3. Hierarchical Regression Predicting T2 English Word Fluency Table 4. Hierarchical Regression Predicting T2 French Word Fluency...........61
Appendix D Table 5. Commonality Analysis Predicting T2 English Word Fluency Table 6. Commonality Analysis Predicting T2 French Word Fluency ............62
Appendix E Table 7. Pearson Correlations between All Variables in Study 2 .....................63
Appendix F Table 8. Hierarchical Regression Predicting T2 English Text Fluency Table 9. Hierarchical Regression Predicting T2 French Text Fluency .............64
Appendix G Table 10. Commonality Analysis Predicting T2 English Text Fluency Table 11. Commonality Analysis Predicting T2 French Text Fluency .............65
Appendix H Table 12. Pearson Correlations between All Variables in Study 3....................66
Appendix I Table 13. Hierarchical Regression Predicting T2 English Text Fluency Table 14. Hierarchical Regression Predicting T2 French Text Fluency ............67
Appendix J Table 15. Commonality Analysis Predicting T2 English Text Fluency Table 16. Commonality Analysis Predicting T2 French Text Fluency ..............68
1
1 Introduction
Within the past few decades, an influx of students exposed to diverse linguistic
backgrounds has entered educational systems, influencing the traditional understanding of literacy
acquisition in monolingual English speakers. This is coupled with the rising popularity of
language immersion programs, which further complicates the linguistic context in which students
learn to read and write today. Interestingly, a large body of research has shown that French
immersion students often attain higher second language skills (e.g., Genesee & Jared, 2008),
higher overall academic achievement (Swain & Lapkin, 1982; Turnbull, Hart & Lapkin, 2001)
and even outperform conventionally educated students in linguistic skills in the first language
(Cheng, Li, Kirby, Quiang & Wade-Woolley, 2010; Turnbull, Hart & Lapkin, 2003). One widely
accepted explanation for this points to the existence of transfer effects between the two languages.
Research on cross-language transfer of language skills supports this facilitatory effect such that
proficiency in linguistic skills in one language can have a positive influence on the development
of that skill in another language (e.g., Chen, Xu, Nguyen, Hong & Wang, 2010; Gottardo, 2002).
Therefore, immersion programs may offer a unique opportunity for students to solidify early
reading skills that can transfer between the multiple languages within their repertoire.
The acquisition of literacy skills is especially imperative in the primary grades as students
transition from “learning to read” to an increased reliance on “reading to learn” (Alexander,
2003). The process of learning to read is built upon many independent underlying skills that need
to work simultaneously to facilitate successful reading. The coordination of these tasks can be
measured by oral reading fluency ability, which requires the skilled execution of these underlying
micro-level processes. Increasingly, oral reading fluency is gaining a more prominent role in
literacy acquisition research, with theoretical and empirical evidence supporting its role as a
2
fundamental indicator of “overall reading competence” (Fuchs, Fuchs & Hosp, 2001). This thesis
explores the cross-language transfer of this important linguistic skill among language immersion
students who are exposed to a unique linguistic environment. In three related studies, this thesis
investigates the development and cross-language transfer of different levels of fluency ability
using both cross-sectional and longitudinal data among primary students learning to read in a
Canadian French immersion program.
2 Literature Review
2.1 French Immersion Education in Canada
With over 350,000 Canadian students enrolled in French immersion programs in 2011
(Office of the Commissioner of Official Languages, 2011), immersion education is rapidly
becoming a popular option for families today. In these immersion programs, students receive
significant amounts of instruction in French (Baker, 2011) and develop literacy skills in both
English and French simultaneously. Immersion education allows children to obtain proficiency in
an additional language with the ultimate aim of achieving additive bilingualism in both languages
of instruction. In the mid-1960s, French immersion was initially introduced in the province of
Quebec with four main objectives: 1) to allow students to become functionally competent in oral
and written French, 2) to continue and sustain the development of students’ first language, 3) to
teach students academic content appropriate to their age and grade level and 4) to allow
Anglophone students to develop an understanding of the Francophone culture and language
(Genesee, 1984). The initial program was designed to capitalize on young children’s ability to
learn language incidentally without impeding their native language development, academic
achievement or general cognitive development (Genesee, 1984).
3
Since its inception, French immersion programs have been implemented across the
country and variations of the program have arisen in response to its growing popularity and
success. Currently, language immersion programs are differentiated by the grade in which the
immersion language is introduced and the amount of instruction received in the immersion
language. Early immersion programs begin in kindergarten or Grade 1 where students receive
100% of academic instruction in the French language. Therefore, children are taught literacy skills
in French before they are formally taught to read and write in English (Genesee & Jared, 2008).
Initially, emphasis is placed on developing listening comprehension skills and teachers are
generally instructed to use French at all times in the classroom (Genesee, 1984). Beginning in
Grade 4, English is introduced and increasingly integrated until about 50% of academic content is
taught in English by Grade 5 or 6. In a different model, middle immersion programs begin to use
French as the main medium of instruction in Grade 4 or 5, with relatively equal instruction
delivered in either language by Grade 7. In the late immersion model, French instruction begins in
Grade 6 or 7, with 50-80% of academic content delivered in French and the remainder taught in
English. By the end of Grade 8, students enrolled in early, middle or late French immersion
programs will have accumulated approximately 6000, 2000 or 1200 total hours of French
instruction respectively (Genesee, 1984). In these programs, students typically enter with very
limited levels of French proficiency, and exposure to French is largely confined to the classroom
environment. At the secondary school level, immersion students are offered the option of
academic courses instructed in the French language, in order to maintain and enhance their
acquired French language skills. Therefore, the French immersion curriculum in Canada
encompasses these three phases: 1) immersion phase, where French is the sole language of
instruction, 2) bilingual phase, with relatively equal amounts of academic instruction delivered in
4
English and in French and 3) follow-up phase, where select courses are taught in French at the
secondary level to maintain their skills (Genesee, 1984).
2.2 Research on the Effects of Immersion Education
In response to rising student populations from diverse linguistic backgrounds, a large body
of international research has sought to explore the consequences of language plurality in an
educational context. Since the 1960s, the impact of French immersion programs on students’
English and French literacy has been a major focus of research. Early research sought to explore
whether immersion education negatively impacted development and proficiency in the L1
language (Genesee, 1978). Decades of empirical evidence sampled across different nations around
the world overwhelmingly point to the fact that this is not the case. In the majority of studies,
linguistic and academic performance of immersion students are compared to those in traditional
English programs and students are matched on age, intellectual ability and socio-economic indices
to the greatest extent possible. This body of research shows that though students may initially lag
on measures of the dominant language (Genesee, 1978; Swain & Lapkin, 1982), both immersion
and non-immersion students are virtually identical on nearly all measures of English language
proficiency once instruction in the dominant language is introduced (Cheng et al., 2010; Swain &
Lapkin, 1982; Turnbull et al., 2003). The initial short-term lag of approximately two years
generally disappears once instruction in the English language is incorporated into the curriculum
(Genesee, 1978; Swain & Lapin, 1982). The only consistent exception to this is found for
vocabulary due to the concept of a limited capacity for words stored within one’s mental lexicon
that must be shared amongst multiple languages. Studies show that monolingual students typically
outperform their bilingual peers on measures of receptive vocabulary (Bialystok, Luk, Peets &
Yang, 2010) and expressive vocabulary (Paez, Tabors & Lopez, 2007).
5
However, in all other areas of English literacy skills, a wealth of evidence suggests that
students in early, middle and late immersion programs obtain comparable, if not higher levels of
English proficiency. For example, Turnbull and colleagues (2001; 2003) analyzed province-wide
data from Ontario’s standardized Education Quality and Accountability Office (EQAO) testing to
explore whether differences existed in English literacy skills among immersion and non-
immersion students. The Grade 3 EQAO test results indicated that immersion students performed
as well, or slightly better than non-immersion peers on tests of English reading and writing
(Turnbull et al., 2001). Interestingly, among Grade 6 students, those enrolled in French immersion
programs significantly outperformed their non-immersion peers on English reading and writing
tasks despite receiving less formal instruction in the English language (Turnbull et al., 2003).
However, these rapid gains in performance are interpreted with caution as some argue that Grade
6 immersion students’ scores may be inflated due to academically weaker students being demitted
from the program. Nevertheless, this adds to a growing body of literature supporting the premise
that linguistic (Genesee & Jared, 2008), cognitive (Hermanto, Moreno & Bialystok, 2012) and
even academic benefits (Rolstad, Mahoney & Glass, 2005; Swain & Lapkin, 1982; Turnbull et al.,
2001) may accrue when individuals acquire proficiency in multiple languages simultaneously.
One explanation for this phenomenon points to the role of transfer effects that may occur
within the context of multilingualism. Due to the interdependence between some aspects of
language proficiency, skills that are acquired in French in an early immersion program may
transfer to English linguistic ability once the English language is formally introduced (Cummins
& Swain, 1986). Therefore, literacy instruction in one language may develop not only language-
specific skills, but also a deeper underlying proficiency that may be transferrable across
languages. In this way, immersion programs may offer a unique opportunity to develop language-
independent literacy skills beyond those in traditional non-immersion programs.
6
2.3 Theory of Transfer
Research on cross-language transfer of language skills supports the hypothesis of a
facilitatory effect such that proficiency in linguistic skills in one language may have a positive
influence on the development of that skill in another language (e.g., August, Calderon & Carlo,
2002). Various theoretical frameworks have been offered to conceptualize the L1-L2 relationship
observed between parallel language tasks, however, no single account alone is adequate to explain
all aspects of the empirical evidence. This line of research is grounded in the theoretical
framework of the interdependence hypothesis that was first postulated by Cummins (1981; 1984).
Cummins’ hypothesis posits that the acquisition of multiple languages is developmentally
interdependent and is linked by common underlying proficiencies (Cummins, 1981). Therefore,
developing competence in certain linguistic skills in one language may facilitate the acquisition of
the corresponding skill in another language. However, it is important to note that transfer is not
believed to be automatic and that certain conditions have been postulated to facilitate or prevent
transfer of skills from an L1 to an L2. For instance, Cummins hypothesized that transfer of skill
may be contingent upon the quality of L1 language instruction, sufficient proficiency in the L2 as
well as adequate exposure to the L2 (Cummins, 1981; Cummins, 2012).
Furthermore, Cummins hypothesized that not all skills acquired in developing language
proficiency are subject to transfer in the same way. In particular, Cummins (2000) makes two
important distinctions that attempt to delineate the language skills that are most likely to transfer
and refer to his construct of common underlying proficiencies. One distinction differentiates
between language that is used for academic or higher-level cognitive purposes from the language
used in day-to-day interpersonal communication. The distinction between these language types is
characterized by the extent of contextual support that is available for the speaker to rely on while
communicating. For instance, language used for academic purposes is typically context-reduced,
7
and as such, requires more thoughtful and deliberate use of language in order to facilitate
communication. Alternatively, the language that is used in day-to-day interpersonal
communication is typically considered to be context-embedded, as both parties can benefit from
support through shared context and mutual experience during verbal communication (e.g.,
speaking to somebody about an event that both parties attended). A second distinction is made
between cognitively demanding and cognitively undemanding communication. The first concept
refers to communication that necessitates language skills that are learned, may not yet be fully
automatized and require cognitive involvement on the part of the speaker. This is contrasted from
cognitively undemanding communication, which draws on more basic, automatized or over-
learned language skills that place little cognitive demand on the speaker. In this way, Cummins
attempts to describe language along these two continuums, which characterizes language by the
amount of contextual support available during communication and the cognitive demands placed
on the speaker. Along these dimensions, Cummins (2000) postulates that skills involved in
context-reduced, cognitively-demanding language use, such as literacy-related language skills, are
the most likely candidates for transfer. These are the common underlying proficiencies that
Cummins discusses, that transcend the typological specificity of reading and writing instruction in
any one particular language. For example, he asserts that “L2 instruction that develops L2 reading
and writing skills is not just developing L2 skills, it is also developing a deeper conceptual and
linguistic proficiency that is strongly related to the development of literacy in the majority
language” (Cummins, 2012, p. 1980). In a review of studies examining language development in
ELLs, Genesee, Geva, Dressler and Kamil (2006) refer to these skills as the ‘procedural’
knowledge that underlies academic or cognitively demanding tasks. This can include
metalinguistic or metacognitive skills that transcend an individual language such as monitoring
skills, meaning-making strategies or the skills involved in defining the meaning of words
8
(Durgunoglu, 2002). For instance, reading comprehension in a given language involves not only
decoding and word reading skills but likely develops proficiency in higher-order thinking skills
such as inferencing, generalizing and predicting skills that can also be transferred to an L2.
Likewise, fluent reading has been posited to result from the successful and efficient
synchronization of phonological, orthographic and semantic processes (Breznitz, 2003; Breznitz,
2006). It may be reasonable to assume that the process of creating links among phonological,
orthographic and semantic patterns develops similarly across different languages, and as such,
proficiency in this integration process can likely be transferred to an L2 as well. Therefore, it can
be hypothesized that these skills (i.e., inferencing, generalizing and predicting skills in reading
comprehension; synchronization of phonological, orthographic and semantic processes in fluent
reading) can be considered examples of ‘procedural’ knowledge that underlie linguistic tasks and
are likely subject to transfer across different languages.
An interesting parallel can be drawn from Cummins’ explanation of higher-order
cognitively demanding language to the theoretical representation of languages in bilinguals that
was proposed by Bialystok and Hakuta (1994). The authors speculate that proficient bilinguals
may have two components to the representations of their two languages in their minds – a
common representation that is the record of general cognitive and linguistic knowledge, and a
second separate representation that records language-specific information (Bialystok & Hakuta,
1994). Therefore, there may be certain ‘procedural’ skills and strategies that only need to be
acquired once to operate across languages, while other skills that involve more language-
dependent knowledge are specific only to the language being acquired (Durgunoglu, 2002). In this
way, linguistic skills that tap into more ‘procedural’ and language-general ability can be
considered more likely candidates for transfer.
