See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/360646493

Longitudinal study of Spanish Dual Language Immersion graduates:

Secondary school academic and language achievement

Article  in  Foreign Language Annals · May 2022

DOI: 10.1111/flan.12615

CITATIONS READS

0 14

5 authors, including:

Amado M. Padilla Margaret Peterson

Stanford University Stanford University
176 PUBLICATIONS   6,368 CITATIONS    4 PUBLICATIONS   4 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Child bilingualism View project

Longitudinal Study of Graduates of a Spanish Dual Immersion Program View project

All content following this page was uploaded by Amado M. Padilla on 23 May 2022.

The user has requested enhancement of the downloaded file.

Received: 23 March 2021 | Accepted: 21 February 2022

DOI: 10.1111/flan.12615

ORIGINAL ARTICLE

Longitudinal study of Spanish Dual Language

Immersion graduates: Secondary school
academic and language achievement

Amado M. Padilla | Xinjie Chen | Elizabeth Swanson |

Margaret Peterson | Amy Peruzzaro

The Challenge
How do graduates of elementary Dual Language Immersion (DLI) programs compare
to non‐DLI students in secondary school? The challenge is the scarcity of longitudinal
research on DLI graduates. We followed 5th grade DLI students through 12th grade
finding that they continue language study and that their achievement in English
Language Arts and Mathematics is similar to non‐DLI students.

Graduate School of Education, Stanford

University, Stanford, California, USA
Correspondence
Xinjie Chen, Graduate School of
Education, Stanford University, Stanford,
CA, USA.
Email: xjchen96@stanford.edu
Abstract
Five cohorts of students (N = 322) who completed a
K–5 Spanish Dual Language Immersion (DLI) program
in an urban school district were followed longitudi-
nally through middle and high school completion.
Academic and language achievement data included
English Language Arts and Mathematics scores on the
Standardized Testing and Reporting (STAR) test for
grades 3–8 and on the California Assessment of
Student Performance and Progress (CAASPP) for 11th
grade, 10th‐grade cumulative high school grade point
average, enrollment in and grades earned in language
classes, and attainment of the California Seal of
Biliteracy. The DLI students were compared to non‐
DLI students who attended the same schools. Most DLI
students continued with advanced level classes in

© 2022 American Council on the Teaching of Foreign Languages.

Foreign Language Annals. 2022;1–27. wileyonlinelibrary.com/journal/flan | 1

2 | PADILLA ET AL.

Spanish in high school, and nearly half earned the

California Seal of Biliteracy. Achievement results
indicate that DLI graduates performed at equivalent
or higher levels than non‐DLI students on standardized
achievement measures for English Language Arts and
Mathematics.

KEYWORDS
dual language immersion, longitudinal study, English language
achievement, math achievement, Spanish, seal of biliteracy

1 | INTRODUCTION

1.1 | Overview of Dual Language Immersion (DLI) program

DLI programs in the United States offer academic content instruction in two languages, English
and another language (e.g., Spanish or Mandarin). Such programs enroll English Learners
(ELs) who are native speakers of the minority language (e.g., Spanish) as well as English Only
(EO) speaking students. DLI programs have expanded rapidly during the past two decades in
the United States. The proliferation of DLI programs in the United States is notable because
most elementary students do not have the opportunity to study a second language (Devlin,
2015). The goal of these programs is to promote bilingualism, biliteracy, academic achievement,
and sociocultural competence (Christian et al., 2000). Indeed, research has shown that students
in well‐implemented DLI programs, both ELs and EO students, reach higher levels of academic
achievement in content areas, become bilingual and biliterate, and gain intercultural
competence (Howard et al., 2018; Padilla et al., 2013).
There are different types of DLI programs: one‐way, principally serving EO students or, if
predominantly serving students classified as ELs, those referred to as developmental bilingual
(see Howard et al., 2018); and two‐way programs that serve both ELs (students with a home
language that is not English) and EOs. Different models have been implemented in practice.
For example, in a 90:10 model, kindergarteners spend 90% of the school day learning academic
content in the minority language and 10% in English. The percentage shifts gradually each year
until by the 4th and 5th grades the percentage of minority language to English instruction
reaches 50:50 and students emerge from elementary school with high levels of proficiency in
both Spanish and English (Rocque et al., 2016). The majority of DLI programs are found in
elementary schools and parents are encouraged to make a long‐term commitment to DLI by
ensuring that their children remain in the program through the completion of elementary
school. Successful DLI programs include the following criteria: (1) Instruction takes place
in English and the minority language with instruction occurring in one language at a
time, without translation; (2) ELs and EO speakers do all content learning in both languages;
and (3) ELs and EO speakers stay together for all academic content instruction
(Lindholm‐Leary, 2012).
After students graduate out of an elementary immersion program, they transition to a
middle school, and, depending on the resources available to the school district, there may or
PADILLA ET AL. | 3

may not be a continuation of the immersion language. Many districts offer one or two classes in
the immersion language; however, most districts do not have the resources to provide a DLI
pathway that continues through middle school. Most commonly, the DLI students will have an
advanced language class (e.g., Spanish) and a content‐based class (e.g., history) taught in
Spanish. Such an arrangement may continue through the completion of middle school. Very
few school districts make any instructional provisions for DLI students once they transition to
high school; students wishing to continue with minority language instruction typically are able
to enroll in advanced language classes (e.g., Level 3 Spanish) because of their proficiency in the
language.

1.2 | Effects of DLI programs

Numerous studies have demonstrated that ELs and EO students who are enrolled in DLI
programs often achieve greater academic growth when compared with peers not enrolled in a
DLI program (e.g., Lindholm‐Leary, 2012; Rocque et al., 2016). For example, in a study of Utah
DLI programs, DLI students at the end of 6th grade were found to markedly outperform their
non‐DLI counterparts on standardized tests in language arts, mathematics and science, and this
held true for EO speakers and ELs (Steele et al., 2019). Burkhauser et al. (2016) found that, after
enrolling in a 4‐year immersion program (K‐3), native Spanish‐speaking students performed as
well as their EO peers in reading and speaking, and even outperformed EOs in listening and
writing. However, the existing studies' results remain inconsistent and positive benefits of
immersion are not always found. It should be noted that many extraneous variables can impact
the results of DLI program graduates when compared to non‐DLI students, including external
factors (e.g., ethnicity, socioeconomic status, gender) and internal factors (e.g., motivation)
(Baker & MacIntyre, 2000; Christian et al., 2000).
Are there continuing positive benefits of the elementary DLI program when students
transition to a secondary school? While research at the secondary level is sparse, there are a few
studies that highlight the benefits of DLI beyond the elementary years. Studies reveal positive
gains in both attitude toward school and academic performance (Lindholm‐Leary & Borsato,
2005; Steele et al., 2017). Lindholm‐Leary and Borsato (2005) found that DLI programs prepare
high school students, both EO and reclassified EL learners, to have positive attitudes towards
school, including mathematics. In their study of 139 high school students, those who attended
DLI programs were more successful than those who did not, and they were also enrolled in
higher level, college preparatory classes (Lindholm‐Leary & Borsato, 2005). Steele et al. (2017)
found that immersion students outperformed their peers on state mandated English Language
Arts (ELAs) tests by 9 months of learning in the 8th grade. In examining Mathematics and
Science scores, Steele et al reported no statistically significant benefit for immersion students,
but no detriment either (Steele et al., 2017). In addition, the study found an immersion
advantage in EL reclassification in later grades with earlier reclassification of EL students to
designation of English proficient.
Overall, there is a significant research gap in the DLI literature in what becomes of DLI
students as they transition from elementary to middle school and then again as they transition
to high school and eventually graduate. While there are a plethora of studies examining
the effects of DLI programs at the elementary level (e.g., Lindholm‐Leary & Block, 2010;
Marian et al., 2013; Padilla et al., 2013), there is a dearth of research at the high school level
4 | PADILLA ET AL.

