Content uploaded by P. David Pearson
Author content
All content in this area was uploaded by P. David Pearson on Aug 04, 2020
Content may be subject to copyright.
PLEASE SCROLL DOWN FOR ARTICLE
7KLVDUWLFOHZDVGRZQORDGHGE\
>6DLORUV0LVW\@
2Q
$SULO
$FFHVVGHWDLOV
$FFHVV'HWDLOV>VXEVFULSWLRQQXPEHU@
3XEOLVKHU
5RXWOHGJH
,QIRUPD/WG5HJLVWHUHGLQ(QJODQGDQG:DOHV5HJLVWHUHG1XPEHU5HJLVWHUHGRIILFH0RUWLPHU+RXVH
0RUWLPHU6WUHHW/RQGRQ:7-+8.
%LOLQJXDO5HVHDUFK-RXUQDO
3XEOLFDWLRQGHWDLOVLQFOXGLQJLQVWUXFWLRQVIRUDXWKRUVDQGVXEVFULSWLRQLQIRUPDWLRQ
KWWSZZZLQIRUPDZRUOGFRPVPSSWLWOHaFRQWHQW W
7KH(IIHFWVRI)LUVWDQG6HFRQG/DQJXDJH,QVWUXFWLRQLQ5XUDO6RXWK
$IULFDQ6FKRROV
0LVW\6DLORUVD-DPHV9+RIIPDQD3'DYLG3HDUVRQE61DWDVKD%HUHWYDVD%HUWXV0DWWKHHF
D7KH8QLYHUVLW\RI7H[DVE8QLYHUVLW\RI&DOLIRUQLDF5($'(GXFDWLRQDO7UXVW
2QOLQHSXEOLFDWLRQGDWH$SULO
7RFLWHWKLV$UWLFOH6DLORUV0LVW\+RIIPDQ-DPHV93HDUVRQ3'DYLG%HUHWYDV61DWDVKDDQG0DWWKHH%HUWXV
7KH(IIHFWVRI)LUVWDQG6HFRQG/DQJXDJH,QVWUXFWLRQLQ5XUDO6RXWK$IULFDQ6FKRROV%LOLQJXDO5HVHDUFK-RXUQDO
٢
7ROLQNWRWKLV$UWLFOH'2,
85/KWWSG[GRLRUJ
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
Bilingual Research Journal, 33: 21–41, 2010
Copyright © the National Association for Bilingual Education
ISSN: 1523-5882 print / 1523-5890 online
DOI: 10.1080/15235881003733241
UBRJ1523-58821523-5890Bilingual Research Journal, Vol. 33, No. 1, Mar 2010: pp. 0–0Bilingual Research Journal
The Effects of First- and Second-Language Instruction
in Rural South African Schools
The Effects of First- and Second-Language InstructionSailors et al. Misty Sailors
The University of Texas
James V. Hoffman
The University of Texas
P. David Pearson
University of California
S. Natasha Beretvas
The University of Texas
Bertus Matthee
READ Educational Trust
In this article, we report on the results of a project devoted to improving literacy in South Africa’s
rural schools; specifically we report the results of an intervention study that centered on improving
Misty Sailors is Associate Professor of Literacy Education in the department of Interdisciplinary Learning and
Teaching at The University of Texas at San Antonio (UTSA). Her research interests are in the role of print rich environ-
ments and comprehension instruction. She teaches undergraduate and graduate literacy education courses. Sailors has
worked in South Africa, Tanzania, and Malawi with classroom teachers for six years and is the Program Director of the
Textbooks and Learning Materials Program at the UTSA (also known as the Malawi Reads! program).
James V. Hoffman is Professor of Language and Literacy Studies at The University of Texas at Austin. His research
interests are in the areas of reading teacher preparation and in the qualities of texts that support reading development.
Hoffman has worked across Africa for fifteen years.
P. David Pearson is the Dean of the Graduate School of Education at University of California, Berkeley. His research
interests include practice and policy in literacy instruction and assessment. He has served as president of the National
Reading Conference and on the boards of directors for the International Reading Association, the National Reading
Conference, and the Association of American Colleges of Teacher Education.
S. Natasha Beretvas is Associate Professor and Chair of the Quantitative Methods program in Educational Psychology
at The University of Texas at Austin. Her research interests lie in the application and evaluation of statistical and psycho-
metric models with a focus on methodological dilemmas in meta-analysis and multilevel modeling.
Bertus Matthee is an Educational Psychologist and has experience in private practice focusing mainly on the cogni-
tive development of young children. For the past 11 years he has worked for the READ Educational Trust, evaluating
the levels of in-service training of educators and the resource provisioning in classrooms at early childhood, primary,
high school and school management levels.
Address correspondence to Misty Sailors, University of Texas at San Antonio, One UTSA Circle, San Antonio, TX
78249. E-mail: Misty.Sailors@utsa.edu
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
22 SAILORS ET AL.
mother-tongue literacy instruction offered to learners in Grades 1 and 2 in South African schools.
Our findings demonstrate that there are positive and strong effects for the home-language initiative
that was tested in this study. Our data suggest that there is a high level of value added to perfor-
mance in both the home-language and English-language learning of the students in this study when a
much more print-rich environment was provided in the home language. These findings challenge the
deficit myth often associated with children who come to school with a first language different from
the medium of instruction. Further, these data suggest that providing instruction in both first and sec-
ond languages can have a positive impact on development in both languages.
INTRODUCTION
The vastness of the African continent is exceeded only by its diversity—geographic, ethnic, cul-
tural, and linguistic. The continent supports over 875 million people with a documented 2,011
languages (UNESCO, 2004). Despite an overt attempt to protect the languages of the indigenous
people by organizations such as United Nations Educational, Scientific and Cultural Organiza-
tion (UNESCO) and some governments, the languages of the former colonial powers continue
to dominate the political and educational systems across the continent.
Koïchiro Matsuura, the Director-General of UNESCO, argued that in a world where English
guides commercialism, protecting the many endangered languages around the world is a way of
safeguarding the world’s linguistic diversity and thus, ensuring the protection of the cultural
diversity of humanity (Matsuura, 2003). One of the ways of protecting these languages, argues
Matsuura, is to promote the practice and use of mother tongues in schools. Defined by UNESCO,
mother tongues are those languages, “learnt by African children through social interactions with
members of communities to which they are linked by parentage” (UNESCO, 1997, p. 15).