9
Though much research on cross-language transfer is explained within the framework of
the interdependence hypothesis, this explanation alone has been found inadequate to account for
all the nuances observed in L1-L2 relationships on parallel language tasks in different languages.
In particular, Cummins’ theoretical model has been criticized for its lack of specificity in
describing his concept of common underlying proficiencies and in identifying what exact skill or
ability is transferring between the L1 and the L2 (Genesee et al., 2006). Some have argued that
this theory is too general and does not address the specific cognitive mechanisms and processes
that facilitate transfer of linguistic skills (Genesee et al., 2006; Geva & Ryan, 1993). In response
to this need for clarification, other theories have surfaced that attempt to expand on and refine the
understanding of relationships between L1 and L2 performance on language tasks. These newer
conceptualizations attempt to elucidate what exactly learners bring from linguistic knowledge
acquired in one language to the process of learning to read in another language. One such
explanation comes from Geva and Ryan (1993) who posit that correlations between L1 and L2
ability on similar language tasks may be due not only to transfer, but also to underlying cognitive
processes that are related to performance across both languages. The authors speculate that a
consideration of individual differences in innate cognitive processes such as working memory,
phonological awareness and rapid automatized naming may also partially account for the high
degree of correlation between performance on similar language tasks in different languages. In
addition, more recent theories such as the transfer facilitation model (Koda, 2008) emphasize the
impact of prior literacy knowledge and experience on reading development in a second language.
In particular, Koda (2008) emphasizes the importance of metalinguistic skills that are acquired in
one language and can be recruited to enhance linguistic skills in other languages. This speaks to
the role of prior linguistic knowledge that is acquired in a first language, which is regarded as a
reservoir of knowledge, skills, and abilities that is available when learning literacy in a new
10
language (Riches & Genesee, 2006). Koda (2008, p.78) defines transfer of this prior linguistic
knowledge (e.g., metalinguistic awareness) as the “automatic activation of well-established first-
language competencies, triggered by second-language input.” In addition, Koda (2008, p.72)
specifically posits that “first-language reading experience not only has lasting impacts on second
language reading development, but also systematically alters processing procedures for second
language print information.” In doing so, she theorizes beyond initial views of transfer as simply
proficiency in one language facilitating the acquisition of a corresponding skill in another
language, and offers a more specific explanation of the mechanism that results in L1 and L2
interrelationships. Though the current review of theories of transfer is limited in its scope, it is
clear that a consensus regarding the specific mechanisms of transfer have yet to fully emerge.
Despite disagreement in the interpretation of transfer effects, it is evident that a broader
consideration of factors beyond those offered in the interdependence hypothesis is required to
understand the mechanisms and extent of transfer that is observed across different linguistic skills.
2.4 Empirical Evidence of Cross-Language Transfer
There is a large body of empirical evidence demonstrating cross-language transfer of
linguistic skills. Some linguistic skills are considered to be language-independent and have been
shown to transfer universally across languages. It is suggested that these skills do not need to be
learned separately for each language in order to operate equally well across different languages.
For instance, phonological awareness is an important linguistic skill referring to one’s
awareness of the units within their spoken language. It refers to the ability to recognize and
manipulate the phonological units of language at the word, syllable and phoneme levels (Wagner,
Torgesen & Rashotte, 1999a). Phonological awareness is understood to be a prerequisite for
successful reading and is highly correlated with word recognition and spelling ability (Cossu,
11
Shankweiler, Liberman, Katz & Tola, 1988; Durgunoglu & Oney, 1999; Lundberg, Olofsson &
Wall, 1980). Conversely, poor phonological skills have been shown to impair decoding ability
and compromise overall reading performance (Torgesen et al., 1999). A large body of research
documents the existence of transfer among structurally similar language pairs including French-
English (Comeau, Cormier, Grandmaison & Lacroix, 1999), Italian-English (D’Angiulli, Siegel &
Serra, 2001) and Spanish-English (Durgunoglu, Nagy & Hancin-Bhatt, 1993). Interestingly, this
ability appears to exhibit universal transfer, even between languages that differ widely in
phonology and orthographic script (Chen et al., 2010; Durgunoglu, 2002; Gottardo, Yan, Siegel &
Wade-Woolley, 2001; McBride-Chang et al., 2008). Therefore, phonological awareness appears
to tap into a cognitive, language-general process involving an ability to manipulate sound units of
a language, regardless of what the specific language and sound units are. For instance, Wang,
Park and Lee (2006) examined predictors of English and Korean word reading among Grade 1,
Grade 2 and Grade 3 students enrolled in a Korean language school in a large American city. The
authors showed that phonological ability in Korean and English were significantly correlated. Of
even greater interest, their results provide evidence for cross-language transfer of phonological
processing in predicting decoding ability. Korean phonological skills were found to predict a
significant amount of unique variance in English pseudoword reading beyond English
phonological and orthographic ability. Thus, this study provides evidence for transfer effects in
phonological ability even between languages that differ widely in their phonology and
orthography. Evidence in other dissimilar language pairs has been documented as well, with Chen
et al. (2010) showing cross-language transfer of phonological ability among Grade 1 and Grade 3
Chinese students enrolled in English programs in China. Taken together, these findings reflect a
large body of literature documenting a robust and universal cross-language transfer effect of
phonological ability between various language pairs.
12
However, other reading skills are considered more dependent on the linguistic features of
the languages being acquired. For instance, research increasingly indicates that orthographic
processing is a literacy skill that is language-dependent and shows limited transfer between
languages. Orthographic knowledge refers to one’s knowledge of permissible letter patterns in the
written system of their language (Perfetti, 1985). This refers to one’s knowledge of acceptable
letter sequences and their relations to letter positions in the words of their language. For example,
within the English language, letter doublets are acceptable in the final position of a word but not
in the initial position. Orthographic processing is understood to make a unique contribution to
reading development as well. For instance, studies have shown that orthographic skills explain
significant unique variance in word reading above the effects of phonological ability (Barker,
Torgesen & Wagner, 1992; Cunningham, Perry & Stanovich, 2001). However, research on cross-
language transfer of orthographic ability suggests that transfer is variable and may depend on the
degree of similarity between the languages in question. For instance, among languages that are
distantly related, transfer of orthographic ability does not appear to occur. Wang, Perfetti and Liu
(2005) examined the role of phonological and orthographic processing on word reading ability in
Chinese-English bilinguals in Grades 2 and 3. The findings indicated that transfer of phonological
ability occurred between Chinese and English, however, orthographic processing ability did not
predict word reading in either direction. The authors speculate that the lack of orthographic
transfer can be explained by the vastly different visual forms that make up English and Chinese
orthography. However, among language pairs with more structural similarity, evidence for
transfer of orthographic processing has been shown to occur. Deacon, Chen, Luo and Ramirez
(2013) examined transfer of orthographic processing among Spanish-English bilingual children in
Grade 4 and Grade 7. The authors hypothesized that transfer of orthographic processing is more
likely to occur when the scripts under acquisition utilize the same alphabet. After controlling for
13
known within-language predictors of word reading, Spanish orthographic processing was found to
account for an additional 5% of unique variance in English word reading, demonstrating transfer
of this linguistic skill. The authors concluded that their results support the idea of transfer of
orthographic processing to reading outcomes across languages that are structurally similar in
orthographic script (Deacon et al., 2013). This suggests that the language-specific structural
aspects of the languages being acquired may be important in determining the nature of transfer in
orthographic processing.
Therefore, the literature on cross-language transfer indicates variability in the extent of
transfer that occurs for different linguistic skills. It is useful to contrast the pattern of transfer of
phonological processing with that of orthographic processing to understand this distinction. While
phonological processing ability appears to transfer between languages regardless of their script,
the transfer of orthographic processing has been shown to depend on the structural similarity of
the languages. Therefore, while language-specific skills such as orthographic processing and
syntactic ability may rely more heavily on language-dependent knowledge, skills within more
general or cognitive domains may arguably be stronger candidates for transfer. This implicates a
variety of higher-level cognitive skills that underlie academic learning and are highly related to
the acquisition of literacy. Within the domain of oral reading fluency, very little is currently
known about the pattern and extent of cross-language transfer and as such, more research is
needed in this area. The relatively few studies that have examined the role of transfer in oral
reading fluency will be discussed in greater detail later.
2.5 Oral Reading Fluency
Oral reading fluency refers to the oral translation of written text (Fuchs et al., 2001) and is
characterized by four key features - speed, effortlessness, autonomy and achievement without
14
much consciousness or awareness (Kuhn, Schwanenflugel & Meisinger, 2010). This task involves
the simultaneous coordination of multiple sub-processes that operate together, resulting in fluent
reading when automaticity in the component skills is achieved. Fluent reading is posited to
involve the creation of strong links among phonological, orthographic, and semantic processes
that occur following repeated exposures to print (Breznitz, 2003; Breznitz, 2006). It involves the
processing of orthographic units, the decoding and recoding of phonological units of sounds into
recognizable whole words and is aided by speeded lexical access (Fuchs et al., 2001). Empirical
evidence shows that oral reading fluency is closely linked to ability in phonological awareness,
rapid naming and word reading accuracy, yet fluent reading is understood to be more than the sum
of its component parts (Fuchs et al., 2001). When fluency in reading is achieved, it is considered a
unique and independent construct. Further, it is purported to be a reliable indicator of overall
reading proficiency by freeing up attentional demands to engage in comprehension (Gough,
Hoover & Peterson, 1996; Jenkins, Fuchs, Espin, van den Broek & Deno, 2000).
The concept of oral reading fluency as an indicator of overall reading proficiency is
grounded in the automaticity model of reading that was first postulated by LaBerge and Samuels
(1974). In this framework, each component task involved in reading requires a certain amount of
attentional load. If all components demand conscious awareness (i.e. not yet automatic), the
combined effort will exceed one’s attentional capacity and reading performance suffers. By
contrast, as lower level processes like phonological coding and orthographic segmentation
become automatic, the demands on attentional resources fall within available limits and permit
fluent reading. Though a strictly bottom-up model of reading has since been found inadequate,
this model provides the framework for interpreting reading fluency as an index of underlying
proficiency. Furthermore, Perfetti (1995) proposed other reading subcomponents that may be
candidates for automaticity including rapid lexical access, integrating background knowledge with
15
text, and inferencing. As automaticity develops in these higher-level processes, additional
attentional resources beyond those involved in decoding text can be released to engage in more in-
depth meaning construction. In addition, the role of top-down contextual facilitation during
reading has been identified in more interactive models of reading (Posner & Snyder, 1975a;
Posner & Snyder 1975b; Stanovich, 2000). Both bottom-up and more interactive models of
reading demonstrate how increasing automaticity enhances the capacity of readers to allocate
attentional resources to comprehension. As such, fluency can serve as an indicator of reading
competence because it captures individual differences across multiple subcomponent skills as well
as serves as an index of the overall comprehension of text.
Empirical evidence lends support to these theoretically driven hypotheses that
conceptualize oral reading fluency as an indicator of reading proficiency. There is substantial
evidence suggesting that fluency ability is strongly related to reading comprehension in samples
of both typical developing children as well as those with reading disabilities (Jenkins et al., 2000;
Levy, Abello & Lysynchuk, 1997; Marston, 1989). In a study by Fuchs, Fuchs and Maxwell
(1988), oral reading fluency was shown to have higher criterion validity with standardized
measures of reading comprehension than other measures of comprehension. Using the Reading
Comprehension subtest of the Stanford Achievement Test as the criterion measure, the authors
administered four alternative measures to determine which was most highly correlated with
comprehension ability. Three of the four alternatives were commonly used direct measures of
reading comprehension in the classroom and included answering questions about a passage,
passage recall, and completing cloze statements. The fourth measure was a test of oral reading
fluency and required students to read a 400-word passage aloud. Interestingly, criterion validity
coefficients generated through Pearson correlations revealed that oral reading fluency showed the
highest correlation (r = 0.91) with the criterion measure of reading comprehension. It is notable
16
that oral reading fluency was more strongly correlated to the criterion comprehension measure
than traditional classroom methods of assessing comprehension, and corroborates theoretical
hypotheses of oral reading fluency as an indicator of overall reading competence. Furthermore, a
large body of research documents positive gains made by students undergoing interventions that
target repeated reading practices on overall reading ability (Ari, 2011; Kuhn & Stahl, 2003;
Meyer & Felton, 1999; Therrien, 2004; Wolf & Katzir-Cohen, 2001). For instance, a meta-
analysis complied by the National Reading Panel (2000) reported moderate effect sizes for gains
across word knowledge (0.55), fluency (0.44) and comprehension (0.35). Wolf and Katzir-Cohen
(2001) suggest that reading interventions focusing on fluency training may become increasingly
important as phonological-based interventions prove inadequate to address the variety of deficits
that impede successful reading.