(Mellgren & Somers, 2008; Robins, 2017). The limited research that exists gives some insight
into the long‐term effects of immersion education on academic outcomes (Steele et al., 2017).
Based on the research gap mentioned above, there is a need for further research on
academic and language outcomes at the secondary level for students who began the DLI
program as EO speakers and students who began the DLI program as English Language
Learners.

2 | RESEARCH AIMS AND Q UESTIONS

The purpose of this study was to investigate longitudinal DLI student data through middle and
high school completion to understand the differential impact of DLI on students' academic and
language outcomes. We followed several cohorts of students in an urban school district who
completed elementary school in a Spanish DLI program. They then transitioned to a middle
school program where they had the option of continuing with a modified DLI program
consisting of two classes in Spanish: (1) a Spanish language and culture class and (2) a social
studies class taught in Spanish. Following completion of middle school, the majority of the
students transitioned to a nearby high school where they had the freedom to continue with
Spanish instruction, switch to a different language, or discontinue language instruction. Three
research questions were developed to examine DLI graduates' long‐term academic and
language outcomes:

1. Do graduates of a Spanish DLI program continue with advanced Spanish language classes
and/or other world language classes in middle school and high school? What are their
language outcomes, as measured by grades in middle school Spanish classes and rates of
earning the California Seal of Biliteracy at high school completion?
2. Do ELs who enter a Spanish immersion program in kindergarten differ in academic
outcomes on state mandated achievement tests and other academic achievement indices
(i.e., Standardized Testing and Reporting [STAR] scores for 5th, 6th, and 7th grade,
California Assessment of Student Performance and Progress [CAASPP]/smarter balanced
assessment consortium [SBAC] scores for 11th grade, grade point average (GPA) from 9th to
12th grade, and completion at high school graduation of required classes for admission to a
California public university [A‐G requirements]) from EO peers as they transition from
elementary to secondary school and progress to high school graduation?
3. Do graduates of a K–5 Spanish DLI program perform at the same levels academically
(as measured by high school GPA, STAR and CAASPP/SBAC test scores, and rates of
earning the California Seal of Biliteracy) as a comparison group of students who did not
attend a DLI program in elementary school?

3 | METHOD

3.1 | Participants

The DLI program featured in this study is located in an urban K‐8 elementary school district in
Northern California. The Spanish DLI program began in kindergarten and continued through
the 5th grade, and was guided by a 90:10 model. Following completion of 5th grade, students
PADILLA ET AL. | 5

were able to transition to a middle school in the same school district that provided a partial
Spanish immersion experience in grades 6–8 exclusively for the former DLI students. In middle
school, two classes were offered—Spanish language arts and social studies—in which content
was taught exclusively in Spanish.
Five cohorts of students (N = 322) who completed the K‐5 DLI program were followed
longitudinally beginning at the end of 5th grade through middle and high school (most
students went to a nearby unified high school district) in the same urban area. Cohorts 1–3
were tracked through their 12th grade graduation, Cohort 4 through the 11th grade, and
Cohort 5 through the 10th grade. These were the latest available data at the time of
analysis.
Table 1 is an overview of the beginning and ending years of the study, as well as the number
of students during relevant years, for each cohort. Over time the number of students with
available data decreased because some students left the school district or enrolled in private
secondary schools.

3.2 | Demographic information

3.2.1 | Gender

The combined sample consisted of 164 female students and 158 male students at the end of the
5th grade.

3.2.2 | Ethnicity

Table 2 reports the ethnicity of the students in the sample. A large majority of
students (over 80%) identified as Hispanic, with the second largest group being White
students (12%).

TABLE 1 Number of students across five cohorts of graduates of the Spanish Immersion Program
5th–12th grade
academic years N (5th grade) N (8th grade) N (12th grade)
Cohort 1 2010–2017 46 41 33
Cohort 2 2011–2018 70 62 48
Cohort 3 2012–2019 60 48 48
Cohort 4 2013–2020 63 53 46a
Cohort 5 2014–2021 83 61 (no data)
Total 322 265 129 (cohorts 1–3)
a
We report data from the end of 5th grade and at the 12th grade graduation because these are the starting and ending years of
the study, as well as data from the end of the 8th grade because this grade provides a useful intermediate reference point. We
report the number of Cohort 4 students in 11th grade rather than 12th grade, and we do not have data for Cohort 5 after 8th
grade due to the early closure of schools due to COVID‐19 2020 and the break in assessment testing.
6 | PADILLA ET AL.

TABLE 2 Ethnicity of DLI graduates in 5th grade

Ethnicity Number of students Percent of sample
Hispanic/Latino 268 83.2
White 39 12.1
Asian/Pacific Islander 9 2.8
Black/African American 4 1.2
Decline to state 2 0.6
Abbreviation: DLI, Dual Language Immersion.

TABLE 3 English language status of DLI graduates at the end of 5th, 8th, and 10th grade
Number of students Number of students Number of students
English language status (5th grade) (8th grade) (10th grade)
English Learner (EL) 94 (29%) 31 (12%) 16 (7%)
English Only (EO) 93 (29%) 77 (29%) 75 (33%)
Initial Fluent English 47 (15%) 40 (15%) 31 (14%)
Proficient (IFEP)
Reclassified Fluent English 88 (27%) 117 (44%) 103 (46%)
Proficient (RFEP)
Abbreviation: DLI, Dual Language Immersion.

3.2.3 | English language status

There were four possible classifications for students' English language proficiency status: EL
students are considered not yet proficient in English. EO students speak English as their only
language at home. Reclassified Fluent English Proficient (RFEP) students were previously
classified as ELs, but are re‐classified as English Proficient based on their scores on the
California English Language Development Test (CELDT). Finally, Initial Fluent English
Proficient (IFEP) students speak a language other than English at home, but are classified as
English proficient the first time they take the CELDT.
Table 3 shows the English language proficiency status of the students in our sample in 5th,
8th, and 10th grade. Students started out fairly split between the four proficiency classifications
in 5th grade, but through 8th grade and 10th grade the number of EL students diminished as
most of these students were redesignated as RFEP. There was some attrition in every category
over time, likely due to students leaving the district or switching to private schools.