One Non-Governmental Organization (NGO) in South Africa, the READ (Read, Educate,
and Develop) Educational Trust, has taken up the call to protect and promote mother-tongue lit-
eracy in that country. READ is deeply committed to raising the literacy levels of South African
learners, especially those who have historically been underserved by public education. We have
served as volunteer evaluation consultants to READ, and in the process, gathered a great deal of
evidence to evaluate the impact of their work. In this article, we report on the results of one
aspect of a much broader project devoted to improving literacy in South Africa’s rural Black
schools; specifically we report the results of an intervention study that centered on improving
mother-tongue literacy instruction offered to learners in Grades 1 and 2 in South African schools.
This research report is prefaced by an account of the educational policies of South Africa and is
followed by our thoughts on the implications of this study for other developing countries in Africa.
EDUCATION IN SOUTH AFRICA
It goes without saying that the country of South Africa has undergone incredible changes. With
the democratic elections of 1994 and the passing of the new constitution in 1996, the political
focus on language and language in education has accelerated, especially for foundation-phase
students (students in Grades 1 through 3) in South Africa. As a result, literacy learning is begin-
ning to take on a more democratic look, differing greatly from past literacy learning, especially for
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 23
African-language-speaking learners. Section 29 of the constitution of 1996 states that everyone has
the right to a basic education, and everyone has the right to receive education in the official language
of their choice in public educational institutions where that education is reasonably practicable.
The move to provide all learners in South Africa an education in the official language of their
choice remains a formidable task, at least in part due to the constitution’s recognition of 11 official
languages: Sepedi, Sesotho, Setswana, siSwati, Tshivenda, Xitsonga, Afrikaans, English, isiNdebele,
isiXhosa, and isiZulu.
The language-in-education policy of South Africa is based on the notion of additive multi-
lingualism, and promotes the use of the home language alongside an additional language. This
policy encourages, according to Desai (2001), a “multilanguage” society with the goal of elevating
the status and advancing the use of indigenous languages. In order to do this, the language pol-
icy of the country calls for “parity of esteem and the equitable use of the official languages”
(Department of Education of South Africa, 2002). Under the language policy of the country,
learning in the Reception Year (Grade R) and Grade 1 is to occur in the learners’ home language.
When the additional language is to be introduced as a subject, the home language continues to
be used alongside the additional language for “as long as possible” (National Curriculum Statement,
p. 5). The National Curriculum gives “School Governing Bodies the responsibility of selecting
school language policies that are appropriate for their circumstances and in line with the policy
of additive multilingualism” (p. 4). The current language-in-education policy is a move, albeit a
gradual one, toward a multilingual society (Brock-Utne & Holmarsdottir, 2004).
Research on Home Language as the Medium of Instruction for Primary
Learners in Africa
The nature of the language-in-education policy in South Africa is based on years of research
around the effectiveness of mother-tongue instruction for literacy acquisition. Heugh (2006)
provides an historical account of the empirical South African research supporting mother-tongue
instruction beginning with a large survey of 18,733 students by E. G. Malherbe in 1938 on the
relationship between education and bilingualism in South Africa. Other research followed,
although none was quite as significant as Macdonald’s work in 1990 (in Heugh, 2006). A central
tenet of Macdonald’s work was that students who switched medium of instruction before they
had sufficiently developed and learned the new target language were not successful. Similarly
stated, Akinnaso (1993) found that effective literacy acquisition and second-language profi-
ciency depended on well-developed first-language proficiency. Heugh (1995) developed this work
and wrote of the importance of building cognitive competence in students prior to introducing a new
language of learning and teaching. Heugh’s (2006) work further suggests that students should
remain engaged in mother-tongue instruction through the first 6 years of their formal education.
Other researchers continue to monitor the role of mother tongue in language and learning.
For example, Broom (2004) most recently assessed reading achievement at the end of Grade 3,
comparing L1, the language of learning and teaching in classrooms, and the history of the school
(“disadvantaged” versus “previously advantaged” schools). In addition to physical differences
in schools, Broom found differences in material resources between the schools in her study. She
attributed this, in part, to the differences in measures of reading achievement between the
schools. Her work suggests that early transition to L2 (usually English in South Africa) may
perpetuate the inequities of South Africa’s past. Their work suggests that language instruction in
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
24 SAILORS ET AL.
primary schools is critical if students are to be successful (Brock-Utne, 2007a, 2007b; Brock-Utne
& Holmarsdottir, 2004).
Other recent work across Africa corroborates this research. For example, in the Six-Year Pri-
mary (Yoruba–English Bilingual) Project in Nigeria, the home language of Yoruba children was
used as the medium of instruction throughout the primary years. Afolayan (1999) showed that
the mother tongue of that region is a more effective medium of instruction for primary learners
than English and that literacy is more easily acquired through Yoruba than through English. In
Zambia, researchers (Tambulukani, Sampa, Musuku, & Linehan, 1999) examined the use of
Icibemba as the language of instruction with primary students; one of the components of this
project capitalized on the Language Experience Approach of Staufer (1967). Learners involved
in this program demonstrated substantive progress in their literacy skills when compared to
learners who received more traditional instruction.
Similarly, special-needs Nigerian students and those who were nonreaders at the end of their pri-
mary school experience were also introduced to the LEA (Umolu, 1999). Umolu reports that learn-
ers who experience reading under an LEA perspective grow in their word recognition skills and
their self-esteem. Finally, Donald, Condy, and Forrester (2003) introduced 3,054 Grade 1 to 3 learn-
ers in South Africa to the Concentrated Language Encounter (CLE), an approach with its roots in
the LEA tradition. This 3-year intervention study focused on texts that were written in mother
tongue and used with students. Students who participated in this study significantly outperformed
nonparticipants on a dictation test and appeared to benefit in terms of their written-language skills.
While it would seem that research has demonstrated the importance of mother-tongue
instruction, the reaction of the community to this pedagogy is mixed. Some studies report that
parents support it (Heugh, 2000; Hunt, 2007) while others demonstrated that parents want their
students in English classrooms (Broom, 2004; De Klerk, 2002a) because more resources exist in
classrooms that support English learning (Maile, 2004). Because of this, there has been a
“surge” toward English as a medium of learning and teaching (Moodley, 2007, p. 708). Teachers
are also mixed in their perceptions about language learning—some support English (Setati,
2008) while others are willing to teach in mother tongue with proper support and training
(Du Plessis & Louw, 2008; De Klerk, 2002b) as some teachers do not feel prepared to teach
languages and language learning (Mentz & Van Der Walt, 2007) and do not feel they have
adequate materials and resources (O’Connor & Geiger, 2009).
In summary, to provide learners in South Africa with instruction in their home language
while they are learning a second language is to value their home experiences and to provide
them with tools that help them gain access to better health, social equity, and international ideas.
