Two-Way Immersion Programs: Features and Statistics.
by Howard, Elizabeth R. - Sugarman, Julie
Since 1991, the Center for Applied Linguistics has monitored the growth
of two-way immersion (TWI) programs in the United States. This information
is published online in the Directory of Two-Way Immersion Programs in the
United States (2000), accessible at www.cal.org/twi/directory.
TWI programs integrate language-minority and language-majority students
for all or most of the school day and strive to promote bilingualism and
biliteracy in addition to grade-level academic achievement for all students
(Christian, 1994). Programs listed in the Directory conform to this general
definition of TWI through adherence to the following criteria: 1) language-minority
and language-majority students are integrated for at least 50% of the day
at all grade levels; 2) content and literacy instruction in both languages
are provided to all students; and 3) language-minority and language-majority
students are balanced, with each group making up one third to two thirds
of the total student population.
The information in the Directory is self-reported; it is based on responses
to a questionnaire completed by representatives of the programs listed.
A new feature of the directory is a search function that makes it possible
to identify programs that share characteristics, such as location, student
demographics, and program design. This digest summarizes findings from
data in the directory regarding features of TWI programs, students, and
staff. National data are presented, along with similar data from the three
states with the most programs--California (86), Texas (34), and New York
The first TWI program in the United States began in 1963. For the next
20 years, the growth of TWI programs was minimal, with fewer than 10 documented
programs in operation before 1981. The majority of programs in existence
today were established during the past two decades. The 2000 Directory
includes 248 TWI programs in 23 states and the District of Columbia. There
has also been considerable expansion within existing programs: Many have
reported adding new grade levels each year, and 40 programs now extend
into middle or high school.
The majority of TWI programs are in public schools; only four are operated
by private schools. Nearly a quarter of the public school programs operate
in specialized environments: 11 are housed in charter schools and 53 in
magnet schools. California has the most programs operating in specialized
environments, with eight charter school programs and 22 magnet school programs.
Relatively few TWI programs (32) are whole-school programs. About three
quarters of the elementary programs (191) operate as strands within schools,
as do all of the secondary programs (32). Twenty-five programs did not
respond to this question.
Languages of Instruction
Most TWI programs are Spanish/English (234). The other programs are
Chinese/English (5), French/English (5), Korean/English (3), and Navajo/English
(2). (One school houses both a Spanish/English and a Chinese/English program.)
The majority of students enrolled in these programs are native speakers
of one or both languages of instruction. In 37 programs, however, more
than 1% of the students are native speakers of a language not used in the
program (i.e., third language speakers). In nine programs, 5% are third
PROGRAM MODELS AND LITERACY INSTRUCTION
A key decision in initiating a TWI program is the choice of a program
model. Nationally, one of the two most common program models is minority-language
dominant, which is used in 104 schools (42%). In these "90/10" or "80/20"
programs, the minority language is used for instruction 80-90% of the time
in the primary grades, with the instructional ratio of the minority language
to English generally reaching 50/50 by fourth grade. An additional 85 programs
(33%) are balanced programs ("50/50"); the amount of instructional time
is equal in the two languages at all grade levels. Only five programs (2%)
separate students by native language for part of the day in the primary
grades and provide differing amounts of instruction in the two languages.
Thirty-two programs (13%) are middle or high school programs, with models
that differ from the elementary model (Montone & Loeb, 2000). Twenty-two
programs (9%) provided no response.
An interesting pattern emerges through examination of the predominant
program models used in California, Texas, and New York. In California,
most programs (63%) are minority-language dominant. In Texas, the percentages
of minority-language dominant programs (41%) and balanced programs (47%)
are roughly equal. In New York, the majority of programs (60%) use the
balanced model. There seem to be regional norms for TWI programs, with
no single program model being dominant across the three states. California
has the highest percentage of middle and high school programs (19% of the
state's TWI programs).
Another essential decision that all TWI programs must make is the language(s)
in which initial literacy instruction will be provided. Nationally, 31%
of the programs use the minority language for initial literacy instruction
for all students, 22% provide initial literacy instruction in both languages
to all students, 20% separate the children by native language for initial
literacy instruction, 1% use English for all students, 14% do not serve
grade levels that require initial literacy instruction, and 12% are unreported.
Compared to the national picture, California has a larger percentage
of programs (53%) that use the minority language to introduce literacy
instruction to all children, and smaller percentages that use both languages
for all students (15%) or separate the students by native language (5%).
The pattern in Texas more closely mirrors that of the nation, with 41%
using the minority language for all students, 26% using both languages
for all students, and 18% separating the students by native language. In
New York, only one program (5%) reported using the minority language for
initial literacy instruction for all students, while 20% separate the students
by native language, and 40% use both languages for all students. It is
interesting to note that while the three main approaches to initial literacy
instruction are practiced in all three states, regional norms appear to
influence the extent to which each of the three options is implemented.
The percentages do not add up to 100%, because they do not include programs
that do not teach primary grades nor programs that did not provide responses
to this question.