3.2.4 | Primary language

Most students spoke Spanish as their primary language at home. In 5th grade, 226 (70%)
students indicated that Spanish was their primary language, and 93 (29%) reported that
English was their primary language. One student spoke primarily French and two spoke
primarily Cantonese.
PADILLA ET AL. | 7

TABLE 4 Parent education level of DLI graduates in 5th grade

Parent education level Number of students Percent of sample
Did not complete high school 72 22.4%
High school graduate 72 22.4%
Some college 78 24.2%
College graduate 53 16.5%
Graduate degree or higher 36 11.2%
Decline to state 11 3.4%
Abbreviation: DLI, Dual Language Immersion.

TABLE 5 Academic and language outcomes

Academic achievement outcomes Grade level
ELA and Mathematics scores on Standardized Testing and 5–8
Reporting (STAR) test
Cumulative high school GPA 10 (highest grade level for which we have data
from all cohorts)
ELA and Math scores on California Assessment of Student 11 (grade level for which CAASPP is most
Performance and Progress (CAASPP) test consistent; does not include Cohort 5)
Completion of UC/CSU A–G high school course 12
requirements
Language achievement outcomes Grade level
Middle school grades in Spanish Language classes and 6–8
Social Studies classes taught in Spanish
High school language classes taken 9–12
Seal of Biliteracy attainment 12

3.2.5 | Parent education level

Parent education level was determined by recording the number of years of schooling for the
parent with the most education in a family. Parent education level of students exiting the 5th
grade was widely distributed, as shown in Table 4.

3.2.6 | Socioeconomic status—Free/reduced lunch status

The majority of students (225, or 70%) were not eligible for free or reduced lunch, meaning that
the family's socioeconomic status was not considered low‐income. In 5th grade, a total number
of 71 (22%) students qualified for free lunch status while another 26 (8%) were eligible for
reduced lunch status.
8 | PADILLA ET AL.

4 | MEASURES

4.1 | Measurement of academic and language outcomes

Table 5 shows the academic and language outcome measures we had access to for the DLI
graduates, as well as the grade levels they were available for.

5 | PROCED U R ES

The data for the study were secured through a collaborative agreement between Stanford
University and a surrounding public elementary (K–8) school district and a separate secondary
district. The K–5 school where the DLI program is situated and its accompanying middle school
which houses the partial immersion program for DLI graduates have been in existence for over
25 years. Upon completion of the 8th grade, most of the DLI graduates transition to the
neighboring high school district in the same urban center. The school records of both districts
are housed at a data center maintained by the Graduate School of Education at Stanford
University where they are available for research if approved by the collaborative partners.

6 | AN AL YTICAL P LAN

Because we had access to the Statewide Student Identifier (SSID) and the local student ID, we
were able to track students' outcomes in middle school and high school for the analysis. We
obtained descriptive results and graphs for the academic and language achievement measures
using R software version 4.0.4 for data analysis (R Core Team, 2021). unde For comparisons
between DLI students of differing EL status and between DLI students and non‐DLI students,
we conducted Tukey's HSD (honest significant difference) tests, which control for false
positives by accounting for the fact that multiple significance tests are carried rout
simultaneously. We also report the Cohen's d measure of effect size for differences between
means (Plonsky & Oswald, 2014).

7 | RESULTS

7.1 | Spanish language outcomes of all DLI students

Research Question 1 was: Do graduates of a Spanish DLI program continue with advanced
Spanish language classes and/or other world language classes in middle school and high school?
What are their language outcomes, as measured by grades in middle school Spanish classes and
rates of earning the California Seal of Biliteracy at high school completion?
In addressing Research Question 1, we focused on DLI graduates' language trajectories after
completing the DLI program. We examined the language classes DLI students elected to take in
middle school and high school, as well as their letter grades in their middle school classes. This
allowed us to determine whether DLI students continued with Spanish language classes
(the target language of the DLI program) and the level of language classes selected by students.
PADILLA ET AL. | 9

F I G U R E 1 Proportion of DLI students earning each letter grade in Spanish language arts. DLI, Dual
Language Immersion. [Color figure can be viewed at wileyonlinelibrary.com]

F I G U R E 2 Proportion of DLI students earning each letter grade in a social studies class taught in Spanish.
DLI, Dual Language Immersion. [Color figure can be viewed at wileyonlinelibrary.com]

7.1.1 | Middle school Spanish classes

The majority of DLI students continued to take Spanish classes after graduating from the DLI
program. In middle school, over 90% of the students for whom we have data participated in a
partial Spanish immersion program each year through grades 6, 7, and 8. Students earned high
letter grades in these classes over the course of middle school. Figure 1 summarizes the grade
10 | PADILLA ET AL.

breakdown of DLI graduates in Spanish Language Arts, and Figure 2 summarizes the grade
breakdown for a Social Studies class taught in Spanish. We analyzed only students' actual letter
grades, such that an A+ or an A− would still both be counted as an A. Each year, at least 80% of
students earned an A or B average in their two Spanish immersion classes, with about 50%
earning an A and about 30% earning a B.

7.1.2 | High school language classes

We examined the high school class enrollment data to determine what language classes and
their respective course levels the 118 former DLI students for whom we have complete data
took throughout all 4 years of high school. This allowed us to understand students' language
class trajectories over the course of their high school career. Spanish class enrollment data is
presented in Figure 3. The language enrollment data confirms that it was common for students
to take Spanish for Native Speakers and to take advanced Spanish classes at some point in high
school. In total, about 40% of students took Spanish 1 and/or 2 for Native Speakers and
interestingly, 60% took Spanish 3 and/or 4 for Native Speakers. Over 60% of students enrolled in
an advanced Spanish class at some point in high school took AP Spanish (34%) or IB
Spanish (27%).
A small number of these DLI graduates elected to take language classes other than Spanish
in high school. About 10% enrolled in French classes, though few (less than 5%) went on to take
AP or IB French. Less than 2% of DLI graduates took high school Chinese classes. In general,

F I G U R E 3 Proportion of DLI students who enrolled in each Spanish Class in high school. DLI, Dual
Language Immersion. [Color figure can be viewed at wileyonlinelibrary.com]
PADILLA ET AL. | 11

TABLE 6 DLI students' rates of earning the California Seal of Biliteracy in 12th grade
Number of students Total number of Percent of students
who earned Seal of students in who earned Seal of
Student group Biliteracy category Biliteracy
DLI English Only (EO) 30 59 50.8%
DLI Initial Fluent English 17 30 56.7%
Proficient (IFEP)
DLI Reclassified Fluent 28 79 35.4%
English Proficient (RFEP)
All DLI students 75 168 44.6%
Abbreviation: DLI, Dual Language Immersion.

the overwhelming majority of DLI graduates continued to take Spanish classes, with many
continuing to advanced levels.

7.1.3 | Seal of biliteracy

Turning our attention now to the California Seal of Biliteracy, the results are summarized
in Table 6. The data show that 44.6% of the DLI students who graduated from high school
received the Seal of Biliteracy. Students classified initially as fully English proficient, aka
IFEP, earned the Seal at the highest rate (56.7%). Meanwhile, 35.4% of students who
were reclassified as English proficient and 50.8% of EO students received the Seal. The few
students who were still classified as ELs in 12th grade were not eligible to earn the
Seal.
This finding suggests that DLI programs achieve their goal in helping students gain high
levels of literacy skills in both their home language and a second language.