Considering the large number of learners, limited learning resources, and a teaching force that
requires an incredible investment in professional development to upgrade their skills, this is a
formidable task (Young, 1995). One nongovernmental organization of South Africa, the READ
Educational Trust, has taken on this challenge and has attempted to provide learners in rural
schools opportunities to learn to read both in their home language and in English.
LEARNING TO READ WITH READ: THE HOME-LANGUAGE INITIATIVE
Since its inception in 1979 in response to the Soweto uprisings, READ has committed itself to
transforming South Africa into a country of literate people through educational interventions.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 25
The most recent project, and perhaps the largest of its kind, is the Business Trust’s Learning for
Living project. This project was supported by the donation of R153 million (approximately 15
million dollars) by the Business Trust, a private foundation in South Africa. With implementa-
tion beginning in 2000 by READ, under the auspices of the Department of Education, this 5-year
project involved the careful resourcing of 957 South African classrooms and the careful mentoring
of 7,500 Foundation Phase teachers (Grades 1–3) and 7,500 Intermediate Phase (Grades 4–7)
teachers. At its completion, this project reached almost 1 million South African learners across all
nine provinces of South Africa. The schools were largely situated in rural and in former township
districts. These were the country’s most disadvantaged schools, with the sparsest of resources
and educational materials. In an effort to approximate a quasi experiment, during the first years
of the Learning for Living study, officials selected control schools that closely mirrored the
demographics of the intervention schools.
During the second year of this project a third purposive sampling of schools was identified.
From within the larger intervention sample, the READ staff nominated schools that had achieved
high levels of program implementation. In other words, these schools were sampled to represent
those schools making commendable progress toward implementing the goals and features of the
Learning for Living project. For the purposes of this report these schools will be referred to as
Select schools.
For the first 3 years of the project learners in Grades 3–7 were assessed. This aspect of the
project and results are described elsewhere (Hoffman, Beretvas, & Pearson, 2004; Sailors, Hoffman,
Pearson, Beretvas, & Matthee, 2007; Sailors, Hoffman, & Matthee, 2007). During the fourth year
of the project we developed an assessment to measure the literacy learning of learners in Grades
1 and 2 in each of the 11 official South African languages. The Home Language Initiative, on
which this article is based, is but one piece of this larger Learning for Living project and focuses
attention on the literacy learning of students at the participating schools in Grades 1 and 2.
In this study we wanted to answer three questions:
1. Compared to control schools, do the professional development and classroom resources
offered in the project schools improve the reading achievement of foundation-phase
students?
2. How does the performance of students in the larger sample of intervention schools com-
pare to the performance of students in the Select schools?
3. Among foundation-phase students, what is the relationship between first/home-language
development and success in learning to read English?
METHOD
The Context: The READ Intervention
READ’s philosophy is rooted in the outcome-based teaching methodologies identified in the
South African National Curriculum. To that end, the work of READ centers on two overarch-
ing ideas: (a) placing high-quality books in the hands of children, and (b) providing teachers
with strategies for engaging students in rich encounters with those books (see, for example,
Elley, 2000).
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
26 SAILORS ET AL.
Because the Learning for Living project focuses on learners in rural areas of South Africa,
READ selects high-quality children books that allow African language learners to see themselves
reflected in the books they read and that are read to them. Participating classrooms received
books for a variety of reasons from a variety of sources.
All participating intervention teachers received books to be used as Read Alouds. Among
the set were 10 high-quality picture books that represented the lives and experiences of the
children of South Africa. The publication of children’s literature is not a profitable business in
South Africa even in English, and it is even riskier for books in the various mother tongues.
Hence the READ staff, in conjunction with outside experts, decided to commission 10 cultur-
ally relevant Grade 1 narratives to be written in English and translated to the home languages
of the participating schools. Grade 2 classrooms received the English version of the10 books.
Finally, all classrooms received 8 wordless, oversized picture books to be used for instruction
in either language.
Teachers also received Learner Materials in the form of 10 developmentally appropriate
mother-tongue leveled readers to be placed in the hands of the children in the classroom. They
were developed/translated, with publisher agreement, from the Australian Sunshine books. READ
purchased the translation rights, and linguistic consultants translated the stories into each of the
official South African languages.
Each of the teachers in Grade 1 received 20 Sunshine Starter home-language books
(based on language of school), 20 Sunshine Big Books (written in English), and 120 “little
books” (6 of each of the Big Book titles). Grade 2 teachers received 40 Sunshine Starter
English books, 40 Big Books (English), and 240 “little books” (6 of each of the Big Book
titles, all in English). In addition, all teachers received a Teacher’s Guide to support their
use of the various books. Additionally, each student received Learner Books, workbooks
that provided children with opportunities to practice the reading skills taught in each of the
leveled readers.
In addition to the materials, the intervention was based on intensive and systematic profes-
sional development for teachers on how to use the materials to maximize student learning. The
teachers’ curriculum, framed in a 10-day teaching cycle, was written by the READ staff to
closely align with the national curriculum. Monthly coaching visits by a member of the READ
training staff ensured maximum support for teachers in the schools. Professional development
included demonstration lessons by the READ mentors, after-school workshops, and one-on-one
teacher reflections in response to the monitoring visits by READ mentors. Because this inter-
vention was school-wide, the READ staff held conversations with the school administrator.
Principal workshops were offered as a means of helping them support their teachers during this
project. Finally, there was close contact and cooperation with the Ministry of Education during
this intervention.
The teachers in this project were understandably enthusiastic about the materials and the
professional development that accompanied them (Sailors, Makalela, & Hoffman, in press).
According to the staff at READ, nothing like the Sunshine series had ever been available any-
where in South Africa; to the contrary, most early home-language courses were textbooks,
entirely phonics based and often culturally irrelevant if not meaningless. For example, chil-
dren might practice the m sound by chanting ma-me-meh— utterances that are not actual
words in any of the home languages (personal communication, Debbie Botha, December 3,
2003).
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 27
The Materials: The Home-Language/English-as-a–Second-Language Assessment
The Language Assessment constructed for this study has two components, home language and
English. For the home-language component, we developed 10 versions, one for each of the
official South African languages represented in this study. We designed this assessment with
several goals in mind. First, it would provide scores for a broad range of first- and second-
language abilities. Second, it could be administered efficiently (in less than 20 minutes) to indi-
vidual students. Third, it would be roughly equivalent in content and difficulty across the 10
home languages and in English. And finally, and perhaps most importantly, it would be sensitive
to the South African cultural context.
The first drafts of the assessment were written in Afrikaans and English, piloted by the
READ staff. The next drafts were written in Xhosa, Sotho, and Zulu and then piloted. We
revised the tests in these five languages in response to pilot data. Then we used the same process
to create versions for the remaining languages. During the administration of the assessment, the
learners were first assessed in their home language, then in English.