Oral reading fluency is typically assessed using curriculum-based measurement (CBM)
whereby students are required to read aloud a passage of text as accurately and quickly as possible
(Fuchs et al., 2001). These measures were originally developed to track reading progress among
students as well as to provide a tool to screen for reading disabilities (Christ, Zopluoglu, Long &
Monaghen, 2012). Achieving fluent reading presents a challenge for many students, though most
are able to attain some proficiency by Grade 2 or 3 (Fuchs et al., 2001). Oral reading fluency
growth as indexed by CBM scores have been shown to follow a distinct trajectory validated by
empirical research. Early studies evaluating fluency trajectories indicate positive growth that is
most substantial in the primary elementary grades (Deno, Fuchs, Marston & Shin, 2001; Fuchs,
Fuchs, Hamlett, Walz & Germann, 1993). By the intermediate grades and into junior high school,
rates of growth are shown to decline dramatically. This negatively accelerating curve in higher
grades is understood to reflect increasingly complex reading demands that require literary analysis
and processing of complex text. Further, the positive impact of word reading skills on reading
17
comprehension seems to decrease by the upper elementary years (Lesaux, Crosson, Kieffer &
Pierce, 2010). Therefore, the role of oral reading fluency as an indicator of reading proficiency is
most salient in the primary grades (Wayman, Wallace, Wiley, Ticha & Espin, 2007).
However, within the early primary grades, there is great variability in performance among
students with vast leaps in development occurring where literacy acquisition is concerned.
Students are typically unable to read connected text until around Grade 2, and as such, reading
fluency in very early readers is often assessed using word lists. Empirical evidence suggests that
preliminary word-level fluency is strongly correlated with later text-level fluency ability in
children (Good, Simmons & Kame’enui, 2001; Jenkins, Fuchs, van den Broek, Espin & Deno,
2003). However, some evidence suggests that oral reading fluency as measured by reading
isolated list words is qualitatively different from passage-level reading fluency. In a study by
Jenkins et al. (2000), criterion validity of word versus text fluency was examined when compared
to reading comprehension scores taken from a standardized assessment. Using the Reading
Comprehension portion of the Iowa Test of Basic Skills as the criterion measure, the authors
assessed passage-level reading fluency and word-level reading fluency to determine how these
skills correlated with comprehension ability. The correlation coefficients between reading
comprehension and word-level and text-level fluency were 0.53 and 0.83 respectively, with the
difference reaching statistical significance. Further, additional regression analysis revealed that
text-level fluency uniquely explained 42% of comprehension scores whereas word-level fluency
explained an additional 1% only. This suggests that though word and text-level fluency may be
highly correlated, text-level fluency is more closely related to overall reading ability as indexed
by comprehension.
18
Furthermore, in a study by Klauda and Guthrie (2008), the authors investigated the unique
contribution of fluency at the word-level, syntactic-level and passage-level on reading
comprehension outcomes within a single language. The authors describe the importance of word-
level fluency as grounded in the automaticity theory and assert that as word recognition becomes
more automatic, attentional demands once devoted to word decoding can be redirected to higher-
level comprehension. Syntactic-level fluency focuses primarily on the fluent processing of
language cues at the phrase or sentence-level that contribute to reading outcomes. This is
supported by research showing that the intervention method of repeated readings (Samuels, 1979)
leads to gains in reading ability, not because of improvements in word recognition but because it
teaches readers to capitalize on syntactic, semantic, morphological and contextual cues in written
text (Dowhower, 1987; Schreiber, 1980). Lastly, passage-level features of reading fluency are
shown to additionally contribute to reading outcomes. This refers to the macrostructure of the text
(Kintsch & Kintsch, 2005) and specifically, points to the processing of features that allow the text
to have an overall organization and coherence. Klauda and Guthrie (2008) showed that after
accounting for control variables, word reading speed explained an additional 10% of variance,
syntactic-level fluency contributed an additional 5% and passage-level fluency accounted for an
additional 2% of variance.
More recently, Geva and Farnia (2012) examined how the construct of reading fluency
changes across time and secondly, the extent to which reading fluency was related to language
comprehension across the primary grades. Students were assessed across a variety of cognitive,
oral language and reading measures, including word-level and text-level reading fluency. A
discriminant factor analysis showed that while word-level and text-level reading fluency loaded
onto a single construct in Grade 2, by Grade 5 these skills yielded two separate factors.
Furthermore, between Grade 2 and Grade 5, the correlation between reading fluency and word
19
identification decreased, while the correlation between text-level fluency and oral language
measures increased. These results lend support to the hypothesis that word and text-level reading
fluency becomes more differentiated over time, with text-level reading fluency aligning more
closely to higher-level language skills (Geva & Farnia, 2012). These studies show that different
levels of the fluency construct contribute differentially to reading outcomes and therefore, an
examination of transfer patterns at each individual level is warranted. Therefore, in the present
study, fluency was assessed across levels of the construct independently in order to account for
the possibility that they are dissimilar and may show different patterns of transfer.
2.6 Cross-Language Transfer of Oral Reading Fluency
Despite its important role in acquiring literacy, relatively little is known regarding cross-
language transfer of oral reading fluency ability. There are few studies that have rigorously
examined its transfer between languages, particularly among students enrolled in language
immersion programs. Although there is a lack of research investigating transfer, there is
substantial evidence showing that fluency ability between languages is highly correlated. For
instance, a four-year longitudinal study conducted by Jared, Cormier, Levy and Wade-Woolley
(2011) examined English language predictors of English and French reading development in a
group of French immersion students from kindergarten to Grade 3. Of interest to this discussion,
findings indicated that cross-language correlations were higher for the reading fluency measures
than for the reading accuracy measures. For instance, word reading fluency correlations were
found to be highly related between English and French in Grade 1 (r = 0.77), Grade 2 (r = 0.74)
and in Grade 3 (r = 0.70). In another study using a longitudinal design, Geva, Wade-Woolley and
Shany (1997) examined the development of accurate and speeded reading processes in students
learning to read English and Hebrew simultaneously. Students were assessed on parallel measures
of letter naming, reading isolated words and reading words in text. The results indicated that both
20
reading speed and reading accuracy were highly correlated in both Grade 1 and Grade 2. The
authors concluded that their findings support previous research on common underlying
proficiencies in reading development that occur in both the first and the second language.
Similarly, Ramirez (2001) reported that Spanish and English oral reading fluency was
significantly correlated (r = 0.72) among Spanish students learning English. These studies suggest
that reading fluency measures are highly related across different language and as such, may be a
likely candidate for transfer.
Though these findings indicate that reading fluency is highly correlated across languages,
only a handful of studies to date have examined transfer of reading fluency ability. For example,
Dominguez de Ramirez and Shapiro (2007) examined oral reading fluency in Spanish-speaking
English language learners studying in bilingual education classrooms in Grade 1 through Grade 5.
The authors were interested in investigating whether reading fluency in the first language was
predictive of later reading outcomes in a second language. Students were administered
curriculum-based measures in reading which assessed their accuracy and speed on both English
and Spanish passages. Results collapsed across grades indicated that Spanish and English reading
performance was highly correlated. Further, regression analysis indicated that Spanish oral
reading fluency was a significant predictor of later English oral reading fluency. The authors
concluded that their study demonstrated evidence for the existence of transfer, and further, that
assessing reading fluency in one language has predictive power in later fluency ability in another
language. However, within-language measures known to be predictive of reading ability were not
controlled for in the study design, and as such, it is unclear whether the cross-language
associations may have been due to these spurious variables. A recent study conducted in German-
English immersion schools (Gebauer, Zaunbauer & Moller, 2013) examined cross-language
pathways of reading fluency and reading comprehension in a sample of Grade 3 and Grade 4
21
students. The findings revealed moderate correlations across languages in both of these constructs
at all time points measured. Furthermore, structural equation modeling revealed cross-language
associations between reading comprehension and reading fluency, with an overall dominance of
transfer from the L2 (English) to the L1 (German) language. The authors concluded that these
skills necessary for successful reading can evidently transfer between languages, though transfer
of fluency ability between languages and across time was not specifically examined.
2.7 Rationale for Present Study
Although oral reading fluency is recognized as an indicator of overall reading competence,
less research has focused on its development and transfer. Furthermore, while linguistic and
cognitive benefits of multiple language acquisition are well documented in studies with bilingual
students, less attention has been directed towards students in a language immersion context. The
extant evidence for cross-language transfer of linguistic skills has focused predominantly on other
abilities, leaving a gap in the literature with respect to transfer of oral reading fluency. Given its
increasingly prominent role as an indicator of reading ability, it is of importance to examine
whether oral reading fluency operates in a cross-linguistic manner such that reading accuracy and
speed in a first language is predictive of the corresponding ability in a second language.
This thesis attempts to address this research gap by examining the development and
transfer of oral reading fluency in a language immersion context. First, it addresses the
development of oral reading fluency by charting the acquisition of fluent reading in both English
and French across one year. Furthermore, it addresses whether bi-directional cross-language
transfer exists in oral reading fluency among language-immersion students. Using both cross-
sectional and longitudinal data of students enrolled in an early French immersion program, both
word-level and text-level transfer of oral reading fluency was examined in three separate studies.
22
1) Study 1 analyzed longitudinal data to chart the development of word-level fluency in
English and French across Grade 2 and Grade 3. This study also examined whether bi-
directional cross-language transfer exists between Grade 2 word-level fluency and Grade 3
word-level fluency across English and French.
2) Study 2 analyzed longitudinal data to examine whether bi-directional cross-language
transfer exists between Grade 2 word-level fluency and Grade 3 text-level fluency across
English and French.
3) Study 3 analyzed cross-sectional data to examine whether bi-directional cross-language
transfer exists in Grade 3 text-level fluency across English and French.
3 Study 1: Word-Level to Word-Level Fluency
This study examined longitudinal data of primary students enrolled in a French immersion
program in Grade 2 and Grade 3. This study charted the development of word-level fluency
ability in English and French, and secondly, investigated whether cross-language transfer exists in
word-level fluency ability. Fluency ability at the word-level examines both accuracy and speed of
reading isolated whole words, and is positively correlated with reading outcomes (Fuchs et al.,
2001). It was hypothesized that word-level fluency in one language would be highly correlated
with the corresponding ability in the other language, and further, that transfer of this ability would
be seen between English and French measures across the two time points.
3.1 Participants
The participants in this study were part of a large-scale longitudinal project that began
when the students were in Grade 2. Students were recruited from an elementary school situated in
a large metropolis in Ontario, Canada where English is the dominant language. The school is
located in a multi-ethnic area in a predominantly middle-class neighbourhood and adopts an early
French Immersion program. From Grades 1 to 3, French is the language of instruction across all
23
subjects throughout the school day. Students were tested at two time points – in the spring of their
Grade 2 year (T1) and in the spring of their Grade 3 year (T2).
At time 1, 83 students participated in the data collection. The following year, attrition was
relatively low and a total of 75 students participated in the final data collection. Attrition occurred
due to students moving schools or dropping out of the language immersion program. Students
who dropped out of the language immersion program had scores within 1.5 standard deviations of
the mean across all measures assessed at time 1, indicating that they were not academically
weaker than the students who continued to participate in the program the following year. For this
study, data was analyzed only for students who participated in data collection at both time points
spaced one year apart. Thus, the final sample consisted of one cohort of 75 students (mean age =
86.11 months at T1, 38 girls). Of the 75 students, 38 met criteria for consideration as ‘English as
First Language’ (EL1) students and the remaining 37 were considered ‘English Language
Learners’ (ELL). French immersion students typically have minimal exposure to the French
language in their home environment, though many students come from a diverse background of
ethnicities. The native language groups represented in the sample include Farsi, Azerbaijan,
Spanish, Hebrew, Russian, Korean, Mandarin, Serbian, Hungarian, Turkish, Punjabi and
Japanese. Students were classified as ‘English as First Language’ (EL1) if their parents were born
in Canada or another English-speaking country and identified English as the language most often
spoken in the home. Students were classified as ‘English Language Learners’ (ELL) if their
parents were born outside of Canada and identified a language other than English as the language
most often spoken in the home (more than 60% of the time). Two EL1 students and four ELL
students were born in a non-English speaking country, though all children were reported to have
arrived in Canada by 2.5 years of age. Data on 30 of the 38 EL1 children indicated that the mean
age at which they first spoke English was 15.18 months (SD = 6.72), and data from 36 of the 37
24
ELL children indicated a mean age of 35.31 months (SD = 12.28). In a one-way ANOVA, the
mean age at which English was first spoken reached statistical significance between EL1 and ELL
students (F (1,64) = 64.40, p = 0.00). Lastly, the highest level of maternal education was
compared between the two language status groups. Among EL1 children, the highest percentage
of students (68%) had mothers who attained a college or undergraduate university degree, and the
second highest proportion (30%) obtained a post-graduate degree. Among ELL students, the
highest proportion (41%) reported obtaining a post-graduate degree followed closely by those
who attained a college or undergraduate university degree (35%).
3.2 Measures
Demographic Questionnaire: A questionnaire was distributed to the parents of all participating
students and completed in the English language. The questionnaire assessed a variety of
demographic variables including early language development (age at which student spoke their
first words), level of parental education (highest level of education received), and language use in
the home (language most often spoken by each member of the family). Data was coded and
analyzed to determine whether there were existing differences between groups in socioeconomic
status, home literacy and language use.
Nonverbal Reasoning: The Matrix Analogies Test (Naglieri, 1985) is a standardized measure of
nonverbal reasoning and includes 64 matrices that are organized into 4 subtests. Within each
subtest, students are presented with incomplete visuospatial patterns and required to select one of
six options that accurately completes each matrix. A discontinue rule of four consecutive incorrect
responses was utilized for each subtest. The total number of correctly completed matrices across
all four subtests was used as the nonverbal reasoning score for analysis. The Cronbach’s alpha
reliability rating of this measure was 0.94.