7.1.4 | Research Question 1 summary

A majority of DLI graduates for whom we have data continued to enroll in Spanish classes
throughout middle and high school. In middle school, the majority of DLI graduates continued
into the partial immersion track and earned high grades in their Spanish classes, with over 80%
of students receiving an A or B. In high school, a majority of students enrolled in an advanced
Spanish class, which may suggest that they aim to reach high levels of proficiency in Spanish
even after the DLI program is complete.
Importantly, DLI students across all proficiency groups (i.e., EO, IFEP, and RFEP)
earned the California Seal of Biliteracy at high rates. The Seal of Biliteracy is gaining
increasing prominence nationwide as a major achievement in the study of a second
language. It is therefore crucial to note that participating in early DLI programs may play an
important role in encouraging and preparing students to earn the Seal later on in their
academic trajectory.
12 | PADILLA ET AL.

7.2 | Academic outcomes and EL classification among DLI students

Research Question 2 was: Do ELs who enter a Spanish immersion program in kindergarten differ
in academic outcomes on state mandated achievement tests and other academic achievement
indices (i.e., STAR scores for 5th, 6th and 7th grade, CAASPP/SBAC scores for 11th grade, GPA
from 9th to 12th grade, and completion at high school graduation of required classes for
admission to a California public university [A‐G requirements]) from EO peers as they transition
from elementary to secondary school and progress to high school graduation?
In Research Question 2, we broadened our focus from DLI graduates' language trajectories
to students' overall academic achievement after exiting the DLI program. To address this
question, we first examined how DLI students performed on academic achievement measures
in middle school and high school. We were particularly interested in how DLI students'
outcomes might differ depending on their English language proficiency status: whether
students were classified as ELs, RFEP, Initially Fluent English Proficient (IFEP), or EO. Lastly,
we examined what proportion of students in the different language groups completed the
classes required for admission to a California public university. These are referred to as the
UC/CSU A–G required classes that denote a student as college ready.

7.2.1 | STAR scores

Because the State of California discontinued STAR testing halfway through the data collection
process, STAR scores were only available for 230 students for grades 5–8.

F I G U R E 4 DLI students' STAR scores in English language arts. Mean scores and bootstrapped confidence
intervals are shown for students from each EL proficiency classification. DLI, Dual Language Immersion; EL,
English Learner; STAR, Standardized Testing and Reporting. [Color figure can be viewed at
wileyonlinelibrary.com]
PADILLA ET AL. | 13

7.2.1.1 | STAR ELA scores

Figure 4 presents the mean STAR scores for students from each EL classification, collapsed
across grades 5–8. Across grade levels, EL students (M = 326.39; SD = 43.27; p < .01; d = 1.28)
had significantly lower STAR ELA scores than EO students (M = 392.59; SD = 56.63). However,
IFEP students (M = 391.47; SD = 48.92; p = .99; d = 0.02), and RFEP students (M = 379.37;
SD = 46.5; p = .20; d = 0.26) did not differ in their scores compared to EO students. Students in
all 3 groups performed similarly on the STAR ELA measure.

7.2.1.2 | STAR mathematics scores

Figure 5 shows the mean STAR mathematics scores in grades 5–8 for each EL classification
group. Similar to STAR ELA scores, EL students (M = 364.07; SD = 84.58; p < .01; d = 0.69) had
significantly lower STAR mathematics scores compared to EO students (M = 425.98;
SD = 92.79). However, IFEP students (M = 431.06; SD = 80.89; p = .98; d = 0.06), and RFEP
students (M = 418.83; SD = 83.07; p = .93; d = 0.08) did not differ statistically from EO students.

7.2.2 | CAASPP scores

The CAASPP replaced the STAR in the later years of the DLI study. We examined 11th‐grade
CAASPP scores for Cohorts 1–4 of the DLI graduates (N = 156), since data for Cohort 5 was not
yet available.

7.2.2.3 | CAASPP English language arts

11th‐grade CAASPP ELA scores are shown in Figure 6. EL students (M = 2449.56;
SD = 94.34; p < .01; d = 1.41) again had significantly lower scores than EO students

F I G U R E 5 DLI students' STAR scores in mathematics. Mean scores and bootstrapped confidence intervals
are shown for students from each EL proficiency classification. DLI, Dual Language Immersion; EL, English
Learner; STAR, Standardized Testing and Reporting. [Color figure can be viewed at wileyonlinelibrary.com]
14 | PADILLA ET AL.

F I G U R E 6 DLI students' CAASPP scores in English language arts. Mean scores and bootstrapped
confidence intervals are shown for students from each EL proficiency classification. CAASPP, California
Assessment of Student Performance and Progress; DLI, Dual Language Immersion; EL, English Learner. [Color
figure can be viewed at wileyonlinelibrary.com]

F I G U R E 7 DLI students' CAASPP scores in math. Mean scores and bootstrapped confidence intervals are
shown for students from each EL proficiency classification. CAASPP, California Assessment of Student
Performance and Progress; DLI, Dual Language Immersion; EL, English Learner. [Color figure can be viewed at
wileyonlinelibrary.com]

(M = 2633.64; SD = 135.61). However, only 9 students remained as long‐term ELs by 11th

grade. There were no significant differences between EO students and IFEP students
(M = 2618.15; SD = 106.07; p = .93; d = 0.12) or RFEP students (M = 2595.97; SD = 89.1;
p = .22; d = 0.34).
PADILLA ET AL. | 15

7.2.2.4 | CAASPP mathematics

The CAASPP mathematics score results for DLI graduates are shown in Figure 7. The
achievement outcomes for mathematics were similar to the CAASPP ELA results. EL students'
(N = 9) scores (M = 2428.78; SD = 77.13; p < .01; d = 1.34) were significantly lower than those of
EO students (M = 2618.09; SD = 147.99), while IFEP students (M = 2620.73; SD = 104.66;
p = .99; d = 0.02) and RFEP students did not show significant differences (M = 2572.93;
SD = 100.13; p = .14; d = 0.37) compared to EO students.

7.2.3 | High school GPA

The end of 10th grade cumulative GPA of DLI graduates was examined by EL status, since 10th
grade was the highest grade for which we had data from all five cohorts. These results are
presented in Figure 8. EO students had the highest GPA on average (M = 3.14; SD = 0.66),
though they were closely followed by IFEP students (M = 3.03; SD = 0.74), whose mean GPA
was not significantly different from EOs (p = .89; d = 0.21). RFEP students had a lower mean
GPA (M = 2.81; SD = 0.81), an effect that was significant but small (p = .03; d = 0.23). EL
students also had a significantly lower average GPA (M = 1.89; SD = 0.95) than EO students
(p < .01; d = 0.21). However, there were only 14 DLI graduates classified as Els in 10th grade for
whom grades were available, so these results are likely driven by a relatively small group of
long‐term English Learners (LTEL).