Each of the two assessments (home language and English) consisted of seven compo-
nents. Table 1 lists all of the components with an explanation of the task involved and the
number of items in each language. Two of the seven components focused on meaning
(Object Identification and Sentence Comprehension) while the other five were code based
(Letter/Sound Identification, Word Identification, Word Production, Sentence Production,
and Spelling). All seven components appeared in both home language and English, and a
practice item was provided for each text. All components except the spelling test were
administered in a one-on-one setting.
It is important to note the change in identified targets and action stimuli across the two
assessments; learners were asked to identify different objects in the two tests. Figure 1 shows
one item from the Object Identification area, the first section of each of the assessments. In the
home-language assessment, learners were asked to identify the shoe and in the English assess-
ment, learners were asked to identify the scissors.
TABLE 1
Assessment Components
Test Area Home Language/English Task # Items in each language
Object Identification Both Listen to a word spoken and
identify the picture.
18 (different targets)
Letter/Sound identification Both Listen to a letter sound
dictated and identify it.
4 (different targets)
Word Identification Both Listen to a word spoken and
identify it.
8 (different targets)
Sentence Comprehension Both Listen to a sentence and
identify a picture.
9 (different action
stimuli)
Word Production Both Students read words from a
word list.
6
Sentence Production Both Read aloud a sentence. 2
Spelling Both Students spell words. Group
administered. 6
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
28 SAILORS ET AL.
Subjects and Procedures
All testing was conducted during the month of July as part of the Learning for Living (LfL)
internal evaluation study. Foundation-level students (Grades 1 and 2) in three types of schools
were assessed in both home language and English:
•Intervention schools (that had been selected by the Ministry of Education to participate in
the Learning for Living project)
•Comparison schools (similar demographics and populations);
•Select project schools
Learners were assessed first in their home language and then in English. A random sample of
20 first and second graders was selected and tested at each of the school sites (randomly drawn
from the larger sample). Table 2 shows the number of Grade 1 (n = 793) and Grade 2 (n = 809)
FIGURE 1 Object identification example.
TABLE 2
Frequencies by Home Language
First Graders Second Graders
Home Language Frequency Percent Frequency Percent
Afrikaans 101 12.7 116 14.3
IsiNdebele 60 7.5 59 7.3
Sepedi 60 7.6 60 7.4
Sesotho 118 14.9 130 16.1
Setswana 95 12.0 98 12.1
Tshivenda 80 1.1 80 9.9
IsiXhosa 167 21.1 155 19.2
Xitsonga 60 7.6 59 7.3
IsiZulu 52 6.6 52 6.4
Total 793 10.0 809 10.0
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 29
learners assessed for each of the nine languages. Tests were administered and scored by READ
staff members, who were speakers of the home language of the students. Unfortunately, we only
had posttest data in this study—we recognize this as a limitation of the study.
Table 3 shows the distribution of ages of Grade 1 and 2 learners who were a part of this
study. The majority of the learners in Grade 1 were 7 years of age at the time of this assessment
and 8 years of age in Grade 2.
Analyses
In order to assess the impact of the READ intervention on home- and English-language instruc-
tion, we used Analyses of Variance (ANOVA) to compare the scores of students in the control
and intervention schools. We used confirmatory factor analyses to examine the infrastructure of
the home-language assessment and the English assessment. Once the factor analyses were
completed, we used ANOVA to identify differences in the subscales (aggregate scores sup-
ported by the factor analysis) between the three groups. We have organized our findings around
these three analyses.
RESULTS
To assess the impact of the READ intervention, we compared the scores of students in the inter-
vention and control schools. Overall, learners from the intervention schools outperformed learn-
ers from the control schools on both the home-language assessment and the English assessment,
as illustrated by the total scores in Tables 4–7. Each of these tables shows the breakdown of the
subtest scores. Table 4 shows the results of the Home Language subtests for Grade 1 learners,
comparing intervention schools with the control schools. Statistically significant differences
favored the performance of intervention schools over control schools in all components of the
assessment except one—Sentence Production, where the pattern of performance was in the
TABLE 3
Distribution of Ages
First Graders Second Graders
Age Frequency Percent Frequency Percent
56.800
613917.52.2
7 493 62.2 129 15.9
8 106 13.4 484 59.8
9293.713416.6
10 11 1.4 28 3.5
11 5 .6 15 1.9
12 2 .3 7 .9
13 0 0 5 .6
Missing 2 .3 5 .6
Total 793 10.0 809 10.0
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
30 SAILORS ET AL.
direction of the control schools, but not at a statistically significant level (p > .01). Table 5
shows the descriptive statistics for the English subtests for Grade 1 learners, comparing inter-
vention schools with the control schools. Again, statistically significant differences exist
between the intervention schools and the control schools on every subtest except the Sentence
Production component.
We found similar differences in Grade 2 learners although these differences are not as
pronounced as in Grade 1. Table 6 provides the descriptive statistics for the Home Language
subtests for Grade 2 learners, comparing intervention schools with control schools. Statistically
significant differences existed for the total score (across components) and for three of the indi-
vidual components (Object Identification, Word Identification, and Sentence Comprehension),
with the intervention schools outperforming control schools. It should be noted that both compo-
nents that centered on meaning (Object Identification and Sentence Comprehension) were sig-
nificantly in favor of intervention schools. Although significant differences did not exist for the
other components, the trends favored the intervention schools over the control schools in three
of the components.
The findings for the English subtests were more consistently in favor of the intervention for
Grade 2 learners, as evidenced by the descriptive statistics in Table 7. Statistically significant
TABLE 4
Descriptive Statistics by Home-Language Subtest for READ versus
Control Groups for First Graders
M(SD)F-ratioh2
Object ID
READ 16.21 (1.61) 77.999* .090
Control 14.92 (2.46)
Letter/Sound ID
READ 3.46 (0.89) 23.926* .029
Control 3.11 (1.03)
Word ID
READ 5.85 (2.17) 67.524* .079
Control 4.43 (2.46)
Sentence Comp.
READ 7.28 (1.46) 31.317* .038
Control 6.60 (1.88)
Word Prod.
READ 2.49 (2.09) 30.244* .037
Control 1.65 (1.81)
Sentence Prod.
READ 1.39 (1.73) .656 .0009
Control 1.50 (1.96)
Spelling
READ 2.43 (2.22) 21.716* .027
Control 1.70 (1.71)
Total
READ 19.38 (4.94) 64.224* .075
Control 16.36 (5.01)
* p < .01.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 31
differences favoring the intervention schools were found for all components of the
assessment.