25
Phonological Awareness: Phonological awareness was measured using the Elision subtest of the
Comprehensive Test of Phonological Processing (Wagner et al., 1999a) in English and a parallel
experimental measure that was adapted for French. This assesses awareness of the sound structure
of one’s language and requires students to remove isolated sound segments from spoken words at
increasingly higher levels of linguistic complexity. A discontinue rule of three consecutive
incorrect responses was used for this measure. In each language, the total number of correct
responses was used as the phonological awareness score for analysis. The Cronbach’s alpha
reliability rating was 0.91 for English and 0.95 for French.
Rapid Automatized Naming (RAN): Rapid Automatized Naming ability was assessed using the
Digits subtest from the Comprehensive Test of Phonological Processing (Wagner et al., 1999a) in
English, and by a parallel experimental measure created to assess the same ability in French. This
measures speed of rapid serial naming and assesses the ability to quickly retrieve phonological
information from long-term memory. The task requires students to read two 9 x 4 matrices of
digits as quickly and accurately as possible. In each language, the total amount of time (in
seconds) required to name the digits across the two matrices was used for analysis. Cronbach’s
alpha reliability rating was 0.84 for ages 7-8 in the testing manual (Wagner et al., 1999a).
Word Level Reading Fluency: Word-level reading fluency was assessed using the Test of Word
Reading Efficiency (Wagner, Torgesen & Rashotte, 1999b) in English and a parallel experimental
measure that was created in French. The French task was a translated version of Form B from the
English TOWRE assessment. This task assesses the ability to read isolated sight words out of
context and with automaticity. Students are asked to read aloud from a list of sight words that
increased in difficulty and length. In each language, the total number of correctly read items
26
within 45 seconds was used as the word-level reading fluency score for analysis. The Cronbach’s
alpha reliability rating was 0.97 for the English version and 0.96 for the French version.
3.3 Procedure
English consent forms were distributed in each participating classroom at the beginning of
Grade 2. For this study, children with parental consent were tested at two points spaced one year
apart – in the spring of Grade 2 and the spring of Grade 3. Testing was conducted by trained
undergraduate and graduate students who travelled to the elementary school to test participants in
a quiet room during the school day. English instructions were used to explain the English
measures, and both English and French instructions were given for the French measures to ensure
students understood each task. The order of the measures administered within a session was
randomized for each student. On average, students required approximately 1 hour to complete all
measures in both languages.
3.4 Results
3.4.1 Preliminary Analyses
Outlier analysis was performed for all variables and any data points that were 2.5 standard
deviations from the mean were removed. This resulted in a final sample of 66 students (mean age
= 86.13 months at T1, 32 girls, 33 EL1s and 33 ELLs) included in all analyses. All independent
and dependent variables were checked for normality, skewness and kurtosis. Several variables
were identified as being skewed and transformations on the data were computed to correct for
normality. For these measures, logarithmic and square root transformations were used to correct
the normality of the distributions. Analyses using the transformed and untransformed data
produced identical results, and as such, results from the untransformed data are presented. All
ANOVAs and regressions satisfied the assumptions of normality, homogeneity and independence.
27
3.4.2 Main Analyses
Table 1 (see Appendix A) presents the mean and standard deviations of the raw scores for
all relevant measures administered in the spring of Grade 2 (T1) and the spring of Grade 3 (T2). A
Pearson correlation matrix is presented in Table 2 (see Appendix B) and indicated that most
variables were significantly correlated. Notably, between-language scores across each language
skill were strongly correlated whereas within-language correlations between different linguistic
skills were moderate at best. In particular, fluency between languages and across time was
strongly associated with correlations between 0.72 to 0.84. This suggests that within the same
linguistic skill, proficiency in one language was highly related to proficiency in the other, and
provides some preliminary support for cross-language transfer between English and French.
The first set of analyses was conducted to determine whether performance on the word-
level fluency task was significantly impacted by language status and time. Two repeated-measures
ANOVAs were conduced with a between-subject factor of ‘language status’ (EL1 vs. ELL) and a
within-subject factor of ‘time’ (Grade 2 vs. Grade 3). The ANOVAs were conducted on English
word fluency and French word fluency ability separately. The results of the analysis revealed no
significant differences between language status groups on either English (F (1,64) = 0.859, p >
0.05) or French fluency (F (1,64) = 0.096, p > 0.05), indicating that EL1 and ELL students
performed comparably on word level fluency in both languages. Furthermore, a significant main
effect of time was found for both English (F (1,64) = 32.08, p = 0.00) and French (F (1,64) =
87.41, p = 0.00) fluency. Performance in word fluency ability significantly improved across time
from Grade 2 to Grade 3 in both English (Grade 2 mean = 59.23, SD = 17.19; Grade 3 mean =
66.67, SD = 11.46; p = 0.00) and in French (Grade 2 mean = 52.29, SD = 13.36; Grade 3 mean =
62.47, SD = 11.03; p = 0.00). An interaction variable between time and language status was not
found to be significant in either ANOVA for English or French word fluency ability.
28
Hierarchical regression analyses were conducted to identify cross-language transfer of
word-level reading fluency ability in English and French. In both analyses, non-verbal reasoning
and language status were entered in the first two steps as control variables. Next, within-language
phonological awareness and rapid naming were entered as within-language control variables in
steps 3 and 4. Within-language word reading fluency was entered next to control for the
autoregressor effect of Grade 2 word reading fluency on Grade 3 ability. Of greatest interest,
cross-language word reading fluency was entered as the next step in the model to test for cross-
language transfer. Lastly, an interaction term between language status and cross-language word
reading fluency was entered into each model, but was dropped from all regressions as it was not
found to be significant.
The regression analysis predicting Grade 3 English word reading fluency is presented in
Table 3 (see Appendix C) and explained 69% of total variance. In this model, non-verbal
reasoning accounted for 9% of variance while language status did not significantly contribute to
the model. Within-language control variables were also found to be significant predictors of
Grade 3 English word reading fluency. Grade 2 English phonological awareness accounted for
12% of additional variance and Grade 2 English rapid naming ability predicted an additional 13%.
Therefore, proficiency in phonological awareness and rapid naming ability in Grade 2 was
associated with higher scores on Grade 3 word reading fluency. The autoregressor variable of
Grade 2 English word reading fluency contributed an additional 33% of variance to the model.
Lastly, Grade 2 French word reading fluency was a significant predictor of Grade 3 English word
reading fluency and contributed an additional 4% to the model. Thus, there was a significant
though small cross-language transfer effect of word reading fluency from French to English.
29
The regression analysis predicting Grade 3 French word reading fluency is shown in Table
4 (see Appendix C) and explained 63% of total variance. Non-verbal reasoning accounted for 6%
of variance while language status did not significantly contribute to the model. Within-language
control variables were significant predictors with Grade 2 French phonological awareness
explaining an additional 9% of variance and Grade 2 French rapid naming ability explaining 14%
of unique variance. The autoregressor variable of Grade 2 French word reading fluency
significantly predicted an additional 31% of variance. Lastly, the cross-language transfer variable
was found to be a significant predictor of Grade 3 French word reading fluency. Performance in
Grade 2 English word reading fluency explained an additional 6% of unique variance above
within-language controls providing evidence for cross language transfer from English to French.
In sum, these analyses demonstrate a significant though small bidirectional cross-language
transfer effect of word-level reading fluency between English and French.
A commonality analysis was conducted to supplement these regressions by specifying the
amount of common variance associated with each combination of predictors (Capraro & Capraro,
2001). The proportion of shared variance between variables can be interpreted as the amount of
variance in the dependent measure that can be explained equally well by any one predictor in the
group. Tables 5 and 6 (see Appendix D) present the commonality data that predict variance in
Grade 3 English word reading fluency and Grade 3 French word reading fluency. In both
analyses, among all combinations of predictors, the fluency constructs shared the greatest
proportion of variance in explaining the dependent measure. The fluency constructs jointly
accounted for 35% of common variance in Grade 3 English word reading fluency and 38% of
common variance in Grade 3 French word reading fluency.
30
3.5 Discussion
The goal of Study 1 was to investigate the development of word-level fluency ability in
English and French across a one-year period from Grade 2 to Grade 3 within a French immersion
population. Secondly, this study examined whether cross-language transfer of oral reading fluency
at the word-level operates between the English and French languages. The results of the study
demonstrated that performance in word-level fluency ability significantly improved over the one-
year period of time in English and in French. This indicated that despite receiving formal literacy
instruction in the French language only, students’ fluency ability showed significant improvement
in both languages. This may be explained by the development of common proficiencies that
underlie fluency ability across different languages, such as the synchronization of phonological,
orthographic and semantic processes involved in reading (Breznitz, 2003; Breznitz, 2006). Lastly,
it is of importance to note that the language status of the students did not impact their
performance, indicating that speaking other language(s) in the home environment did not appear
to impede English or French language development. Concerns have been expressed regarding
whether minority language children in French immersion programs may experience incomplete
development of all their languages if they are instructed solely in French (Genesee & Jared,
2008). The results of this study address some of these concerns and add to a growing body of
literature in support of language immersion education. In this study, students who spoke a non-
English language in the home demonstrated equivalent performance across all linguistic measures
when compared to students who were raised in English-speaking households. This is consistent
with other research suggesting that speaking other language(s) in the home environment does not
negatively impact learning English and French in a French immersion program (Cummins, 1998).
Furthermore, the results of this study demonstrated that significant bidirectional cross-
language transfer exists in word-level reading fluency between the English and French languages.
31
After controlling for the contribution of within-language predictors and autoregressor effects,
French word-level reading fluency ability predicted an additional 4% of variance in English word-
level reading fluency one year later. Similarly, using the same stringent controls, English word-
level reading fluency predicted an additional 6% of variance in French word-level reading fluency
one year later. In addition, the supplementary commonality analysis indicated that a large amount
of variance in word-level fluency ability is shared between English and French word-level
fluency. Taken together, this suggests that fluency ability is largely shared between languages,
making it possible for performance in one language to be predictive of performance in another
language, even when measured across time. This indicates that transfer occurs such that fluency
skill in one language can have a positive facilitatory effect on the corresponding ability in another
language. By using longitudinal data and within-language controls, these findings extend the
current literature by providing evidence for cross-language transfer of word-level fluency ability
between these two alphabetic languages using a stringent methodological design.
4 Study 2: Word-Level to Text-Level Fluency
This study examined longitudinal data of primary grade students enrolled in a French
immersion program and investigated whether cross-language transfer exists between the word-
level and the text-level of fluency ability. Fluency ability at the word-level examines accuracy and
speed of reading isolated whole words, and at the text-level, assesses the same ability in reading
connected passages of text. To date, no other known research has systematically investigated
transfer of fluency ability at the word-level and the text-level separately. This study attempts to
fill this gap in the literature by differentiating between these two levels of the fluency construct
and examining transfer from one level to the other. Some evidence suggests that word-level and
text-level fluency may operate differently as predictors of reading ability (Geva & Farnia, 2012)
32
and as such, may display differential patterns of transfer. However, though findings suggest that
word-level and text-level fluency become increasingly differentiated, the association between
word-level and text-level fluency within the same language is still found to be significantly strong
with correlation values exceeding 0.90 (Geva & Farnia, 2012). Therefore, it was hypothesized that
word-level fluency in one language would be highly correlated with text-level fluency ability in
the other language. The high degree of correlation between word-level and text-level fluency
suggests that these constructs rely on some common underlying processes. Though text-level
fluency may benefit additionally from higher-level linguistic processes, both word-level and text-
level fluency are dependent on the automatization of word-level skills such as the processing of
orthographic units, the decoding and recoding of phonological units, and further, are aided by
speed of lexical access (Fuchs et al., 2001). Therefore, the skills implicated in these two
constructs are not mutually exclusive, but rather, each construct may rely on both shared and
distinct abilities. As such, it was also hypothesized that some evidence for transfer would likely be
seen between English and French word-level to text-level fluency across the two time points.
4.1 Participants
For this study, longitudinal data was analyzed for students who participated at two time
points – in the spring of Grade 2 and the spring of Grade 3. The final sample in Study 2 was
identical to the sample used in Study 1 and consisted of one cohort of 75 students (mean age =
86.11 months at T1, 38 girls, 38 EL1s and 37 ELLs). Demographic information for the final
sample of students is provided in the Participants section of Study 1.
4.2 Measures
This study used the same measures as described in Study 1. An additional measure
assessing text-level oral reading fluency was given in the spring of Grade 3.
33
Text Level Reading Fluency: Text-level reading fluency was assessed using the Oral Reading
Fluency subtest of the Wechsler Individual Achievement Test – Third Edition (Wechsler, 2009) in
English and the Test de Lecture: 3e Annee (Barik & Swain, 1979) in French. Students were
presented with two narrative passages (average words per passage: English = 109, French = 108)
in each language and instructed to read the text aloud as quickly and accurately as possible. Any
mispronunciations, omissions, additions or substitutions of words were marked as errors. If
students hesitated on a word for more than 3 seconds, it was provided by the examiner and
marked as an error. Words that were spontaneously self-corrected within 3 seconds were not
considered errors. The time in seconds taken to read each passage was recorded. The number of
words read correctly per minute was determined for each language and used as the text-level
reading fluency score for analysis. Words correct per minute (WCPM) was calculated by the total
number of words read correctly divided by the total amount of time in minutes to read the two
passages. The Cronbach’s alpha reliability rating for ages 7-8 was 0.92 in the testing manual for
the English measure (Wechsler, 2009) and 0.85 in the testing manual for the French measure
(Barik & Swain, 1979).
4.3 Procedure
The experimental procedure was identical to Study 1 with the inclusion of a text-level oral
reading fluency measure administered in both languages in the spring of Grade 3. Consistent with
the administration of the other measures, English instructions were used to explain the English
version of the text-level reading fluency measure, and both English and French instructions were
given for the French version of the measure.