F I G U R E 8 DLI students' average high school GPA. Mean GPA and bootstrapped confidence intervals are
shown for students from each EL proficiency classification. DLI, Dual Language Immersion; EL, English
Learner. [Color figure can be viewed at wileyonlinelibrary.com]
16 | PADILLA ET AL.

TABLE 7 Number of students who met UC/CSU approved course requirements by EL classification.
Total number
Number of students of students in Percent of students
EL status who met requirements category who met requirements
English Only (EO) 25 37 67.0%
Initial Fluent English 14 23 60.9%
Proficient (IFEP)
Reclassified Fluent English 33 55 60.0%
Proficient (RFEP)
All students 72 120 60.0%
Abbreviations: EL, English Learner; UC, University of California/California State University.

7.2.4 | University of California/California State University (UC/CSU)

approved courses

For the 120 DLI graduates for whom we have 12th grade data, Table 7 reports the number of
students who completed the UC/CSU approved course requirements from each English
proficiency classification group. The UC/CSU course requirements are significant because
completing these required courses qualifies a student for admission into a UC or CSU campus.
As shown in Table 7, approximately the same percent of students in the three language
categories qualified for admission to a California university based on required content classes.

7.2.5 | Research Question 2 summary

Across all the academic outcome measures examined—completion of UC/CSU requirements,

STAR and CAASPP scores, and high school GPA—DLI students demonstrated consistent
patterns in their results. The vast majority of Els in the DLI sample were reclassified by high
school, and once reclassified, there were no significant differences in any of the academic
achievement measures when compared to IFEP and EO students. However, a small number of
students who were still classified as long‐term Els (N = 9) performed at a lower level on average
than other DLI graduates.

7.3 | Comparison of DLI and non‐DLI students

Research Question 3 was: Do graduates of a K‐5 Spanish DLI program perform at the same levels
academically, as measured by high school GPA and STAR and CAASPP/SBAC test scores, as a
comparison group of students who did not attend a DLI program in elementary school?
The results reported for Research Questions 1 and 2 have demonstrated that there were no
significant differences in English Language Arts and Math achievement between DLI graduates
who were classified as EO, IFEP, or RFEP. Furthermore, DLI graduates' language trajectories
were not complete at the end of the program: large numbers of students continued to take
PADILLA ET AL. | 17

Spanish classes during the years afterwards and reached advanced levels of Spanish classes in
high school.
These findings are focused specifically within the sample of DLI graduates. However, in
Research Question 3, we also wished to investigate how DLI graduates' academic outcomes
would compare to students from the same district who did not enroll in a DLI program in
elementary school. This would shed light on whether completing the DLI program was
associated with higher or lower patterns of achievement among students.
To examine how the outcomes of DLI students compared to non‐DLI students, we created
three comparison groups of students and tracked their academic performance through middle
and high school. The first group consisted of 200 non‐DLI students classified as Els in 5th
grade, some of whom were reclassified in later grades. The second group consisted of 200 non‐
DLI EO students, and the third group was made up of 200 non‐DLI, non‐Hispanic EO students.
All three groups were sampled from the same district as the DLI students and attended 5th
grade during the same years as the DLI cohorts.
In addition, the groups were matched to the DLI students on several important
demographic characteristics. Table 8 presents the three comparison groups and the
demographic measures for which they were matched with DLI students. In the non‐DLI EL/
RFEP and non‐DLI EO samples, 80% of each comparison group was made up of students who
identified as Hispanic, with the remaining 20% of students identifying as another ethnicity.
This matched the ethnicity breakdown for the DLI students. The third comparison group was
composed entirely of students who did not identify as Hispanic, to explore how DLI students
would perform compared to a group that was not matched on ethnicity, but which was
controlled for socioeconomic status using free/reduced lunch data with about 30% of students
eligible for free or reduced lunch. However, on average, the parent with the highest education
level in the non‐DLI EL/RFEP group had a lower education level than the parent with the
highest education level in the other three groups. Twenty‐eight percent of non‐DLI EL/RFEP
students had a parent who attended college compared to 67% of non‐DLI EO students; for DLI
students 51% of their parents attended college; and 76% of non‐DLI, non‐Hispanic EO students
had a parent who attended college.
We compared DLI graduates with the three non‐DLI comparison groups using the same
academic outcome measures which were used to compare DLI students from different EL

TABLE 8 Non‐DLI student samples that were compared to DLI graduates

Group name
200 non‐DLI EL/RFEP students
200 non‐DLI EO students
200 non‐DLI, non‐Hispanic EO
students
Abbreviations: DLI, Dual Language Immersion; EO, English Only; RFEP, Reclassified Fluent English Proficient.
Characteristics matched to DLI student sample
• Same school district and grade levels
• Ethnicity; 80% of students identified as Hispanic
• Socioeconomic status: 30% of students were eligible for free or
reduced lunch
• Same school district and grade levels
• Ethnicity; 80% of students identified as Hispanic
• Socioeconomic status: 30% of students were eligible for free or
reduced lunch
• Same school district and grade levels
• Socioeconomic status: 30% of students were eligible for free or
reduced lunch
18 | PADILLA ET AL.

F I G U R E 9 DLI and non‐DLI students' average STAR ELA scores. Mean scores and bootstrapped confidence
intervals are shown for DLI students by EL proficiency classification, and for the three non‐DLI comparison
groups. DLI, Dual Language Immersion; EL, English Learner; ELA, English Language Art; STAR, Standardized
Testing and Reporting. [Color figure can be viewed at wileyonlinelibrary.com]

classification groups: middle school STAR scores, 11th grade CAASPP scores, and high school
cumulative GPA.

7.3.1 | STAR ELA scores

Figure 9 illustrates that DLI students as a whole also had significantly higher STAR ELA scores
than non‐DLI Els/RFEPs (p < .01; d = 1.05). This held true specifically for DLI EOs, IFEPs, and
RFEPs (all p < .01; d between 1.08 and 1.25), while there was not a significant difference
between the scores of DLI ELs and non‐DLI ELs/RFEPs (p = .35; d = 0.25).
Overall, DLI students' scores were higher than those of non‐DLI EOs, but the difference was
not statistically significant (p = .09; d = 0.19). Further inspection revealed that DLI EOs
(p < .01; d = 0.46) and IFEPs (p < .01; d = 0.45) had significantly higher scores than non‐DLI
EOs, but DLI RFEPs did not differ from non‐DLI EOs on ELA scores (p = .26; d = 0.26).
There were no significant differences between the scores of DLI EOs, IFEPs, or RFEPs
compared to those of non‐DLI, non‐Hispanic EOs (all p > .75; d between 0.05 and 0.16).