As a very stringent test of the intervention, we compared the scores received by students in
the Select schools (those schools that were found to be high-implementing schools) to the stu-
dents in the other two groups (intervention and control schools) at both grade levels (Table 8).
We found significant differences in several places; these differences are summarized in Table 9.
First, Grade 1 learners in the Select schools outperformed the other groups on all measures of the
English assessment and Sentence Comprehension and Sentence Production on the home-language
assessment. Second, the Select and intervention schools significantly outperformed the control
group on the Object and Word Identification on the home-language assessment, although there
were no significant differences in the Select and intervention schools. Third, the Select schools
outperformed the intervention and control schools (which did not differ significantly from each
other) on the home-language Sentence Production. Interestingly, the intervention schools scored
significantly greater than the control and Select schools (which did not differ significantly from
each other) on the home-language word-identification assessment.
In Grade 2 we found similar differences when comparing the Select schools to the interven-
tion and control schools (see Table 10).
TABLE 5
Descriptive Statistics by English-Language Subtest for READ versus
Control Groups for First Graders
M(SD)F-ratioh2
Object ID
READ 9.64 (3.84) 11.175* .014
Control 8.65 (3.91)
Letter/Sound ID
READ 3.29 (1.14) 22.539* .028
Control 2.87 (1.27)
Word Id.
READ 4.48 (2.38) 16.18* .020
Control 3.75 (2.34)
Sentence Comp.
READ 4.64 (2.69) 19.859* .024
Control 3.76 (2.46)
Word Prod.
READ 1.11 (1.60) 8.109* .010
Control 0.77 (1.39)
Sentence Prod.
READ 0.84 (1.51) 0.279 .0004
Control 0.91 (1.81)
Spelling
READ 0.87 (1.20) 16.697* .021
Control 0.53 (0.88)
Total
READ 17.83 (4.51) 19.502* .024
Control 16.30 (4.62)
* p < .01.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
32 SAILORS ET AL.
As in Grade 1, we found differences between the groups in several places; we present those
differences in Table 11. First, there were significant differences favoring the Select group
over the other two groups on the Object Identification, Sentence Comprehension, Word Pro-
duction, and Spelling on the English assessment. The Select and intervention schools differed
significantly from the control schools (although the Select and intervention schools did not
differ significantly from each other) on the Object Identification and Sentence Comprehen-
sion components of the home-language assessment and on the Word Identification and Sen-
tence Production components of the English assessments. The intervention schools scored
significantly greater than the Select and control schools on the Word Identification measures
for the home-language assessment. The intervention and control schools scored significantly
higher than the Select schools on two of the home-language assessments, including the Word
Production and Spelling components. There was only one component where we found no
significant differences between the three groups—the Sentence Production on the home-
language assessment.
In summary, intervention schools tended to outperform control schools, especially on home-
language measures, but they still fell short of the standard of the Select schools, especially on
the English measures.
TABLE 6
Descriptive Statistics by Home-Language Subtest for READ versus
Control Groups for Second Graders
M(SD)F-ratioh 2
Object ID
READ 16.24 (1.78) 19.159* .023
Control 15.60 (2.17)
Letter/Sound ID
READ 3.51 (1.06) 2.389 .003
Control 3.63 (0.76)
Word ID
READ 6.91 (1.83) 7.476* .009
Control 6.51 (2.12)
Sentence Comp.
READ 7.68 (1.27) 27.331* .033
Control 7.12 (1.65)
Word Prod.
READ 3.87 (1.72) 2.062 .003
Control 3.68 (1.77)
Sentence Prod.
READ 1.41 (0.90) 0.138 .0002
Control 1.39 (1.11)
Spelling
READ 4.20 (1.94) .646 .001
Control 4.08 (1.85)
Total
READ 21.86 (4.36) 9.238* .011
Control 20.87 (4.11)
*p < .01.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 33
To assess the relationship between the students’ first/home language and their success in
learning to read English, we examined the first-order correlations between home- and English-
language test scores (total score and all component scores) across Grades 1 and 2. Table 12 pro-
vides the entire correlation matrix. All correlations differed significantly from zero (p < .001).
We used confirmatory factor analysis to examine the infrastructure of the two assessments.
For each language, we extracted two principle axis factors and used an oblique rotation to model
the factors as being correlated. The results (see Tables 13 and 14) support the extraction of a
Coding factor (consisting of Word Identification, Word Production, Sentence Production, and
Spelling) and a Meaning factor (Object Identification and Sentence Comprehension) as underly-
ing responses to the subscales. (Loadings stronger than 0.7 were considered substantial). Tables
13 and 14 list the factor loadings. In calculating the Coding and Meaning subscale scores, each of
the six subtests were first standardized (converted to Z-scales) We used a factor analysis to reduce
the number of variables and investigated the differences that existed between the two factors that
emerged: Coding (Word Identification, Word Production, Sentence Production, and Spelling)
and Meaning (Object Identification and Sentence Comprehension). In calculating the Coding and
Meaning subscale scores, each of the six subtests were first standardized (converted to Z-scores)
for each of Grades 1 and 2. The resulting Z-scores were then averaged together to provide a
TABLE 7
Descriptive Statistics by English-Language Subtest for READ versus
Control Groups for Second Graders
M(SD)F-ratioh2
Object ID
READ 12.12 (3.10) 53.999* .063
Control 10.29 (3.58)
Letter/Sound ID
READ 3.79 (0.63) 29.318* .035
Control 3.46 (1.04)
Word ID
READ 6.72 (1.80) 49.620* .058
Control 5.64 (2.44)
Sentence Comp.
READ 6.42 (2.04) 43.521* .051
Control 5.32 (2.50)
Word Prod.
READ 3.74 (1.54) 55.335* .064
Control 2.79 (1.95)
Sentence Prod.
READ 1.66 (.73) 22.993* .028
Control 1.34 (1.11)
Spelling
READ 3.49 (1.84) 146.760* .154
Control 1.83 (1.70)
Total
READ 23.58 (3.13) 128.852* .138
Control 20.48 (4.46)
*p < .01.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
34 SAILORS ET AL.
Coding subscale score. Then the Z-scores for the Object Identification and for Sentence Compre-
hension were combined to obtain the Meaning subscale score. The average of standard scores is
zero (by definition) and thus the overall sample mean for each variable and grade was zero.
Analyses of variance (ANOVAs) were conducted to identify whether there were significant
differences between the three groups’ performances on the Meaning and Coding subscales, by
grade. If a significant F-ratio resulted (indicating that at least one of the three groups differed signif-
icantly from the others), then follow-up pairwise comparisons were conducted using Bonnferroni’s
adjustment to preserve the nominal ∝-level of .05 across the three contrasts per subscale.