34
4.4 Results
4.4.1 Preliminary Analyses
Preliminary analyses were conducted in an identical manner as the previous study and
descriptions are provided in the Results section of Study 1. All regressions satisfied the
assumptions of normality, homogeneity and independence.
4.4.2 Main Analyses
Table 1 (see Appendix A) presents the means and standard deviations for the raw scores of
all relevant measures administered in the spring of Grade 2 (T1) and the spring of Grade 3 (T2). A
Pearson correlation matrix is presented in Table 7 (see Appendix E) and show that the majority of
variables were significantly correlated. Notably, fluency variables were significantly correlated
with one another, both between levels of fluency (correlations between 0.73 to 0.75) and between
languages (correlation between 0.76 to 0.83).
Hierarchical regression analyses were conducted to test for cross-language transfer from
word-level to text-level fluency ability between English and French. In both analyses, the control
variables nonverbal reasoning and language status were entered in the first two steps. Next,
within-language phonological awareness and rapid naming ability were entered as steps 3 and 4.
Within-language word reading fluency was entered next to control for the effect of Grade 2 word
reading fluency on Grade 3 text reading fluency ability. Word-level reading fluency was used to
control for the autoregressor effect of fluency ability because text-level fluency was not assessed
in Grade 2. Of greatest interest, Grade 2 cross-language word reading fluency was entered to test
for cross-language transfer from word-level to text-level reading fluency ability. Lastly, an
interaction term between language status and cross-language word reading fluency was entered
into each model but was not found to be significant and thus, dropped from all regressions.
35
The regression analysis predicting Grade 3 English text reading fluency is shown in Table
8 (see Appendix F) and accounted for 66% of total variance. Nonverbal reasoning accounted for
17% of variance while language status did not significantly contribute to the model. The within-
language controls were significant predictors of Grade 3 English text reading fluency with Grade
2 English phonological awareness accounting for an additional 8% of variance and Grade 2
English rapid naming ability explaining 10% of unique variance. Grade 2 English word reading
fluency explained 25% of variance in the model and is highly associated with Grade 3 reading
fluency ability. Notably, Grade 2 French word reading fluency was found to be a significant
predictor as well, explaining 10% of additional variance above the stringent controls. Thus, there
was significant transfer of word-level to text-level reading fluency ability from French to English.
The regression analysis predicting Grade 3 French text reading fluency is shown in Table
9 (see Appendix F) and accounted for 55% of total variance. Nonverbal reasoning and language
status controls were not found to be significant predictors in this model. Grade 2 French
phonological awareness was a marginally significant predictor, and Grade 2 rapid naming ability
explained an additional 14% of unique variance. Similar to the above regression, Grade 2 French
word reading fluency was the strongest predictor of Grade 3 text reading fluency, explaining 34%
of unique variance in the model. Lastly, Grade 2 English word reading fluency did not contribute
any unique variance to Grade 3 French text reading fluency. Thus, there was significant cross-
language transfer of word-level to text-level fluency from French to English, but no evidence for
transfer from English to French.
Tables 10 and 11 (see Appendix G) present the commonality data for the two criterion
measures in this study. In examining the shared contributions of predictor pairs, the word-level
fluency constructs across the two languages share the largest proportion of common variance. In
36
predicting Grade 3 English text-level fluency, the constructs of Grade 2 word-level fluency in
English and in French explained a total of 33% of shared variance in the criterion measure.
Similarly, in explaining variance in Grade 3 French text-level fluency, these same constructs
(Grade 2 English and French word-level fluency) shared a total of 33% of variance, which is the
largest mutually exclusive partition of variance in the dependent measure.
4.5 Discussion
The goal of Study 2 was to investigate whether cross-language transfer of oral reading
fluency ability occurred from the word-level to the text-level in English and French. To date, very
little research distinguishes between the two levels of this construct, and no research has
investigated transfer across levels of this ability. Using a rigorous methodological design, this
study attempted to address this gap by examining transfer of fluency ability across levels of the
construct using longitudinal data. The results of this study demonstrated that transfer exists from
the word-level to the text-level of fluency ability from French to English only. Controlling for
within-language predictors of fluency and autoregressor effects, performance on French word-
level fluency significantly predicted an additional 10% of variance in English text-level fluency
measured one year later. In addition, supplementary commonality analysis indicated that the
largest partition of shared variance in text-level fluency outcomes was shared between English
and French fluency performance. Taken together, this study provides evidence for some transfer
across levels of fluency ability that occurs between languages and across time.
Interestingly, the crossover effect of fluency ability was found only from French to
English, but not in the other direction. Though traditional interpretations understood transfer to
occur primarily from the majority to the minority language, a considerable amount of research has
shown that transfer of literacy-related skills can occur in both directions (Cashion & Eagan, 1990;
37
Harper, 2010; Verhoeven, 1991). However, evidence of bidirectional transfer has not been
consistently reported, particularly within language immersion populations. Rather, some studies
have found immersion-specific relations that show a dominance of transfer from minority
(immersion language) to majority language instead (Gebauer et al., 2013). In an extension to his
original interdependence theory, Cummins (1998) surmised that “transfer may be more likely to
occur from minority to majority language because of the greater exposure to literacy in the
majority language outside of school and the strong social pressure to learn it,” referring
specifically to the unique directional transfer observed in an immersion context. Therefore, the
unidirectional transfer observed in this study can be understood within this immersion-specific
context as well. The unidirectional effect from minority (French) to majority language (English) is
likely related to the fact that French immersion students in Grade 2 and Grade 3 receive formal
instruction in the French language only, but receive adequate exposure to the majority language in
their home and community environments. This finding suggests that shared skills in word –level
and text-level reading fluency can be acquired in a minority language context and may transfer to
the majority language, even when there is limited formal exposure to the majority language in an
academic setting. The fact that transfer occurred after limited formal exposure to the English
language may be due to the abundance of literacy exposure in the majority language outside of
school (Cummins, 1998). The students in this study are exposed to the English language in a
variety of informal ways outside of the school environment (e.g., exposure in the home
environment, peer interactions, extra-curricular activities, and media outlets including radio,
television, internet etc.), which can account for the dominance of minority to majority language
transfer that was observed in this study.
38
5 Study 3: Text-Level to Text-Level Fluency
This study examined cross-sectional data of primary grade students enrolled in a French
immersion program in Grade 3. This study evaluated text-level reading fluency in English and in
French, and further, investigated whether cross-language transfer exists in text-level fluency
across the two languages. Fluency ability at the text-level examines both accuracy and speed of
reading connected passages of text, and is purported to be an overall indicator of reading
proficiency (Fuchs et al., 2001). It was hypothesized that text-level fluency in one language
would be highly correlated with the corresponding ability in the other language, and further, that
transfer of this ability would be seen between English and French measures.
5.1 Participants
For this study, cross-sectional data was analyzed for students at one time point only – in
the spring of Grade 3. The final sample in Study 3 was identical to the sample used in the above
studies and consisted of one cohort of 75 students (mean age = 86.11 months, 38 girls, 38 EL1s
and 37 ELLs). Demographic information for the final sample of students is provided in the
Participants section of Study 1.
5.2 Measures
This study used the same measures as described in Study 1 with the inclusion of a text-
level oral reading fluency measure administered in both languages in the spring of Grade 3. A
description of the measure is provided in the Measures section of Study 2.
5.3 Procedure
The experimental procedure was identical to Study 1 with the inclusion of a text-level oral
reading fluency measure administered in both languages in the spring of Grade 3. Consistent with
the administration of the other measures, English instructions were used to explain the English
39
version of the text-level reading fluency measure, and both English and French instructions were
given for the French version of the measure. Study 3 used a cross-sectional design and as such,
the procedure was implemented for all measures collected in the spring of Grade 3 only.
5.4 Results
5.4.1 Preliminary Analyses
Preliminary analyses were conducted in an identical manner as the previous study and
descriptions are provided in the Results section of Study 1. All ANOVAs and regressions satisfied
the assumptions of normality, homogeneity and independence.
5.4.2 Main Analyses
Table 1 (see Appendix A) presents the means and standard deviations for the raw scores of
all relevant measures administered in the spring of Grade 3 (T2). Pearson correlations for all
measures are reported in Table 12 (see Appendix H) and indicate that most variables were
significantly correlated. Similar to Study 1, correlations between English and French scores of the
same linguistic skill were notably strong. The correlations between different linguistic skills in
each language individually were more variable, and fall in the weak to moderate range. In
particular, fluency between English and French was strongly associated with significant
correlations within levels of fluency (between 0.83 to 0.84) and across levels of fluency (between
0.67 to 0.79). The strong between-language correlations found in the data support the hypothesis
of cross-language transfer within the same linguistic skill in English and French.
The first set of analyses was conducted to determine whether performance in text-level
fluency was significantly impacted by language status at one time point. Two one-way ANOVAs
were conducted on English and French text-level fluency scores in Grade 3 with one between-
subject factor of ‘language status’. The analyses revealed no significant differences between the
40
two groups in either language (English: F (1,65) = 0.000, p > 0.05; French; F (1,65) = 0.632, p >
0.05), indicating that EL1 and ELL students demonstrated commensurate performance in text-
level fluency ability in both English and French.
Hierarchical regression analyses were conducted to identify cross-language transfer of
text-level reading fluency ability between English and French using cross-sectional data. In both
analyses, the control variables of nonverbal reasoning and language status were entered in as the
first two steps. Next, within-language predictors of oral reading fluency were entered to control
for variables known to be associated with fluency ability including phonological awareness, rapid
naming ability and word-level reading fluency. Of greatest interest, cross-language text-level
reading fluency ability was entered next to test for cross-language transfer of reading fluency
ability. Lastly, an interaction term between language status and cross-language reading fluency
was entered and dropped from all regressions as it was not found to be significant.
The regression analysis predicting Grade 3 English text reading fluency is presented in
Table 13 (see Appendix I) and explained 87% of total variance. Nonverbal reasoning accounted
for 17% of variance while language status did not significantly contribute to the model. Consistent
with the literature on the development of reading fluency, within-language control variables were
found to be significant predictors of Grade 3 English text reading fluency. Grade 3 English
phonological awareness accounted for 15% of additional variance, Grade 3 rapid naming ability
contributed an additional 5% and Grade 3 word-level reading fluency predicted 41% of unique
variance to the model. Therefore, proficiency in these within-language linguistic skills is highly
associated with higher scores on text-level reading fluency and combined, account for 75% of
variance in the model. Lastly, cross-language transfer was tested to examine whether cross-
language ability in text fluency predicted any additional variance beyond known within-language
41
predictors. Grade 3 French text reading fluency was found to significantly account for an
additional 11% of unique variance in the model, providing evidence for cross-language transfer of
text reading fluency from French to English.
The regression analysis predicting Grade 3 French text reading fluency is shown in Table
14 (see Appendix I) and explained 73% of total variance. In this model, the control variables
nonverbal reasoning and language status did not significantly account for any variance. Similar to
the above regression for English text reading fluency, within-language control variables were
found to be significant predictors of Grade 3 French text reading fluency. Grade 3 French
phonological awareness contributed an additional 22%, Grade 3 rapid naming ability accounted
for 11% of unique variance and Grade 3 word reading fluency predicted an additional 26% of
unique variance in the model. Consistent with previous findings, phonological awareness, rapid
naming ability and word reading fluency accounted for a combined 62% of variance in the model,
and are highly associated with proficiency in within-language text reading fluency. Lastly,
English text reading fluency significantly predicted an additional 11% of unique variance in the
model beyond the known within-language predictors of this ability. In sum, these analyses
demonstrate some evidence for bidirectional cross-language transfer of text-level reading fluency
between English and French.
Supplementary commonality analysis was conducted and results are presented in Tables
15 and 16 (see Appendix J). Notably, in both sets of analyses, the fluency constructs across
languages shared the largest proportion of common variance in the criterion measures. In
predicting Grade 3 English text-level fluency, the largest partition of common variance (31%) was
shared between the predictors Grade 3 English word-level fluency and Grade 3 French text-level
fluency. In predicting Grade 3 French text-level fluency, the constructs Grade 3 French word-
42
level fluency and Grade 3 English text-level fluency accounted for a similarly large proportion of
variance (31%). This suggests that across different languages (English and French) and different
levels of fluency (word-level and text-level), there is a significant amount of common explained
variance across fluency constructs and likely, draws on some common underlying ability that is
not language-specific.
5.5 Discussion
The goal of Study 3 was to investigate whether cross-language transfer of oral reading
fluency occurred at the text-level between English and French. First, the results of the study
demonstrated that the language status of the students did not impact text-level fluency
performance, indicating that speaking other language(s) in the home environment did not appear
to impede their English or French fluency development. Additionally, this study examined cross-
language transfer of oral reading fluency at the text-level between English and French. As
mentioned previously, few studies have investigated the different levels of this construct. Using a
cross-sectional design, this study provides evidence for significant bidirectional cross-language
transfer in text-level reading fluency between the English and French languages. Controlling for
within-language predictors of fluency ability, French text-level reading fluency predicted an
additional 11% of variance in concurrent English text-level reading fluency. Similarly, using the
same stringent controls, English text-level reading fluency explained 11% of unique variance in
concurrent French text-level reading fluency. Furthermore, the supplementary commonality
analysis indicated that the largest partition of shared variance in text-level fluency outcomes is
shared between English and French fluency. This suggests that text-level reading fluency is
strongly associated across languages, allowing performance in one language to be predictive of
concurrent performance in another language. Furthermore, performance on English and French
fluency tasks were highly correlated despite the fact that students received formal literacy
43
instruction in French only. This suggests that creating strong links among phonological,
orthographic and semantic processes in one language may have a positive and facilitatory effect
on fluent reading in the other language. It is also important to note the role of cognitive processes
such as phonological awareness and rapid automatized naming that underlie reading fluency tasks
across languages, and likely contribute to the high degree of correlation observed in this study.