7.3.2 | STAR mathematics scores

DLI students had consistently higher STAR mathematics scores than both non‐DLI ELs and
non‐DLI EOs, as shown in Figure 10. Except for ELs (p = .99; d = 0.08), DLI students scored
higher than non‐DLI EO students regardless of EL status (all p < .01; d between 0.55 and 0.69).
PADILLA ET AL. | 19

F I G U R E 10 DLI and non‐DLI students' average STAR mathematics scores. Mean scores and bootstrapped
confidence intervals are shown for DLI students by EL proficiency classification, and for the three non‐DLI
comparison groups. DLI, Dual Language Immersion; EL, English Language; STAR, Standardized Testing and
Reporting. [Color figure can be viewed at wileyonlinelibrary.com]

There was not a significant difference between the overall scores of DLI students and those
of non‐DLI, non‐Hispanic EO students. However, there were differences when we examined
DLI students by EL status. Compared to non‐DLI, non‐Hispanic EO students, DLI IFEPs had
significantly higher scores (p = .047; d = 0.37), and DLI EOs had scores that were marginally
higher, but which fell short of statistical significance (p = .061; d = 0.3). Finally, DLI RFEPs did
not differ significantly in mathematics scores from the non‐DLI, non‐Hispanic EO students
(p = .32; d = 0.23).

7.3.3 | CAASPP ELA scores

Similarly, as shown in Figure 11, DLI students had significantly higher CAASPP ELA scores
than non‐DLI ELs (p < .01; d = 0.83). This held true for students classified as EOs, IFEPs, and
RFEPs. There was no significant difference between the scores of DLI students and non‐DLI
EOs (p = .88; d = 0.08). Non‐DLI, non‐Hispanic EOs had scores that were somewhat higher, but
not significantly higher, than DLI students (p = .066; d = 0.28). These patterns held true across
all groups of DLI students, except students classified as long‐term ELs, whose scores were lower
than the two EO comparison groups.

7.3.4 | CAASPP mathematics scores

As shown in Figure 12, DLI students had significantly higher 11th‐grade CAASPP mathematics
scores than non‐DLI ELs/RFEPs (p < .01; d = 0.86). DLI students also had higher scores on
20 | PADILLA ET AL.

F I G U R E 11 DLI and non‐DLI students' average CAASPP ELA scores. Mean scores and bootstrapped
confidence intervals are shown for DLI students by EL proficiency classification, and for the three non‐DLI
comparison groups. CAASPP, California Assessment of Student Performance and Progress; DLI, Dual Language
Immersion; EL, English Learner; ELA, English Language Art. [Color figure can be viewed at
wileyonlinelibrary.com]

F I G U R E 12 DLI and non‐DLI students' average CAASPP math scores. Mean scores and bootstrapped
confidence intervals are shown for DLI students by EL proficiency classification, and for the three non‐DLI
comparison groups. CAASPP, California Assessment of Student Performance and Progress; DLI, Dual Language
Immersion; EL, English Learner. [Color figure can be viewed at wileyonlinelibrary.com]
PADILLA ET AL. | 21

average than non‐DLI EOs, though this difference did not reach significance (p = .12; d = 0.25).
Non‐DLI, non‐Hispanic EOs had significantly higher scores than DLI students
(p < .01; d = 0.37).
All DLI students except ELs had significantly higher scores than non‐DLI EL students
(all p < .01; d between 0.84 and 1.29). There was no significant difference in the scores of
non‐DLI EOs and DLI IFEPs and RFEPs; however, DLI EOs had higher scores than non‐DLI
EOs (p = .021; d = 0.46).
Also, non‐DLI, non‐Hispanic EOs' scores were not significantly different from those of DLI
EOs and IFEPs. Their scores were, however, significantly higher than those of DLI RFEPs
specifically (p < .01; d = 0.52).

7.3.5 | Cumulative 10th grade high school GPA

DLI students had significantly higher cumulative 10th grade GPAs than non‐DLI ELs (p < .01;
d = 0.57). There was not a significant difference between DLI students' GPAs and non‐DLI EO
students' GPAs (p = .65; d = 0.32). However, non‐DLI, non‐Hispanic EOs had significantly
higher GPAs than DLI students (p < .01; d = 0.39). These results are presented in Figure 13.
Furthermore, we can examine differences in cumulative high school GPA by EL status.
Most DLI students who started as ELs in 5th grade were reclassified, so we will focus on DLI
RFEPs compared to the 5th grade non‐DLI EL comparison group (many of whom were also
later reclassified). There was no significant difference between the GPA of DLI RFEPs and the
non‐DLI EO comparison group. The same pattern held for DLI, though DLI EOs had an
average GPA that was marginally significantly higher than non‐DLI EOs (p = .053; d = 0.42).

F I G U R E 13 DLI and non‐DLI students' average cumulative high school GPA. Mean GPA and bootstrapped
confidence intervals are shown for DLI students by EL proficiency classification, and for the three non‐DLI
comparison groups. DLI, Dual Language Immersion; EL, English Language [Color figure can be viewed at
wileyonlinelibrary.com]
22 | PADILLA ET AL.

There was no significant difference between the GPAs of non‐DLI, non‐Hispanic students and
those of DLI EOs or IFEPs, though non‐DLI, non‐Hispanic students had significantly higher
GPAs than DLI RFEPs (p < .01; d = 0.49).

7.3.6 | Research Question 3 summary

Across academic achievement outcomes, DLI students performed at higher levels on average
than the non‐DLI EL/RFEP comparison group. Furthermore, DLI graduates had similar or
higher levels of achievement compared to the non‐DLI comparison group that was matched for
ethnicity, as well as compared to the non‐DLI, non‐Hispanic comparison group for all measures
except cumulative high school GPA. These patterns held for DLI students regardless of EL
status, except for the small number of longer‐term ELs who tended to perform lower than
comparison groups.

8 | DISCUS SION

In this study, we examined the academic and language trajectories of students who had been
enrolled in a Spanish K–5 DLI program. We present summaries of our findings and discussion
in relation to each of our three major research questions.
Regarding the first research question, we sought to examine whether graduates of the
elementary Spanish DLI program would continue with advanced Spanish language classes in
secondary school and to explore their language outcomes. Our results suggested the persistence
of second language learning of students who graduated from the DLI program. Elementary
immersion programs begin early and offer the opportunity for a long sequence of study in the
second language. Both of these are critical elements of becoming bilingual and biliterate. Most
of the students in this study took advantage of continuing Spanish language study. About 90%
of the students for whom we have data participated in a partial Spanish immersion program in
middle school. 80% of those earned an A or B in the immersion classes. Many of the students
continued to study higher levels of Spanish language and/or native speaker classes in high
school. Over 60% took Spanish or Spanish for Native Speakers as upper division classes (level 3
or 4), and over 60% took either Advanced Placement Spanish or advanced IB Spanish in high
school. It was rare for students to take classes in languages other than Spanish, with only about
10% doing so. These results are aligned with the literature. For example, Mellgren and Somers
(2008) conducted a study by surveying the graduates of a K–5 Spanish language immersion
program. They found that many former immersion students took advanced Spanish classes
during high school, and most of the graduates reached high levels of fluency in reading,
listening, speaking and writing in Spanish.
Importantly, regarding the language outcomes, our findings show that 42% of DLI graduates
earned the California Seal of Biliteracy by their high school graduation, a rate significantly
higher than any comparison groups. Furthermore, statewide, only 11% of all high school
graduates in 2019 earned the California Seal of Biliteracy (Californians Together, 2019). This
supports the observation that many DLI students continue to exhibit proficiency in both
English and Spanish through high school. When students develop high levels of language and
cultural competence, they are gaining skills which help them become globally competent
citizens. The California World Languages Framework states that “creating and sustaining long
PADILLA ET AL. | 23