On the home-language assessments, Grade 1 learners in intervention schools scored signifi-
cantly higher on both the Coding (F[1, 791] = 33.532, p < .0001, h = .041) and the Meaning sub-
tests (F[1, 791] = p <.0001, n2 = .084) than did the control group on the home language
assessments (see Table 15). For second graders, students in the READ program significantly
outperformed students in Control schools on the Meaning subtest (F [1, 805] = 34.121, p = <
.0001, h2 = .041). On the Coding subtest although the difference favored the students in READ
school, the difference was not significant (F[1, 806] = 3.225, p = .073, h2 = .004).
A different pattern characterized the English assessments. Grade 1 learners in Select schools
performed significantly better than students in the intervention and control schools on both the
Coding and Meaning subscales of the English assessment (see Table 16). For first graders,
READ scored significantly higher on both the Coding (F[1, 791] = 9.190, p < .01, h2 = .011) and
TABLE 8
Means (and Standard Deviations) of Home- and English-Language
Test Scores by Type of School for Grade 1, Year 3
Type of School
Subtest Project Comparison Select
Object Identification
Home Language 16.07 (1.60) 14.85 (2.71) 16.39 (1.76)
English Language 9.14 (3.79) 7.22 (3.55) 10.77 (4.43)
Letter/Sound Identification
Home Language 3.47 (.95) 3.04 (1.06) 3.34 (0.97)
English Language 3.39 (1.56) 2.56 (1.43) 3.23 (1.19)
Word Identification
Home Language 5.89 (2.21) 4.28 (2.81) 5.53 (2.21)
English Language 4.25 (2.36) 3.04 (2.17) 4.91 (2.36)
Sentence Comprehension
Home Language 7.19 (1.48) 6.05 (1.91) 7.56 (1.45)
English Language 4.16 (2.67) 2.90 (2.34) 5.48 (2.61)
Word Production
Home Language 2.68 (2.13) 1.90 (2.00) 2.07 (2.13)
English Language 1.03 (1.55) 0.48 (1.28) 1.51 (1.91)
Sentence Production
Home Language 1.80 (2.43) 2.14 (3.03) 3.02 (2.93)
English Language 1.28 (2.30) 0.44 (1.44) 2.31 (2.99
Spelling
Home Language 2.43 (2.01) 1.62 (1.73) 2.10 (2.32)
English Language 0.73 (1.15) 0.37 (.85) 1.13 (1.35)
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 35
the Meaning subtests (F[1, 791] = 18.171, p < .0001, h2 = .022) than did the Control group. Sim-
ilarly, for second graders, students in the READ program significantly outperformed students in
Control schools on both the Coding subtest (F[1, 805] = 114.123, p < .0001, h2 = .124) and on
the Meaning subtest (F[1, 805] = 61.129, p < .0001, h2 = .071).
DISCUSSION
We set out to answer three questions in this analysis. First we wanted to know whether the
READ intervention impacted student performance. The answer to that question is clear: The
differences revealed in this analysis suggest positive and strong effects for the READ Home
TABLE 9
Statistical Significance of Home- and English-Language Test
Scores by Measures by Type of School for Grade 1, Year 3
Measure Note
Significant
Differences
Object Identification
Home Language * (S R)>C
English Language 3 S>R>C
Letter/Sound Identification
Home Language * (S R)>C
English Language * (S R)>C
Word Identification
Home Language * (S R)>C
English Language 3 S>R>C
Sentence Comprehension
Home Language 3 S>R>C
English Language 3 S>R>C
Word Production
Home Language + R> (S C)
English Language 3 S>R>C
Sentence Production
Home Language @ S>(R C)
English Language 3 S>R>C
Spelling
Home Language + R> (S C)
English Language 3 S>R>C
Any conditions inside the same set of parenthesis do not differ
from one another. S = Select Schools, R = READ, C = Control.
3: all three significantly differ from each other (with Select >
READ > Comparison.
*: Select and READ each sig. > Comp though Select and READ
not sig. different.
+: READ sig. greater than Select and Comparison (which do not
differ significantly from each other).
@: Select sig. greater than READ and Comparison (which do not
differ significantly from each other).
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
36 SAILORS ET AL.
Language Initiative. Second, we wanted to know how the project schools compared to a group
of Select schools known for their high levels of implementation. On this score, the project
schools in the intervention clearly have a way to go to match the Select school standards, but
more often than not, the project schools are closer to the Select schools in their performance than
they are to the control schools. Third, we wanted to know whether learning to read in the home
language translated into better performance in English. Our initial data suggest that there is a
high level of value added to performance in both languages when a much more print-rich envi-
ronment can be provided in the home language. These findings challenge the deficit myth often
associated with children who come to school with a first language different from the medium of
instruction. Further, these data suggest that providing instruction in both first and second lan-
guages can have a positive impact on development in both languages.
For South Africa, the short-term implications for this study are transparent. Foundation-phase
English language learners in South Africa perform better in their home language when their
teachers are provided with high-quality instructional materials and accompanying professional
development; moreover, while performance in English does not favor the intervention as consis-
tently as performance in the home language, any differences that do emerge favor the intervention.
These findings are certainly consistent with earlier reviewed tradition of African research
TABLE 10
Means (and Standard Deviations) of Home- and English-Language Test
Scores by Type of School for Grade 2, Year 3
Type of School
Subtest PROJECT Comparison Select
Object Identification
Home Language 16.14 (1.89) 15.46 (1.93) 16.25 (1.89)
English Language 11.46 (3.00) 8.85 (3.12) 13.45 (2.90)
Letter/Sound Identification
Home Language 3.75 (0.63) 3.55 (0.81) 3.11 (1.49)
English Language 3.81 (0.86) 3.20 (1.20) 3.85 (0.55)
Word Identification
Home Language 7.08 (1.86) 6.42 (2.14) 6.31 (2.05)
English Language 6.55 (1.86) 5.15 (2.55) 6.87 (1.58)
Sentence Comprehension
Home Language 7.70 (1.30) 6.87 (1.39) 7.82 (1.22)
English Language 6.04 (1.98) 4.62 (2.13) 7.15 (1.68)
Word Production
Home Language 4.13 (1.58) 3.75 (1.67) 3.07 (2.04)
English Language 3.44 (1.72) 2.45 (1.82) 4.01 (1.58)
Sentence Production
Home Language 3.51 (2.91) 3.18 (3.00) 2.88 (3.02)
English Language 4.30 (2.95) 2.73 (3.01) 4.76 (2.88)
Spelling
Home Language 4.46 (1.71) 4.09 (1.97) 3.16 (2.19)
English Language 3.33 (1.84) 1.31 (1.35) 3.96 (1.76)
Word List
English Language 26.03 (12.89) 13.83 (11.61) 27.60 (11.36)
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 37
validating the efficacy of home-language literacy instruction (Donald et al., 2003; Tambulukani
et al., 1999; Umolu, 1999).