These findings extend the extant literature on cross-language transfer by providing some evidence
for concurrent transfer of text-level fluency between these two alphabetic languages.
However, a major limitation of Study 3 was its cross-sectional design in examining
transfer of text-level fluency ability between English and French. Data across all variables was
measured at one time point only and as such, only tentative conclusions can be drawn with respect
to transfer. The results indicated strong cross-language associations between text-level reading
fluency ability across English and French at the same point in time, however, it is not possible to
speak to the predictive value of this ability across languages over time. Further research should
investigate the existence of transfer in text-level fluency ability using a longitudinal design to
ascertain whether text-level reading fluency can be predictive of the same ability measured at a
different point in time.
6 General Discussion
This set of studies provides evidence for cross-language transfer of oral reading fluency
ability among language immersion students in the primary grades. The results indicated that
bidirectional transfer between the two languages was evident at both levels of the fluency
construct when investigated separately, and unidirectional transfer from word-level to text-level
fluency was found from the L2 (French) to the L1 (English) language. Taken together, these three
studies demonstrate that reading fluency both at the word-level and text-level is highly correlated
44
between languages, and that this synchronization of phonological, orthographic and semantic
processes is subject to transfer between the English and French languages. In addition, it is
important to note the contribution of individual differences in underlying cognitive processes (i.e.,
phonological awareness, rapid automatized naming) that play a role in the high degree of
correlation between oral reading fluency in English and in French.
The results of these studies add to a large body of research documenting cross-language
transfer effects between various language pairs across many linguistic skills. Furthermore, this
study utilized a stringent research design, which controlled for known within-language predictors
of reading ability as well as auto-regressor effects of fluency ability. However, a closer discussion
considering what specific skill or ability is subject to transfer is certainly warranted here. Fluency
in reading is comprised of various building blocks of literacy development including decoding
and recoding of phonological units of sounds, processing of orthographic patterns, whole word
recognition and rapid lexical access (Fuchs et al., 2001). While proficiency in these individual
skills (i.e., phonological, orthographic and semantic processes) is certainly a precursor to
achieving fluency, the execution of these skills alone does not automatically result in fluent
reading. Rather, the crux of fluent reading is the synchronization of these component processes,
and the achievement of automaticity in the coordination of these individual processes can be
understood as a distinct skill. More specifically, Breznitz (2003; 2006) posits that rapid word
reading results from the successful and efficient integration of phonological, orthographic and
semantic processes. Through the gradual development of automaticity or ‘fluency’ in these
component skills, it is believed that more attentional resources will become available to engage in
higher-level linguistic processes that aid in fluent reading such as syntactic knowledge or
contextual facilitation. Therefore, it is only when the simultaneous coordination of these sub-
processes is achieved, that the four key features of fluent reading – speed, effortlessness,
45
autonomy, and achievement without consciousness or awareness - are realized (Kuhn et al.,
2010). This process of synchronization is likely to be similar across different languages, and in
this way, lends support to the idea of “procedural” knowledge that can be acquired in one
language and drawn upon to support reading development in another language. This can be
captured in Cummins’ (2012, p. 1980) theoretical explanation of developing a “deeper conceptual
and linguistic proficiency” that underscores academic tasks across different languages. This
ability may tap into a language-general ‘procedural’ skill that is understood to be acquired in one
language and permeate into the development of another language (Genesee et al., 2006). In
addition, the synchronization of skills involved in reading fluency can be further conceptualized
as a metalinguistic skill as described by Koda (2008), where prior experience in one language
systematically alters the processing of a similar procedure in a second language. For instance, the
recognition and formation of links between phonological, orthographic and semantic knowledge
that facilitate fluent reading in one language can be considered a skill that transcends any
particular language (i.e., a metalinguistic skill), and one that arguably can shape the processing of
fluent reading in another language. Metalinguistic ability involves an understanding of the rules,
patterns and processes that are inherent across languages in general, and research suggests that
bilinguals possess an advantage in these aspects of language processing over their monolingual
peers (Bialystok, 2001). This can be applied to an understanding of fluency in that a deeper
conceptual proficiency in the recognition and formation of connections between phonological,
orthographic and semantic aspects of language is likely to underlie fluent reading across
languages. This is the mechanism that is speculated to transfer in fluent reading, and the results of
these studies are consistent with this explanation of cross-language transfer in reading fluency.
An interpretation of the results of these studies would not be complete without also
considering underlying cognitive processes that may play a role in the interrelationships between
46
L1 and L2 reading fluency performance. Geva and Ryan (1993) argued that transfer may not be
the only mechanism that accounts for the high degree of correlation on similar language tasks in
different languages. Rather, the authors suggested that underlying cognitive skills that are part of
one’s innate cognitive endowment (i.e., working memory, phonological awareness, rapid
automatized naming) may also underlie performance on these language tasks (Geva & Ryan,
1993). With respect to reading fluency in particular, both phonological awareness and rapid
automatized naming are expected to be closely related to overall reading performance, and are
considered underlying cognitive abilities that facilitate fluent reading. Consistent with this view,
the results across all three studies conducted in this thesis indicated a high degree of correlation
on phonological awareness and rapid automatized naming across parallel languages tasks.
Furthermore, the results of the regression models indicated that within-language phonological
awareness and rapid automatized naming consistently predicted unique variance in reading
fluency performance. This lends support to the notion put forth by Geva and Ryan (1993), and
implicates the role of individual differences in underlying cognitive processes in partially
accounting for the relationship between parallel language tasks in L1 and L2. Therefore, while the
results of these studies provide evidence for transfer, it is evident that a consideration of other
factors (i.e., innate cognitive abilities that underlie performance on parallel tasks across
languages) is also necessary to adequately interpret the results.
To date, very little research distinguishes between levels of the fluency construct, and no
research has investigated transfer within and between levels of this ability separately. In the study
by Klauda and Guthrie (2008) discussed earlier, the authors showed the unique contribution of
fluency at the word-level, syntactic-level and passage-level on reading comprehension outcomes
within a single language. Therefore, it is evident that different levels of the fluency construct
contribute differentially to reading outcomes, which speaks to both the relatedness and
47
distinctiveness of these variables. Based on this premise, this thesis was the first to demonstrate
transfer at the word and text level of fluency independently, and further, showed evidence for
transfer between the word-level and the text-level of the construct as well. While bidirectional
transfer was evidenced when examining word-level fluency (Study 1) and text-level fluency
(Study 3) separately, the asymmetrical pattern of transfer shown across levels of fluency (Study 2)
provide further evidence that these constructs may not operate in the same way. Rather, a
dominance of L2 (French) to L1 (English) transfer was shown when examining for transfer from
the word-level to the text-level of fluency.
The differential patterns of transfer further suggest that word-level and text-level fluency
may be tapping into related, but distinct skills. Fluent reading is characterized by speed,
effortlessness and autonomy (Kuhn et al., 2010), though the influences that impede or facilitate
fluent reading of words versus passages of text may not be identical. In particular, higher-level
language skills play a significant role in text reading fluency, but not in word reading fluency
(Geva & Farnia, 2012). While it is certainly the case that word-level and text-level reading draw
upon many similar underlying skills (e.g., decoding and recoding of phonological units,
processing of orthographic units, whole word recognition), these constructs may also be impacted
by level-specific influences. For instance, empirical data informing models of top-down
influences on reading suggest that higher-level processes of language comprehension such as
syntactic knowledge, vocabulary, and listening comprehension are closely associated with reading
fluency at the text-level, but not at the word-level (Geva & Farnia, 2012; Rumelhart, 1994).
Therefore, while word-level fluency may rely more heavily on bottom-up decoding processes,
text-level fluency appears to additionally benefit from cues at the syntactic and passage-level of
text, and thus rely less on strict decoding ability (Geva & Farnia, 2012). This might partially
account for the fact that word to text level transfer of fluency ability occurred from French to
48
English only, as students in the study received formal direct instruction in early reading skills
such as decoding in the French language. Perhaps by Grade 3, students had developed sufficient
decoding skills in French such that word-level processes were able to transfer to facilitate English
reading, and thus, allow students to capitalize on contextual cues in the passages of text. It is
possible that the reverse was not true for English decoding ability, which may have prevented
transfer to French text reading fluency.
Furthermore, the unidirectional transfer from the L2 to the L1 may also be partially
explained by other theoretical accounts. This pattern of transfer is consistent with some existing
evidence for an immersion-specific effect with transfer occurring from a minority language
(immersion language) to a majority language (e.g., Gebauer et al., 2013). Cummins (1998) posited
that transfer is more likely to occur from the minority to the majority language due to abundant
exposure to literacy in the majority language outside of the school environment such as in the
home or in the media. It is very likely the case that the students involved in this study had
plentiful experiences for exposure to reading and writing in the English language in the home
environment. English is the dominant language in this large Canadian metropolitan city whereas
exposure to the French language for these students was largely confined to the school
environment. Further evidence for the abundant exposure of French immersion students to
English comes from studies suggesting that even during unstructured or recreational periods
within the school day (e.g., lunch, recess, physical education class etc.), students typically prefer
to use the English language in their interactions with one another (Tarone & Swain, 2011).
One serious limitation to this set of studies is that transfer was investigated between a
single language pair only. Therefore, although transfer was evidenced at different levels of the
construct between languages and across time, it is not possible to conclude that fluency ability is a
49
language-independent construct that shows universal transfer. The similarities between the
English and French languages which are both grounded in alphabetic scripts may have made
transfer effects more likely than if different language pairs were chosen. Research has found
strong associations in fluency ability between similar language pairs including Spanish-English
(Dominguez de Ramirez & Shapiro, 2007) and German-English (Gebauer et al., 2013), but also
among less similar languages such as Hebrew-English (Geva et al., 1997). Given the high degree
of correlation in fluency ability between languages, it can be hypothesized that transfer will likely
also occur between languages that differ more widely. Furthermore, if a theoretical understanding
of the automatization of fluent reading as an acquired, higher-level linguistic process is in fact
accurate, then transfer ought to be seen across a variety of language pairs. However, additional
research investigating transfer of fluency using language pairs that differ more widely in
phonology and orthography will be required to further ascertain whether fluency ability shows
universal language-independent transfer.
Another factor that limits the generalizability of these results is the sample of participants
from which the data was collected. The students who participated in these studies came from a
single school in a relatively high socioeconomic area of a Canadian metropolitan city. As a result,
parental education for the majority of students was relatively high and most participants came
from home environments with fairly high levels of English proficiency. This may have impacted
results of these studies that indicated no significant differences between EL1 and ELL students.
Perhaps, among students with lower levels of home literacy in the majority language, immersion
programs that are conducted solely in the French language may have differential impact on
students’ language development in English and French.
50
Despite its limitations, the primary contribution of these studies to the extant literature lies
in providing evidence for transfer both within and between different levels of oral reading
fluency. Further, the evidence suggests that word-level and text-level fluency are highly related,
but not identical constructs. Taken together, the results of these studies can contribute to an
existing body of literature that documents cross-language transfer effects across a variety of
linguistic skills and between large numbers of language pairs. In addition, this research may
contribute to a more comprehensive model of multilingual literacy development that accounts for
both language-specific and language-universal processes that operate when students learn to read
multiple languages simultaneously.
These findings have many important implications in educational and applied contexts as
well. Within school systems, this can inform best practices for teachers in predicting literacy
outcomes, especially in highly multilingual environments with students who may be acquiring
more than one language simultaneously. For instance, this can inform predicted trajectories of
language development for students who come from minority language backgrounds and are
acquiring a second or third language within our school systems. The results of this thesis
contribute to empirical evidence indicating that oral reading fluency is highly related between
language pairs and as such, reading fluency assessed in one language is highly predictive of
ability in another language. This could allow for fluency measures in one language to be used in a
diagnostic manner to inform the corresponding ability in a second language. In some situations
and with appropriate caution, this may even be used to aid in the identification of children who
are at risk for difficulties in learning to read in different languages. In these ways, current
practices that discourage systematic assessment of English language ability amongst recently-
arrived English language learners can be circumvented, and students can still be evaluated for risk
of reading difficulties. This can allow for earlier assessment and intervention to occur, rather than
51
allowing reading difficulties to compound and exacerbate academic delays. Lastly, this research
speaks to the value of language immersion programs that may allow certain skills developed in
one language to support literacy acquisition in another. This can have important implications at
the policy level in impacting educational planning as well as in informing decisions regarding
funding for language programs within our school system.
52
References Alexander, P. A. (2003). Profiling the developing reader: The interplay of knowledge, interest,
and strategic processing. In C. M. Fairbanks, J. Worthy, B. Maloch, J. V. Hoffman, & D. L. Schallert (Eds.), The fifty-first yearbook of the National Reading Conference. Oak Creek, WI: National Reading Conference.
Ari, O. (2011). Fluency interventions for developmental readers: Repeated readings and wide reading. Research & Teaching in Developmental Education, 28, 5-15.
August, D., Calderon, M., & Carlo, M. (2002). Transfer of skills from Spanish to English: A study of young learners. Report for practitioners, parents, and policy makers. Washington, DC: Center for Applied Linguistics.
Baker, C. (2011). Foundations of bilingual education and bilingualism (5th ed.). North York, ON: Multilingual Matters.
Barik, H. C., & Swain, M. (1979). Test de Lecture. Toronto, ON: OISE Press.
Barker, T. A., Torgesen, J. K., & Wagner, R. K. (1992). The role of orthographic processing skills on five different reading tasks. Reading Research Quarterly, 27, 335-345.