sequences of language learning pathways provides opportunities for students to develop global
competence and a functional range of proficiency for the workplace” (California Department of
Education, 2020). The findings in this study demonstrate that DLI students' language
trajectories often continue all the way through middle and high school, frequently concluding
with students enrolling in advanced classes in the minority language. Attaining the California
Seal of Biliteracy provides support for the claim that DLI students can reach high levels of
bilingual proficiency.
Regarding the second research question, we first sought to understand whether EL learners
who entered a Spanish immersion program in kindergarten differed from EO peers on state
mandated achievement tests and other academic achievement indices (i.e., STAR scores for 5th
to 8th grade, CAASPP/SBAC scores for 11th grade, GPA from 9th to 12th grade) as all students
progress through school. Our findings revealed that almost all graduates of the DLI program
who were classified as ELs in kindergarten are assessed and reclassified as Reclassified Fluent
English proficient (RFEP) before entering high school. Further, there were not significant
differences between the scores of RFEPs, IFEPs, and EOs on the achievement measures
weexamined, except for high school GPA in which RFEP students' scores were lower than EO
students' scores (though this was not a large difference). In other words, this finding
demonstrated that most students who begin a DLI program as ELs are reclassified as English
proficient before entering high school. Once they are reclassified, they perform similarly to EOs
on state mandated achievement tests in English Language Arts and Mathematics.
By program design, in kindergarten DLI students are roughly split into 50% students who
speak the minority language (Spanish) at home, many of whom are ELs, and 50% EO students.
In our sample, the number of students classified as EL dropped each year. By 5th grade the
percentage of students who were still designated as EL was down to 29% and by 10th grade the
percentage was a mere 7%. The few remaining long‐term ELs (LTELs) had significantly lower
outcomes across all academic achievement measures. While the number of LTEL students was
small—only 16 students out of 5 cohorts of DLI graduates (N = 225)—by the time they reached
high school, an examination of student records indicates that many of these students had
additional support needs that prevented them from being reclassified, such as needing special
education services.
Of particular importance, the data show that RFEP students generally performed at a level
similar to EO and IFEP students across all academic measures. In other words, students who
were classified as ELs in kindergarten, but who were later reclassified, performed similarly to
students who started the DLI program speaking only English or who had a non‐English home
language, but who were classified at school entrance as IFEP.
Most of the achievement outcome research on DLI programs has been limited to the
elementary level (Lindholm‐Leary & Block, 2010; Marian et al., 2013; Padilla et al., 2013) and
only a few studies have examined student outcomes through the middle school years
(Cobb et al., 2006; Steele et al., 2017). This study, however, is one of the first to examine the
academic achievement of DLI program graduates all the way through high school completion.
These findings are significant because they demonstrate that DLI programs are associated with
high levels of academic achievement for both EO students and ELs, and may contribute to
preparing students for college and careers where they may use their bilingual skills.
Regarding the third research question, we sought to understand whether graduates of DLI
programs performed at the same levels academically as those students who did not attend a DLI
program in elementary school. On average, in comparing graduates of the DLI program at the
end of the 5th grade with a comparison group of students who were ELs in kindergarten, but
24 | PADILLA ET AL.

who did not attend a DLI program, the data indicate that DLI graduates matched the
achievement levels of the EO comparison group across outcome measures. Graduates of the
DLI program also had consistently higher middle school STAR scores, high school CAASPP
scores, and high school GPAs than a comparison group of students who were ELs in 5th grade.
Our findings in this study corroborated previous research that asserts that graduates of DLI
programs perform at the same level or surpass peers on academic measures who do not attend
immersion programs in elementary school (Lindholm‐Leary & Howard et al., 2008; Thomas &
Collier, 2002), but importantly our findings extended the results beyond elementary school
through to high school graduation. Previous research has largely discussed the beneficial
outcomes of two‐way immersion programs at the elementary level. For example, Marian et al.
(2013) found that two‐way immersion education is beneficial for both minority language and
majority language elementary students, as students from the DLI program outperformed peers
in reading and mathematics by the completion of elementary school. Some studies examine
students beyond the elementary level, but not over the course of multiple school years. For
example, Cobb et al. (2006) conducted a study that examined the effects of a two‐way
elementary school immersion program on later 6th and 7th grade junior high school
achievement. The findings from this study showed that the greatest effect for native EO
students was on Reading achievement, while for native Spanish speakers the largest effect was
on Writing and Mathematics achievement. However, no research studies could be identified
that compared students' academic achievement from elementary through the 12th grade for
DLI and non‐DLI students.
Looking at DLI students by language classification, all students except the few long‐term
ELs performed at the same level or higher than the comparison groups across all achievement
measures. Once again, this indicates that when former EL students of a DLI program are
reclassified, they tend to have higher achievement outcomes than students who were ELs in 5th
grade, but who were not enrolled in a DLI program. These results extend previous findings by
Steele et al. (2017) whose study ended with 8th grade outcomes, while the current study
followed former DLI students through high school graduation. These results also support the
general belief educators hold that being reclassified from EL to fluent English proficient is one
of the most vital milestones in an EL's academic life (Umansky & Reardon, 2014). In sum, DLI
students generally achieved at equivalent or higher levels than the non‐DLI comparison groups.

9 | LIMITATION S A N D FU TU RE RE S E A RC H

There are certain limitations of this study. First, this study explored existing longitudinal
student data in a specific area in California; thus, the findings may not be widely generalizable.
Future research is encouraged to explore DLI program results in other areas of the United
States. Second, the study presented here focused on a single Spanish DLI program; future
research is encouraged to compare DLI programs with a variety of minority languages
(e.g., Mandarin or French). Third, future research should investigate larger samples of DLI
graduates to determine what demographic factors may contribute to academic success. Finally,
the comparison groups of students at the secondary level often attended different elementary
schools in the same school district, therefore confounding variables may have been introduced
in the study. Accordingly, future research could aim to eliminate possible confounding
variables by studying DLI and non‐DLI students who attended the same school where DLI and
non‐DLI strands co‐exist (e.g., Padilla et al., 2013).
PADILLA ET AL. | 25

In spite of these possible limitations in the study, the academic achievement as observed on
multiple measures of DLI students is remarkable. Students graduating from elementary DLI
programs perform at equivalent or higher levels across academic content measures in English
Language Art and Mathematics compared to non‐DLI students.