These findings are also supported by our personal experiences as we have traveled through-
out rural South Africa collecting data for this study, working with teachers in both intervention
and control schools. This may seem simple to those of us who teach in places where we have
access to such materials, but such is not the case in most South African schools. We have been
to nonparticipating schools in rural South Africa where students have very limited access to
TABLE 11
Statistical Significance of Home- and English-Language Test
Scores by Measures by Type of School for Grade 1, Year 3
Measure Note
Significant
Differences
Object Identification
Home Language * (S R)>C
English Language 3 S>R>C
Letter/Sound Identification
Home Language ? S>(R C)
English Language * (S R)>C
Word Identification
Home Language + R> (S C)
English Language * (S R)>C
Sentence Comprehension
Home Language * (S R)>C
English Language 3 S>R>C
Word Production
Home Language ? S>(R C)
English Language 3 S>R>C
Sentence Production
Home Language 0 (S R C)
English Language * (S R)>C
Spelling
Home Language ? S>(R C)
English Language 3 S>R>C
Word List
English Language * (S R)>C
Any conditions inside the same set of parenthesis do not differ
from one another. S = Select Schools, R = READ, C = Control.
3: all three significantly differ from each other (with Select >
READ > Comparison.
*: Select and READ each sig. > Comp though Select and READ
not sig different.
+: READ sig. greater than Select and Comparison (which do not
differ significantly from each other).
@: Select sig. greater than READ and Comparison (which do not
differ significantly from each other).
?: READ and Comparison schools each sig. > Select schools
(though READ and comparison do not differ sig.).
0: no sig. diffs.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
38 SAILORS ET AL.
instructional materials, and their teachers receive hardly any professional development on what
to do with the few materials they do have.
We have seen Grade 6 learners carefully placing the only books owned at their school safely
in the school’s vault at the end of the day. We have seen libraries that contain collections of very
outdated books. Nonproject teachers express strong motivation to be part of projects such as this
one, projects that will bring them the precious resource of books. Over and over we have heard
TABLE 12
Correlations Between Test Scores Across Grades
Home-Language Test
Home-Language Test 1 2 3 4 5 6 7 8
1. Object ID .60
2. Letter/Sound ID .22 .66
3. Word ID .32 .47 .83
4. Sentence Comp. .41 .13 .30 .50
5. Word Prod. .28 .39 .66 .29 .90
6. Sentence Prod. .12 .21 .26 .19 .32 .81
7. Spelling .27 .43 .62 .25 .70 .24 .85
8. Total Test Score .57 .62 .79 .56 .79 .51 .76 .90
English-Language Test
English-Language Test 1 2 3 4 5 6 7 8
1. Object ID .79
2. Letter/Sound ID .41 .74
3. Word ID .49 .54 .81
4. Sentence Comp. .68 .37 .50 .75
5. Word Prod. .51 .42 .62 .50 .89
6. Sentence Prod. .40 .34 .48 .39 .50 .87
7. Spelling .44 .36 .53 .42 .68 .37 .81
8. Total Test Score .76 .67 .80 .74 .82 .67 .73 .93
All correlations were statistically significant (p < .001). Internal consistency reliability estimates appear in bold/
italics along the diagonal.
TABLE 13
Pattern Coefficients for Two-Correlated Factors (r = .373) Model
Explored for Six Home-Language Subscale Scores for All Cases
Factor
Test Coding Meaning
Word ID .805 .090
Word Prod. .886 .007
Sentence Prod. .509 −.040
Spelling .861 −.006
Object ID −.014 .852
Sentence Comp. .020 .824
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 39
from teachers, administrators, and community members in the project schools express apprecia-
tion and gratitude for the materials provided by this project. Teachers and administrators express
similar feelings about the professional development, talking expansively about how it has
changed teaching in their schools since the inception of the project. We have come to the con-
clusion that literacy learning cannot take place, at least in rural South Africa, without the proper
TABLE 14
Pattern Coefficients for Two-Correlated Factors
(r = .601) Model Explored for Six English-Language
Subscale Scores for All Cases
Factor
Test Coding Meaning
Word ID .726 .130
Word Prod. .866 .029
Sentence Prod. .67 8 .032
Spelling .894 −.109
Object ID .030 .895
Sentence Comp. −.005 .923
TABLE 15
Descriptive Statistics for Home-Language Coding and
Meaning Subtests by Grade and Group
Coding Subtest Meaning Subtest
NM(SD)M(SD)
First Grade
READ 538 −.18 .79 .07 .72
Control 255 −.51 .68 −.48 1.06
Second Grade
READ 560 .31 .66 .21 .73
Control 248 .22 .69 −.13 .85
TABLE 16
Descriptive Statistics for English-Language Coding and
Meaning Subtests by Grade and Group
Coding Subtest Meaning Subtest
NM(SD)M(SD)
First Grade
READ 538 −.42 .64 −.22 .93
Control 255 −.57 .66 −.52 .91
Second Grade
READ 560 .62 .56 .45 .71
Control 248 .12 .73 −.003 .85
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
40 SAILORS ET AL.
instructional materials—books to read, for sure. But not just any books—books that are develop-
mentally appropriate, interesting, grounded in the cultural and linguistic experiences of the
learners, and accompanied by professional development to help teachers learn how to use them
effectively.
Policy makers are and must remain both encouraged by and concerned with the findings
from this project. The encouragement comes from the cost-effectiveness of this approach. It
has been documented that the costs for the overall Learning for Living from which the current
project emanated were approximately 50 Rand per student. In today’s currency exchange envi-
ronment 50 Rand is less than 7 dollars per learner. The data demonstrate that significant stu-
dent achievement can be seen as a function of an additional investment of 7 dollars per learner.
The concern comes from uncertainty about whether government officials and philanthropic
organizations will be willing to invest even this modest sum in South Africa’s rural students
and teachers.
And finally, as Nelson Mandela said, “If you talk to a man in a language he understands, that
goes to his head. If you talk to him in his language, that goes to his heart.” Because English
guides commerce around the world, ongoing conversations aimed at protecting indigenous
home languages from extinction must continue; we need these languages to support children’s
literacy learning, but we also need them to preserve our planet’s vast and precious linguistic
diversity. Projects like this go a long way toward fulfilling that goal, but they need support—
monetary and political. Hopefully our findings will provide support for organizations and gov-
ernments to continue to support linguistic preservation and diversity.