Bialystok, E. (2001). Metalinguistic aspects of bilingual processing. Annual Review of Applied Linguistics, 21, 169-181.
Bialystok, E., & Hakuta, K. (1994). In other words: The science and psychology of second-language acquisition. New York, NY: Basic Books.
Bialystok, E., Luk, G., Peets, K. F., & Yang, S. (2010). Receptive vocabulary differences in
monolingual and bilingual children. Bilingualism: Language and Cognition, 13, 525-531. Breznitz, Z. (2003). Speed of phonological and orthographic processing as factors in dyslexia:
Electrophysiological evidence. Genetic, Social and General Psychology Monographs, 129, 183-206.
Breznitz, Z. (2006). Fluency in Reading: Synchronization of Processes. Mahwah, NJ: Lawrence
Erlbaum Associates.
Capraro, R. M., & Capraro, M. M. (2001). Commonality analysis: Understanding variance contributions to overall canonical correlation effects of attitude towards mathematics on geometry achievement. Multiple Linear Regression Viewpoints, 27, 16-23.
Cashion, M., & Eagan, R. (1990). Spontaneous reading and writing in English by students in total French immersion: Summary of final report. English Quarterly, 22, 30-44.
Chen, X., Xu, F., Nguyen, T-K., Hong, G., & Wang, Y. (2010). Effects of cross-language transfer on first-language phonological awareness and literacy skills in Chinese children receiving English instruction. Journal of Educational Psychology, 102, 712-728.
53
Cheng, L., Li, M., Kirby, J. R., Quiang, H., & Wade-Woolley, L. (2010). English language immersion and students’ academic achievement in English, Chinese and mathematics. Evaluation & Research in Education, 23, 151-169.
Christ, T. J., Zopluoglu, C., Long, J. D., & Monaghen, B. D. (2012). Curriculum-based measurement of oral reading: Quality of progress monitoring outcomes. Exceptional Children, 78, 356-373.
Comeau, L., Cormier, P., Grandmaison, E., & Lacroix, D. (1999). A longitudinal study of phonological processing skills in children learning to read in a second language. Journal of Educational Psychology, 91, 29-43.
Cossu, G., Shankweiler, D., Liberman, I. Y., Katz, L., & Tola, G. (1988). Awareness of phonological segments and reading ability in Italian children. Applied Psycholinguistics, 9, 1-16.
Cummins, J. (1981). The role of primary language development in promoting educational success for language minority students. In California state Department of Education, Office of Bilingual Bicultural Education (Ed.), Schooling and language minority students: A theoretical framework (pp. 3-49). Los Angeles, CA: California State University.
Cummins, J. (1984). Bilingualism and special education: Issues in assessment and pedagogy. Clevedon, UK: Multilingual Matters.
Cummins, J. (1998). Immersion education for the millennium: What have we learned from 30 years of research on second language immersion? In M. R. Childs & R. M. Bostwick (Eds.), Learning through two languages: Research and practice. Numazu, Japan: Katoh Gakuen.
Cummins, J. (2000). Language, power and pedagogy: Bilingual children in the crossfire. Clevedon, UK: Multilingual Matters.
Cummins, J. (2012). The intersection of cognitive and sociocultural factors in the development of reading comprehension among immigrant students. Reading and Writing, 25, 1973-1990.
Cummins, J., & Swain, M. (1986). Bilingualism in education. London and New York: Longman.
Cunningham, A. E., Perry, K. E., & Stanovich, K. E. (2001). Converging evidence for the concept of orthographic processing. Reading and Writing: An Interdisciplinary Journal, 14, 549-568.
D’Angiulli, A., Siegel, L. S., & Serra, E. (2001). The development of reading in English sand Italian bilingual children. Applied Psycholinguistics, 22, 479-507.
Deacon, S. H., Chen, X., Luo, Y., & Ramirez, G. (2013). Beyond language borders: Orthographic processing and word reading in Spanish-English bilinguals. Journal of Research in Reading, 36, 58-74.
54
Deno, S. L., Fuchs, L. S., Marston, D., & Shin, J. (2001). Using curriculum-based measurement to establish growth standards for students with learning disabilities. School Psychology Review, 30, 507-524.
Dominguez de Ramirez, R., & Shapiro, E. S. (2007). Cross-language relationship between Spanish and English oral reading fluency among Spanish-speaking English language learners in bilingual education classrooms. Psychology in the Schools, 44, 795-806.
Dowhower, S. L. (1987). Effects of repeated reading on second-grade transitional readers’ fluency and comprehension. Reading Research Quarterly, 22, 389-406.
Durgunoglu, A. Y. (2002). Cross-linguistic transfer in literacy development and implications for language learners. Annals of Dyslexia, 52, 189 -204.
Durgunoglu, A. Y., & Oney, B. (1999). A cross-linguistic comparison of phonological awareness and word recognition. Reading and Writing, 11, 281-299.
Durgunoglu, A. Y., Nagy, W. E., & Hancin-Bhatt, B. J. (1993). Cross-language transfer of phonological awareness. Journal of Educational Psychology, 85, 455-465.
Fuchs, L. S., Fuchs, D., & Hosp, M. (2001). Oral reading fluency as an indicator of reading competence: A theoretical, empirical and historical analysis. Scientific Studies of Reading, 5, 239-256.
Fuchs, L. S., Fuchs, D., & Maxwell, L. (1988). The validity of informal measures of reading comprehension. Remedial and Special Education, 9, 20-28.
Fuchs, L. S., Fuchs, D., Hamlett, C. L., Walz, L., & Germann, G. (1993). Formative evaluation of academic progress: How much growth can we expect? School Psychology Review, 22, 27-48.
Gebauer, S. K., Zaunbauer, A. C. M., & Moller, J. (2013). Cross-language transfer in English immersion programs in Germany: Reading comprehension and reading fluency. Contemporary Educational Psychology, 38, 64-74.
Genesee, F. (1978). A longitudinal evaluation of an early immersion program. Canadian Journal of Education, 3, 31-50.
Genesee, F. (1984). French Immersion Programs. In S. M. Shapson & V. D’Oyley (Eds.), Bilingual and multicultural education: Canadian perspectives (Vol. 15). Multilingual matters.
Genesee, F., & Jared, D. (2008). Literacy development in early French immersion programs. Canadian Psychology, 49, 140-147.
Genesee, F., Geva, E., Dressler, C., & Kamil, M. (2006). Cross-linguistic relationships in working memory, phonological processes, and oral language. In D. August & T. Shanahan (Eds.), Developing literacy in second-language learners: A report of the National Literacy Panel on language-minority children and youth (pp. 153-174). Mahwah, NJ: Erlbaum.
55
Geva, E., & Farnia, F. (2012). Developmental changes in the nature of language proficiency and reading fluency paint a more complex view of reading comprehension in ELL and El1. Reading and Writing, 25, 1819-1845.
Geva, E., & Ryan, E. B. (1993). Linguistic and cognitive correlates of academic skills in first and second languages. Language Learning, 43, 5-42.
Geva, E., Wade-Woolley, L., & Shany, M. (1997). Development of reading efficiency in first and second language. Scientific Studies of Reading, 1, 119-144.
Good, R. H., Simmons, D. C., & Kame’enui, E. J. (2001). The importance and decision-making utility of a continuum of fluency-based indicators of foundational reading skills for third-grade high-stakes outcomes. Scientific Studies of Reading, 5, 257-288.
Gottardo, A. (2002). Language and reading skills in bilingual Spanish-English speakers. Topics in Language Disorders, 23, 42-66.
Gottardo, A., Yan, B., Siegel, L. S., & Wade-Woolley, L. (2001). Factors related to English reading performance in children with Chinese as a first language: More evidence of cross-language transfer of phonological processing. Journal of Educational Psychology, 93, 530-542.
Gough, P. B., Hoover, W., & Peterson, C. L. (1996). Some observations on the simple view of reading. In C. Cornoldi & J. Oakhill (Eds.), Reading comprehension difficulties (pp. 1-13). Mahwah, NJ: Lawrence Erlbaum Associates Inc.
Harper, S. N. (2010). Narrowing the gap in early literacy for French immersion students: The effects of a family literacy intervention on grade 1 children’s English and French literacy development. Dissertation Abstracts International Section A: Humanities and Social Sciences, 72, 2261.
Hermanto, N., Moreno, S., & Bialystok, E. (2012). Linguistic and metalinguistic outcomes of intense immersion education: How bilingual? International Journal of Bilingual Education and Bilingualism, 15, 131-145.
Jared, D., Cormier, P., Levy, B. A., & Wade-Woolley, L. (2011). Early predictors of biliteracy development in children in French immersion: A 4-year longitudinal study. Journal of Educational Psychology, 103, 119-139.
Jenkins, J. R., Fuchs, L. S., Espin, C., van den Broek, P., & Deno, S. L. (2000). Effects of task format and performance dimension on word reading measures: Criterion validity, sensitivity to impairment, and context facilitation. In Pacific Coast Research Conference, San Diego, CA.
Jenkins, J. R., Fuchs, L. S., van den Broek, P., Espin, C., & Deno, S. L. (2003). Sources of individual differences in reading comprehension and reading fluency. Journal of Educational Psychology, 95, 719-729.
56
Kintsch, W., & Kintsch, E. (2005). Comprehension. In S. G. Paris & S. A. Stahl (Eds.), Children’s reading comprehension and assessment (pp.71-92). Mahwah, NJ: Erlbaum.
Klauda, S. L., & Guthrie, J. T. (2008). Relationship of three components of reading fluency to reading comprehension. Journal of Educational Psychology, 100, 310-321.
Koda, K. (2008). Impacts of prior literacy experience on second-language learning to read. In K. Koda & A. M. Zehler (Eds.), Learning to read across languages: Cross-linguistic relationships in first- and second-language literacy development (pp. 68-96). Mahwah, NJ: Routledge.
Kuhn, M. R., & Stahl, S. A. (2003). Fluency: A review of developmental and remedial practices. Journal of Educational Psychology, 95, 3-21.
Kuhn, M. R., Schwanenflugel, P. J., & Meisinger, E. B. (2010). Aligning theory and assessment of reading fluency: Automaticity, prosody and definitions of fluency. Reading Research Quarterly, 45, 230-251.
LaBerge, D., & Samuels, S. J. (1974). Toward a theory of automatic information processing in reading. Cognitive Psychology, 62, 293-323.
Lesaux, N. K., Crosson, A. C., Kieffer, M. J., & Pierce, M. (2010). Uneven profiles: Language minority learners’ word reading, vocabulary, and reading comprehension skills. Journal of Applied Developmental Psychology, 31, 475-483
Levy, B. A., Abello, B., & Lysynchuk, L. (1997). Transfer from word training to reading in context: Gains in reading fluency and comprehension. Learning Disability Quarterly, 20, 173-188.
Lundberg, I., Olofsson, A., & Wall, S. (1980). Reading and spelling skills in the first school years predicted from phonemic awareness skills in kindergarten. Scandinavian Journal of Psychology, 1, 159-173.
Marston, D. (1989). A curriculum-based measurement approach to assessing academic performance: What is it and why do it. In M. R. Shinn (Ed.), Curriculum-based measurement: Assessing special children (pp. 18-78). New York: Guilford.
McBride-Chang, C., Tong, X., Shu, H., Wong, A. M-Y., Leung, K-W., & Tardif, T. (2008). Syllable, phoneme and tone: Psycholinguistic units in early Chinese and English word recognition. Scientific Studies of Reading, 12, 171-194.
Meyer, M. S., & Felton, R. H. (1999). Repeated reading to enhance fluency: Old approaches and new directions. Annals of Dyslexia, 49, 283-306.
Naglieri, J. A. (1985). Matrix Analogies Test – Expanded Form. San Antonio, TX: The Psychological Corporation.
57
National Reading Panel. (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Washington, DC: National Institute of Child Health and Human Development.
Office of the Commissioner of Official Languages (2011). Provincial Enrollment Statistics 2010-2011. Ottawa: Minister of Public Works and Government Services Canada.
Paez, M. M., Tabors, P. O., & Lopez, L. M. (2007). Dual language and literacy development of Spanish-speaking preschool children. Journal of Applied Developmental Psychology, 28, 85-102.
Perfetti, C. A. (1985). Reading Ability. New York, NY: Oxford University Press.
Perfetti, C. A. (1995). Cognitive research can inform reading education. Journal of Research in Reading, 18, 106-115.
Posner, M. I., & Snyder, C. R. R. (1975a). Attention and cognitive control. In R. Solso (Ed.), Information processing and cognition: The Loyola Symposium (pp. 55-85). Hillsdale, NJ: Lawrence Erlbaum Associates Inc.
Posner, M. I., & Snyder, C. R. R. (1975b). Facilitation and inhibition in the processing of signals. In P. M. A. Rabbitt & S. Dornic (Eds.), Attention and performance (pp. 669-682). New York: Academic.
Ramirez, M. C. (2001). An investigation of English language learners and reading skills on reading comprehension for Spanish-speaking English language learners. Dissertation Abstracts International, 62, 2716.
Riches, C., & Genesee, F. (2006). Crosslanguage and crossmodal influences. In F. Genesee, K. Lindholm-Leary, W. Saunders, & D. Christian (Eds.), Educating English language learners: A synthesis of research evidence. New York: Cambridge University Press.
Rolstad, K., Mahoney, K., & Glass, G. V. (2005). The big picture: A meta-analysis of program effectiveness research on English language learners. Educational Policy, 19, 572-594.
Rumelhart, D. E. (1994). Toward an interactive model of reading. In R. B. Ruddell, M. R. Ruddell, & H. Singer (Eds.), Theoretical models and processes of reading (4th ed.). Newark: DE: International Reading Association.