The stereotypical TWI program is composed of two populations: Latino,
low-income, native Spanish speakers and White, middle class, native English
speakers. Although there are certainly programs that conform to this stereotype,
directory searches indicate there is greater diversity in the student populations
of current TWI programs than the stereotype suggests.
Because students who are classified as native speakers of the minority
language generally belong to a single racial or ethnic group (e.g., Latino
in the case of native Spanish speakers), the racial and ethnic make-up
of the native English speakers is a more useful indicator of the overall
diversity of TWI programs. Nationally, a majority of TWI programs (54%)
have a mixture of ethnicity, with no one ethnic group making up more than
75% of the native English speakers. Another 34% of programs have a predominant
racial or ethnic group among their native English speakers, but the specific
racial or ethnic make-up of that group varies across programs:
Percent of programs Race/ethnicity of more than 75% of native English
1% Native American
12% no response from program
Like the national norm, both California and New York have a majority
of programs with no clear racial or ethnic majority among native English
speakers; however, the percentages of such programs in these states are
higher than the national figure (54%), with 66% and 60% respectively. At
35%, Texas has a much lower percentage of programs with no clear racial
or ethnic majority among its native English speakers. It also has a much
higher percentage of programs where more than 75% of the native English
speakers are Latino (35% compared to 7% for California and 10% for New
York) and a slightly higher percentage of programs where more than 75%
of the native English speakers are White (21% compared to 13% for California
and 15% for New York). There are no programs in Texas or New York that
have a majority of African Americans or Asians, and no programs in any
of the three states have a majority of Native Americans.
Because eligibility for free or reduced lunch is determined by family
income, children who participate are often classified as being at risk
for low academic performance due to poverty. Working with this definition,
TWI programs appear to serve a sizable at-risk population of both native
English speakers and language minority students. Nationally, about one
third of programs (80 schools or 32%) report that more than half of both
native English speakers and language minority students participate in a
free or reduced school lunch program. California has 19 programs (22%)
in which more than half of both the native English speakers and language
minority students receive free or reduced lunch, Texas has 17 (50%), and
New York has 9 (45%).
Data confirm that there are more low-income language minority students
than low-income native English speakers enrolled in TWI programs. Nearly
one quarter of the programs (60 programs or 24%) report that more than
half of their language minority students and less than half of their native
English speakers receive free or reduced lunch, but no schools report that
more than half of their native English speakers and less than half of their
language minority students receive free or reduced lunch. This profile
holds for each of the three states. In California, 30 programs (35%) report
that more than half of their language minority students and less than half
of their native English speakers receive free or reduced lunch. Texas has
8 such programs (24%), and New York has one (5%).
A serious concern of TWI programs is the limited availability of qualified
bilingual teachers and support staff. (Support staff includes instructional
assistants, bilingual program coordinators, parent liaisons, and so forth.)
However, more than half (54%) of all TWI programs reported that 100% of
their teachers are proficient in both program languages. The percentage
in California is even higher, with 70% of the programs reporting that 100%
of their teachers are bilingual in the languages of instruction. In Texas
(40%) and New York (45%), the percentages are just below the national figure,
with slightly less than half of the programs in each state reporting that
all teachers are bilingual.
Nationally, only 29% of TWI programs report that 100% of their support
staff are proficient in both languages of instruction. Again, California
has a higher percentage than the national figure, with 33% of programs
reporting that 100% of the support staff are bilingual. Texas also has
a higher percentage than the nation (44%), and interestingly, a slightly
higher percentage of programs in which 100% of support staff are bilingual
than programs in which 100% of teachers are bilingual. New York is just
below the national average, with only 20% of its programs reporting that
all teachers and support staff are bilingual in the languages of instruction.
Fewer than 10% of programs, nationally and for all three states, report
that fewer than half of their teachers and staff are proficient in both
This digest provides a more detailed description of TWI programs, students,
and staff on a national level than has previously been documented. It also
describes the variation in programs by geographical region, something that
has not been discussed in the literature to this point. This digest can
thus serve as a useful starting point for those conducting research on
TWI programs. Practitioners in TWI programs may also find this document
helpful as a way of placing their local situation in a national context.
Finally, as many TWI programs are funded in part by federal grants from
the U.S. Department of Education, this digest may be of interest to policymakers
who want to know more about the types of programs and students that are
supported through such grants.
Christian, D. (1994). "Two-way bilingual education: Students learning
through two languages" (Educational Practice Rep. No. 12). Santa Cruz,
CA, and Washington, DC: National Center for Research on Cultural Diversity
and Second Language Learning.
Directory of Two-Way Bilingual Immersion Programs in the United States.
(2000). Washington, DC: Center for Applied Linguistics. Available: www.cal.org/twi/directory
Montone, C., & Loeb, M. (2000). "Implementing two-way immersion
programs in secondary schools" (Educational Practice Rep. No. 5). Santa
Cruz, CA, and Washington, DC: Center for Research on Education, Diversity