REFERENCES
Afolayan, A. (1999). The alienated role of the mother tongue in literacy education for sustainable national development:
The Western Nigerian Yoruba example. In S. Manaka (Ed.), Proceedings of the 1st PanAfrican Conference on
Reading for All (pp. 70–88). Pretoria, South Africa: International Reading Association, READ, & UNESCO/
DANIDA.
Akinnaso, F. N. (1993). Policy and experiment in mother tongue literacy in Nigeria. International Review of Education,
39, 255–285.
Brock-Utne, B. (2007a). Language of instruction and student performance: New insights from research in Tanzania and
South Africa. International Review of Education, 53, 509–530.
Brock-Utne, B. (2007b). Learning through a familiar language versus learning through a foreign language—A look into
some secondary school classrooms in Tanzania. International Journal of Educational Development, 27, 487–498.
Brock-Utne, B., & Holmarsdottir, H. B. (2004). Language policies and practices in Tanzania and South Africa: Problems
and challenges. International Journal of Educational Development, 24, 67–83.
Broom, Y. (2004). Reading English in multilingual South African primary schools. Bilingual Education in Bilingualism,
7, 506–523.
De Klerk, V. (2002a). Language issues in our schools: Whose voice counts? Part 1: The parents speak. Perspectives in
Education, 20, 1–14.
De Klerk, V. (2002b). Language issues in our schools: Whose voice counts? Part 2: The teachers speak. Perspectives in
Education, 20, 15–27.
Department of Education of South Africa. (2002). Revised national curriculum statement grades R–9. Pretoria, South
Africa. Retrieved September 1, 2009, from http://www.info.gov.za/view/DownloadFileAction?id=70257
Donald, D., Condy, J., & Forrester, J. (2003). Literacy development: Lessons learned from a pilot project in South
Africa. The Reading Teacher, 56, 484–492.
Du Plessis, S., & Louw, B. (2008). Challenges to preschool teachers in learner’s acquisition of English as Language of
Learning and Teaching. South African Journal of Education, 28, 53–75.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
THE EFFECTS OF FIRST- AND SECOND-LANGUAGE INSTRUCTION 41
Elley, W. B. (2000). The potential of book floods for raising literacy levels. International Journal of Educational
Research, 46, 233–255.
Heugh, K. (1995). From unequal education to the real thing. In K. Heugh, A. Siegruhn, & P. Pluddermann (Eds.), Multi-
lingual education for South Africa (pp. 42–53). Johannesburg, South Africa: Heinemann.
Heugh, K. (2000). The case against bilingual and multilingual education in South Africa. Cape Town, South Africa:
PRAESA.
Heugh, K. 2006. Theory and practice: Language education models in Africa. In H. Alidou, A. Boly, B. Brock-Utne, Y. S.
Diallo, K. Heugh, & H. E. Wolff (Eds.), Association for the development of education in Africa (pp. 56–84).
Libreville, Gabon: Association for the Development of Education in Africa, UNESCO Institute for Education.
Hoffman, J. V., Beretvas, N. & Pearson, P. D. (2004). Improving literacy instruction in South African schools: The
Business Trust’s “Learning for living project.” Thinking Classrooms, 5, 26–33.
Hunt, G. (2007). Failure to thrive? The community literacy strand of the Additive Bilingual Project at an Eastern Cape
community school, South Africa. Journal of Research in Reading, 30, 80–96.
Maile, S. (2004). School choice in South Africa. Education and Urban Society, 37, 94–116.
Matsuura, K. (2003, March). Address to the United Nations Educational, Scientific, and Cultural Organization. Paper
presented at the International Expert Meeting on the UNESCO Program Aimed at the Safeguarding of Endangered
Languages.
Mentz, K., & Van Der Walt, J. L. (2007). Multicultural concerns of educators in the Western Cape Province of South
Africa. Education and Urban Society, 39, 423–449.
Moodley, V. (2007). Codeswitching in the multilingual English first language classroom. International Journal of
Bilingual Education and Bilingualism, 10, 707–722.
O’Connor, J., & Geiger, M. (2009). Challenges facing primary school educators of English second (or other) language
learners in the Western Cape. South African Journal of Education, 29, 253–269.
Sailors, M., Hoffman, J. V., and Matthee, B. (2007). South African schools that promote literacy learning with students
from low-income communities. Reading Research Quarterly, 42, 364–387.
Sailors, M., Hoffman, J. V., Pearson, P. D., Beretvas, N. S., & Matthee, B. (2007). “Learning to read with READ:
Testing the effectiveness the “Learning for Living” project.” In C. S. Sunal & K. Mutua (Eds.), Research on
Education in Africa, the Caribbean, and the Middle East. Greenwich, CN: Information Age Press.
Sailors, M., Makalela, L., & Hoffman, J. V. (in press). Opportunity Matters: The unintended successes of The Ithuba
Writing Program. To appear in Voices in the Middle, 18.
Setati, M. (2008). Access to mathematics versus access to the language of power: The struggle in a multilingual mathe-
matics classroom. South African Journal of Education, 28, 103–116.
Stauffer, R. G. (1967). The language-experience approach to the teaching of reading. New York, NY: Harper & Row.
Tambulukani, G., Sampa, F., Musuku, R., & Linehan, S. (1999). Reading in Zambia: A quiet revolution through the pri-
mary reading programme. In S. Manaka (Ed.), Proceedings of the 1st Pan-African Conference on Reading for All
(pp. 170–175). Pretoria, South Africa: International Reading Association, READ, & UNESCO/DANIDA.
Umolu, J. (1999). Strategies for teaching reading to children with special needs: The Nigerian Perspective. In S. Manaka
(Ed.), Proceedings of the Ist Pan-African Conference on Reading for All (pp. 212–217). Pretoria, South Africa:
International Reading Association, READ, & UNESCO/DANIDA.
UNESCO (1997). Summary of the Intergovernmental Conference on Language Policies in Africa. Retrieved March 22,
2004, from http://portal.unesco.org/education/ev.php?URL_ID=13132&URL_DO=DO_TOPIC&URL_SECTION=201
UNESCO. (2004). UNESCO guidelines on language and education. Retrieved March 22, 2004, from http://portal.
unesco.org/culture/en/ev.php@URL_ID=18571&URL_DO=DO_TOPIC&URL_SECTION=201.html
Young, Y. (1995). Preparing teacher trainees to teach in multilingual classes. In K. Heugh, A. Siegruhn, & P. Pluddermann
(Eds.), Multilingual education for South Africa (pp. 63–70). Johannesburg, South Africa: Heinemann.
Downloaded By: [Sailors, Misty] At: 22:51 27 April 2010