Samuels, S. J. (1979). The method of repeated readings. Reading Teacher, 32, 403-408.
Schreiber, P. A. (1980). On the acquisition of reading fluency. Journal of Reading Behaviour, 12, 177-186.
Stanovich, K. E. (2000). Progress in understanding reading: Scientific foundations and new frontiers. New York: Guilford.
Swain, M., & Lapkin, S. (1982). Evaluating bilingual education: A Canadian case study. Clevedon, UK: Multilingual Matters.
58
Tarone, E., & Swain, M. (2011). A sociolinguistic perspective on second language use in immersion classrooms. The Modern Language Journal, 79, 166-178.
Therrien, W. J. (2004). Fluency and comprehension gains as a result of repeated reading: A meta-analysis. Remedial and Special Education, 25, 252-261.
Torgesen, J. K., Wagner, R. K., Rashotte, C. A., Rose, E., Lindamood, P., Conway, T., & Garvan, C. (1999). Preventing reading failure in children with phonological processing difficulties: Group and individual responses to instruction. Journal of Educational Psychology, 81, 579-593.
Turnbull, M., Hart, D., & Lapkin, S. (2001). Grade 3 immersion students’ performance in reading, writing, and mathematics: Province-wide results from Ontario (1998/99). Canadian Modern Language Review, 58, 9-26.
Turnbull, M., Hart, D., & Lapkin, S. (2003). Grade 6 French immersion students’ performance on large-scale reading, writing, and mathematics tests: Building explanations. The Alberta Journal of Educational Research, 1, 6-23.
Verhoeven, L. (1991). Acquisition of biliteracy. AILA Review, 8, 61-74.
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1999a). Comprehensive test of phonological processing. Austin, TX: PRO-ED.
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1999b). Test of word reading efficiency. Austin, TX: PRO-ED.
Wang, M., Park, Y., & Lee, K. R. (2006). Korean-English biliteracy acquisition: Cross-language phonological and orthographic transfer. Journal of Educational psychology, 98, 148-158.
Wang, M., Perfetti, C. A., & Liu, Y. (2005). Chinese-English biliteracy acquisition: Cross-language and writing system transfer. Cognition, 97, 67-88.
Wayman, M. M., Wallace, T., Wiley, H. I., Ticha, R., & Espin, C. A. (2007). Literature synthesis on curriculum-based measurement in reading. The Journal of Special Education, 41, 85-120.
Wechsler, D. (2009). Wechsler Individual Achievement Test – Third Edition. London: The Psychological Corp.
Wolf, M., & Katzir-Cohen, T. (2001). Reading fluency and its intervention. Scientific Studies of Reading, 5, 211-239.
59
Appendix A Table 1. Descriptive Statistics for All Measures at T1 and T2 T1 (Grade 2) T2 (Grade 3) Mean SD Mean SD Nonverbal reasoning 20.06 11.99 - - - - English phonological awareness 14.20 4.75 16.27 4.16 English rapid naming ability 38.55 10.61 32.98 7.34 English word reading fluency 59.23 17.19 66.67 11.46 English text reading fluency - - - - 120.48 42.83 French phonological awareness 14.29 5.03 18.18 6.81 French rapid naming ability 46.43 11.53 36.70 7.52 French word reading fluency 52.29 13.36 62.47 11.03 French text reading fluency - - - - 95.54 29.89 Note. All values are presented as raw scores (for description of what raw scores represent, see Measures section in Study 1, Study 2 and Study 3).
60
Appendix B Table 2. Pearson Correlations between All Variables in Study 1 ______________________________________________________________________________ NVR T1 T1 T1 T1 T1 T1 T2 T2
En PA Fr PA En RAN Fr RAN En WF Fr WF En WF Fr WF NVR 1 .43** .36** -.17 -.12 .34** .37** .30* .24* T1 En PA 1 .79** -.33** -.33** .35** .51** .44** .41** T1 Fr PA 1 -.18 -.21 .22 .38** .37** .37** T1 En RAN 1 .78** -.44** -.42** -.50** -.44** T1 Fr RAN 1 -.42** -.43** -.47** -.45** T1 En WF 1 .76** .79** .72** T1 Fr WF 1 .77** .76** T2 En WF 1 .84** T2 Fr WF 1 Notes. * p < 0.05; ** p < 0.01 T1 – Spring of Grade 2; T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency
61
Appendix C Table 3. Hierarchical Regression Predicting T2 English Word Fluency Step Variable Adjusted R2 ∆R2 ß 1 NVR 0.08* 0.09* -0.03* 2 Language Status 0.07 0.01 0.01 3 T1 En PA 0.18** 0.12** 0.08** 4 T1 En RAN 0.30** 0.13** -0.14** 5 T1 En WF 0.65** 0.33** 0.47** 6 T1 Fr WF 0.69** 0.04** 0.33** Notes. * p < 0.05; ** p < 0.01 T1 – Spring of Grade 2; T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency Table 4. Hierarchical Regression Predicting T2 French Word Fluency Step Variable Adjusted R2 ∆R2 ß 1 NVR 0.04* 0.06* -0.10* 2 Language Status 0.03 0.00 0.11 3 T1 Fr PA 0.11* 0.09* 0.14* 4 T1 Fr RAN 0.25** 0.14** -0.10** 5 T1 Fr WF 0.57** 0.31** 0.38** 6 T1 En WF 0.63** 0.06** 0.41** Notes. * p < 0.05; ** p < 0.01 T1 – Spring of Grade 2; T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency
62
Appendix D Table 5. Commonality Analysis Predicting T2 English Word Fluency Construct Coefficient % Total Common to: NVR & T1 English PA 0.00 0.10% NVR & T1 English RAN 0.00 0.07% NVR & T1 English WF 0.00 0.09% NVR & T1 French WF 0.00 0.06% T1 English PA & T1 English RAN 0.00 0.46% T1 English PA & T1 English WF 0.00 0.47% T1 English PA & T1 French WF 0.02 2.16% T1 English RAN & T1 English WF 0.02 2.99% T1 English RAN & T1 French WF 0.00 0.46% T1 English WF & T1 French WF 0.25 34.67% Table 6: Commonality Analysis Predicting T2 French Word Fluency Construct Coefficient % Total Common to: NVR & T1 French PA 0.01 0.76% NVR & T1 French RAN 0.01 0.23% NVR & T1 French WF 0.00 0.52% NVR & T1 English WF 0.01 0.76% T1 French PA & T1 French RAN 0.00 0.36% T1 French PA & T1 French WF 0.02 4.15% T1 French PA & T1 English WF 0.01 1.32% T1 French RAN & T1 French WF 0.01 1.40% T1 French RAN & T1 English WF 0.01 1.35% T1 French WF & T1 English WF 0.25 37.91%
63
Appendix E
Table 7. Pearson Correlations between All Variables in Study 2 ______________________________________________________________________________ NVR T1 T1 T1 T1 T1 T1 T2 T2
En PA Fr PA En RAN Fr RAN En WF Fr WF En TF Fr TF NVR 1 .43** .36** -.17 -.12 .34** .37** .41** .23 T1 En PA 1 .79** -.33** -.33** .35** .51** .42** .37** T1 Fr PA 1 -.18 -.21 .22 .38** .35** .30* T1 En RAN 1 .78** -.44** -.42** -.44** -.41** T1 Fr RAN 1 -.42** -.43** -.38** -.43** T1 En WF 1 .76** .73** .62** T1 Fr WF 1 .80** .75** T2 En TF 1 .83** T2 Fr TF 1 Notes. * p < 0.05; ** p < 0.01 T1 – Spring of Grade 2; T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency TF – Text-Level Fluency
64
Appendix F
Table 8. Hierarchical Regression Predicting T2 English Text Fluency Step Variable Adjusted R2 ∆R2 ß 1 NVR 0.15** 0.17** 0.12** 2 Language Status 0.14 0.00 0.04 3 T1 En PA 0.21* 0.08* -0.02* 4 T1 En RAN 0.30** 0.10** -0.11** 5 T1 En WF 0.56** 0.25** 0.25** 6 T1 Fr WF 0.66** 0.10** 0.53** Notes. * p < 0.05; ** p < 0.01 T1 – Spring of Grad 2, T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency TF – Text-Level Fluency Table 9. Hierarchical Regression Predicting T2 French Text Fluency Step Variable Adjusted R2 ∆R2 ß 1 NVR 0.04‡ 0.05‡ -0.06‡ 2 Language Status 0.03 0.01 0.11 3 T1 Fr PA 0.07‡ 0.05‡ 0.02‡ 4 T1 Fr RAN 0.20** 0.14** -0.11** 5 T1 Fr WF 0.56** 0.34** 0.62** 6 T1 En WF 0.55 0.01 0.13 Notes. ‡ p = 0.06, * p < 0.05; ** p < 0.01 T1 – Spring of Grade 2, T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency TF – Text-Level Fluency
65
Appendix G
Table 10. Commonality Analysis Predicting T2 English Text Fluency Construct Coefficient % Total Common to: NVR & T1 English PA 0.00 0.03% NVR & T1 English RAN 0.00 0.15% NVR & T1 English WF 0.00 0.73% NVR & T1 French WF 0.00 0.45% T1 English PA & T1 English RAN 0.00 0.03% T1 English PA & T1 English WF 0.00 0.16% T1 English PA & T1 French WF 0.02 1.56% T1 English RAN & T1 English WF 0.02 1.09% T1 English RAN & T1 French WF 0.00 0.59% T1 English WF & T1 French WF 0.25 32.73% Table 11. Commonality Analysis Predicting T2 French Text Fluency Construct Coefficient % Total Common to: NVR & T1 French PA 0.00 0.11% NVR & T1 French RAN 0.00 0.18% NVR & T1 French WF 0.00 0.43% NVR & T1 English WF 0.00 0.14% T1 French PA & T1 French RAN 0.00 0.11% T1 French PA & T1 French WF 0.02 3.48% T1 French PA & T1 English WF 0.00 0.08% T1 French RAN & T1 French WF 0.02 3.01% T1 French RAN & T1 English WF 0.00 0.43% T1 French WF & T1 English WF 0.19 32.69%
66
Appendix H
Table 12. Pearson Correlations between All Variables in Study 3 ______________________________________________________________________________ NVR T2 T2 T2 T2 T2 T2 T2 T2
En PA Fr PA En RAN Fr RAN En WF Fr WF En TF Fr TF NVR 1 .28* .38** -.07 -.05 .30* .24* .41** .23 T2 En PA 1 .78** -.45** -.48** .53** .49** .48** .40** T2 Fr PA 1 -.36** -.45** .53** .52** .55** .52** T2 En RAN 1 .81** -.51** -.50** -.41** -.44** T2 Fr RAN 1 -.60** -.61** -.49** -.53** T2 En WF 1 .84** .86** .67** T2 Fr WF 1 .79** .79** T2 En TF 1 .83** T2 Fr TF 1 Notes. * p < 0.05; ** p < 0.01 T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency TF – Text-Level Fluency
67
Appendix I
Table 13. Hierarchical Regression Predicting T2 English Text Fluency Step Variable Adjusted R2 ∆R2 ß 1 NVR 0.15** 0.17** 0.14** 2 Language Status 0.14 0.00 -0.01 3 T2 En PA 0.28** 0.15** 0.01** 4 T2 En RAN 0.33* 0.05* 0.09* 5 T2 En WF 0.75** 0.41** 0.55** 6 T2 Fr TF 0.87** 0.11** 0.47** Notes. * p < 0.05; ** p < 0.01 T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency TF – Text-Level Fluency Table 14. Hierarchical Regression Predicting T2 French Text Fluency Step Variable Adjusted R2 ∆R2 ß 1 NVR 0.04 0.05 -0.11 2 Language Status 0.03 0.01 0.08 3 T2 Fr PA 0.25** 0.22** 0.07** 4 T2 Fr RAN 0.35** 0.11** -0.03** 5 T2 Fr WF 0.62** 0.26** 0.29** 6 T2 En TF 0.73** 0.11** 0.59** Notes. * p < 0.05; ** p < 0.01 T2 – Spring of Grade 3 En – English; Fr – French NVR – Nonverbal Reasoning
PA – Phonological Awareness RAN – Rapid Automatized Naming WF – Word-Level Fluency TF – Text-Level Fluency
68
Appendix J
Table 15. Commonality Analysis Predicting T2 English Text Fluency Construct Coefficient % Total Common to: NVR & T2 English PA 0.00 0.10% NVR & T2 English RAN 0.00 0.39% NVR & T2 English WF 0.02 2.23% NVR & T2 French TF 0.01 0.56% T2 English PA & T2 English RAN 0.00 0.01% T2 English PA & T2 English WF 0.01 1.41% T2 English PA & T2 French TF 0.00 0.01% T2 English RAN & T2 English WF 0.01 0.59% T2 English RAN & T1 French TF 0.00 0.55% T2 English WF & T1 French TF 0.27 30.78% Table 16. Commonality Analysis Predicting T2 French Text Fluency Construct Coefficient % Total Common to: NVR & T2 French PA 0.00 0.25% NVR & T2 French RAN 0.00 0.17% NVR & T2 French WF 0.00 0.43% NVR & T2 English TF 0.01 1.03% T2 French PA & T2 French RAN 0.00 0.12% T2 French PA & T2 French WF 0.00 0.29% T2 French PA & T2 English TF 0.01 1.10% T2 French RAN & T2 French WF 0.01 0.99% T2 French RAN & T2 English TF 0.00 0.10% T2 French WF & T2 English TF 0.23 30.81%