10 | C ONC LUSION AND I MPLICATIONS

In conclusion, this remains one of the few studies to examine DLI students' academic
achievement through high school graduation. In this DLI program, the majority of ELs were
reclassified as fully English proficient by 5th grade, and all but a few by 10th grade.
Importantly, reclassified ELs performed at the same level as EO students on nearly all
academic measures. In general, DLI graduates performed at equivalent or higher levels than
non‐DLI students. Finally, of particular significance is that DLI programs promote high
levels of biliteracy. Most DLI students continue at advanced levels of the minority language
in high school, and nearly half went on to earn the Seal of Biliteracy.
The findings in this study suggest that elementary students who begin in a DLI program
in elementary school and continue with language instruction through high school attain
high levels of language proficiency and academic achievement. Recommendations for
education practitioners and researchers to consider are: expand the implementation of
robust DLI programs into other elementary schools and languages; recruit and offer
research‐based professional learning for DLI teachers to support the growth in DLI
programs; plan and execute DLI program pathways from elementary through high school to
maximize dual language proficiency; measure the outcomes of second language proficiency
at key transitions (e.g., 5th and 8th grades) in K–8/12 programs; and identify and use
assessment instruments that measure content knowledge (e.g., math and science) in the
languages of the DLI programs.

ACKNOWLEDGMENTS
The completion of this project could not have been possible without the collaboration and
assistance of administrators from the Redwood City School District, Sequoia Union High
School District and Stanford University. Their contributions are sincerely appreciated and
gratefully acknowledged. We would especially like to express appreciation to Superintend-
ent John Baker, Assistant Superintendent Linda Montes and Assistant Superintendent
Bonnie Hansen. Appreciation is also extended to the Stanford University School of
Education for funding this study with a grant through its initiative for research projects
with Sequoia K‐12 Research Collaborative and to Michelle Nayfack, Executive Director of
the Collaborative, for her guidance in navigating two school districts to complete this
project.

ORCID
Xinjie Chen http://orcid.org/0000-0002-5492-225X
Elizabeth Swanson https://orcid.org/0000-0001-8004-4289

REFERENCES
Baker, S. C., & MacIntyre, P. D. (2000). The role of gender and immersion in communication and second
language orientations. Language Learning, 50(2), 311–341.
26 | PADILLA ET AL.

Burkhauser, S., Steele, J., Li, J., Slater, R., Bacon, M., & Miller, T. (2016). Partner‐language learning trajectories
in dual‐language immersion: Evidence from an urban district. Foreign Language Annals, 49(3), 415–433.
https://doi.org/10.1111/flan.12218
California Department of Education. (2020). World languages framework for public schools, kindergarten through
grade twelve. https://www.cde.ca.gov/ci/fl/cf/
Californians Together. (2019). Who are the recipients of the seal of biliteracy? [Infographic]. https://www.
californianstogether.org/product/the-seal-of-biliteracy-infographic/
Christian, D., Howard, E. R., & Loeb, M. I. (2000). Bilingualism for all: Two‐way immersion education in the
United States. Theory into Practice, 39(4), 258–266. https://doi.org/10.1207/s15430421tip3904_9
Cobb, B., Vega, D., & Kronauge, C. (2006). Effects of an elementary dual language immersion school program on
junior high school achievement. Middle Grades Research Journal, 1, 27–47.
Devlin, K. (2015). Learning a foreign language a “must” in Europe, not so in America. Fact Tank. https://www.
pewresearch.org/fact-tank/2015/07/13/learning-a-foreign-language-a-musin-europe-not-so-in-america/
Howard, E. R., Lindholm‐Leary, K. J., Rogers, D., Olague, N., Medina, J., Kennedy, B., Sugarman, J., &
Christian, D. (2018). Guiding principles for dual language education (3rd ed.). Center for Applied
Linguistics.
Lindholm‐Leary, K. (2012). Success and challenges in dual language education. Theory into Practice, 51,
256–262. https://doi.org/10.1080/00405841.2012.726053
Lindholm‐Leary, K., & Block, N. (2010). Achievement in predominantly low SES/Hispanic dual language
schools. International Journal of Bilingual Education and Bilingualism, 13(1), 43–60. https://doi.org/10.
1080/13670050902777546
Lindholm‐Leary, K., & Borsato, G. (2005). Hispanic high schoolers and mathematics: Follow‐up of students who
had participated in two‐way bilingual elementary programs. Bilingual Research Journal, 29(3), 641–652.
https://doi.org/10.1080/15235882.2005.10162856
Marian, V., Shook, A., & Schroeder, S. (2013). Bilingual two‐way immersion programs benefit academic
achievement. Bilingual Research Journal, 36(2), 167–186. https://doi.org/10.1080/15235882.2013.818075
Mellgren, M., & Somers, E. (2008). Graduates of a language immersion program: What are they doing now? ACIE
Newsletter, The Bridge: From Research to Practice. https://citeseerx.ist.psu.edu/viewdoc/download?doi=
10.1.1.739.1520&rep=rep1&type=pdf
Padilla, A. M., Fan, L., Xu, X., & Silva, D. (2013). A Mandarin/English two‐way immersion program: Language
proficiency and academic achievement. Foreign Language Annals, 46, 661–679. https://doi.org/10.1111/
flan.12060
Plonsky, L., & Oswald, F. L. (2014). How big is “Big”? Interpreting effect sizes in L2 research. Language
Learning, 64(4), 878–912. https://doi.org/10.1111/lang.12079
Robins, C. (2017). The impact of dual language programs on Latino high school students (Order No. 10284407)
[Doctoral dissertation, Northwest Nazarene University]. ProQuest Dissertations Publishing. https://whdl.
org/sites/default/files/Chelle%20Robins%20FINAL%20DISSERTATION.pdf
Rocque, R., Ferrin, S., Hite, J. M., & Randall, V. (2016). The unique skills and traits of principals in one‐way and
two‐way dual immersion schools. Foreign Language Annals, 49(4), 801–818.
Steele, J., Slater, R., Zamarro, G., Miller, T., Li, J., Burkhauser, S., & Bacon, M. (2017). Effects of dual‐language
immersion programs on student achievement: Evidence from lottery data. American Educational Research
Journal, 54(1), 282–306. https://doi.org/10.3102/0002831216634463
Steele, J., Walzinger‐Tharp, J., Slater, R., Roberts, G., & Bowman, K. (2019). Research Brief: Student performance
under dual language immersion scale‐up in Utah. American Councils for International Education. Retrieved
March 23, 2021, from http://americancouncils.org/research-assessment/studies-and-reports
Thomas, W. P., & Collier, V. P. (2002). A national study of school effectiveness for language minority students'
long‐term academic achievement. Center for Research on Education, Diversity, and Excellence, University
of California at Santa Cruz.
Umansky, I. M., & Reardon, S. F. (2014). Reclassification patterns among Latino English learner students in
bilingual, dual immersion, and English immersion classrooms. American Educational Research Journal,
51(5), 879–912.
R Core Team. (2021). R: A language and environment for statistical computing. https://www.R-project.org/
PADILLA ET AL. | 27

How to cite this article: Padilla, A. M., Chen, X., Swanson, E., Peterson, M., &
Peruzzaro, A. (2022). Longitudinal study of Spanish Dual Language Immersion
graduates: Secondary school academic and language achievement. Foreign Language
Annals, 1–27. https://doi.org/10.1111/flan.12615

View publication stats