IMPACT OF EARLY SECOND LANGUAGE LEARNING ON FIRST

Report
IMPACT OF EARLY SECOND
LANGUAGE LEARNING ON
FIRST LANGUAGE
A Study of the
Confucius Institute in Edmonton
In Partnership with
Edmonton Public Schools
June 2009
Researchers
Study Sponsored by
The Confucius Institute in Edmonton
 Primary Researcher:


Dr. John Macnab:


Research Support Services – Edmonton Public
Schools
Editor:

Stuart Wachowicz

Chairman: Confucius Institute in Edmonton
Purpose of the Study

To examine the reciprocal effect of second
language study from an early age on first
language

Hypothesis:

That students participating in intensive second
language training from an early age will have
comparable or superior performance in
formalized English language studies than
students who are in unilingual programs.
Rationale for the Study:

To grow demand for second language education
in school systems (especially in North America),
and in particular to grow enrollment in Chinese
language, a number of sincerely held, but
possibly erroneous beliefs need to be addressed:




L2 should wait until L1 is firmly established
Early L2 can damage L1 acquisition
L2 students struggling in L1 should be withdrawn
from L2
L2 learning is difficult, therefore limit to strong
students
Rationale for the Study:
Such beliefs contribute to an educational
culture in which L2 is not valued as other
core subjects.
 Hence promotion of the idea that all
students should gain a measurable
proficiency in L2 is a challenge.
 Subsequently promoting Chinese is even
more difficult given the idea that Asian
languages are harder to learn for western
students

Basis for the research
Edmonton Public Schools (EPS) has a 30
year history of intensive L2 education, and
is the Canadian pioneer in “Bilingual”
education. (80 000 students)
 EPS has carefully collected years of
performance data of students on Alberta
government exams at grades 3, 6, 9 and
12.
 The data was segregated for students in
Immersion and Bilingual programs

EPS Language Programs
Program
Age and grade
Time in L2
French Immersion K-2
(age 5-17)
3-6
7-9
10-12
100%
85%
70%
40%
Bilingual (age 517)
50%
30-35%
15%
K-6
7-9
10-12
Second Language 4-9
Courses
10-12
10%
12.5%
Languages Offered
Immersion
Bilingual
Second
Language
French (3400)
Chinese (2000)
Arabic
(1000)
German ( 800)
Spanish (300)
Ukrainian (300)
Hebrew
(150)
ASL
(100)
French (26 000)
Spanish (3 000)
Chinese (1 000)
German (1 000)
Japanese (800)
Punjabi (200)
ASL
(150)
Ukrainian (150)
Cree
(300)
Research Design

Local conditions and time issues ruled out a “controlled”
experiment






Open boundaries
Programs of Choice
Student mobility
Access was available to many years worth of data of
student performance on provincially standardized tests in
English and core subjects. These test are written by all
students in the province at grades 3, 6, 9 and 12
English tests are of high quality assessing all strands of
Language Arts
Given the purpose, the research design was limited to the
English results.
Research Design

Based on student performance on tests the
provinces sets cut scores to determine students
who achieve:





Below Acceptable Standard
Within Acceptable Standard
Above Acceptable Standard
Grade 3 English results were used as “pre-test”
Grade 6 and 9 results were used as “post-test”
scores for analysis
Research Design:
Controlling Confounding Variables
Provincial Achievement Tests represent
the total population of Alberta students,
therefore true population parameters are
available (normal population) =Control
 Only students remaining in the L2
program from k to 9 were included
 Individual student results were converted
into Z scores: Z=(X-μ)/σ This permits
tracking of growth, based upon what
would be expected progress for that
student.

Research Design:

Data analyzed in two ways:



Statistical tests analyzed in an R computing environment
(statistical computing and graphics language interface)
Graphical representation
Results not broken down by specific language:



Some language programs are too small
The study considered instruction in L2 in an English
environment the common treatment
Programs varied slightly in content and approach in
different schools, making the total cohort better
approximate a random sample than if we focused on
smaller groups, yielding greater generalization.
Research Design: The Student Sample

Cohort 1:


Cohort 2:


304 students who began their schooling in
September 1997
286 students who began their schooling in
September 1998
Total sample size 590
Cohort Composition
Cohort 1
Cohort 2
Total
Chinese
Bilingual
126
122
248
French
Immersion
128
101
229
German
Bilingual
35
41
76
Ukrainian
Bilingual
9
10
19
Arabic Bilingual
6
9
15
Hebrew
Bilingual
0
3
3
TOTAL
590
Rationale for Confucius Institute
Sponsorship




The study has raised interest across Canada in
the media, and academic and education circles,
and will play a key role in the language education
debate in Canada
The largest population in the cohort are students
learning Chinese
A study that could demonstrate a positive
correlation between Chinese language learning
and increased proficiency in L1 would be useful in
promoting Chinese.
National coverage in the national media would
profile the Confucius Institute as making a
valuable contribution to language learning in
Canada and beyond
Results:
Course
Standard Provincial
Results
%
Results for
Bilingual or
Immersion
students %
Difference
from
Province
ELA 3
Below
10
5
-5
ELA 3
Acceptable
72
72
0
ELA 3
Excellence
18
23
+5
ELA 6
Acceptable
12
4
-8
ELA 6
Below
69
70
+1
ELA 6
Excellence
19
26
+7
ELA 9
Below
12
2
-10
ELA 9
Acceptable
72
62
-10
ELA 9
Excellence
16
36
+20
Change in English Performance
70
Language Arts Achievement Levels
Standard
50
40
30
20
10
0
Percent of Students at Achievement Level
60
Acceptable
Below
Excellence
LA3
LA6
LA9
Provincial Achievement Test
Tracking Z Scores
Z>0
Z = 0
Z < 0


Above Provincial Average
At Provincial Average
Below Provincial Average
Result indicates that even with LESS
instruction in English, Immersion and
Bilingual students increased in English
proficiency relative to the total population.
15
10
Percent of Students
10
0
5
5
0
Percent of Students
15
20
20
25
Grade 3 vs 9 Z Score Distribution
-3
-2
-1
0
1
Grade 3 English Language Arts z-score
2
3
-3
-2
-1
0
1
Grade 9 English Language Arts z-score
2
3
Observations

In grade 3 the student population is only
slightly above the normal distribution for
the province. (This after 3 years of being
in the intense L2 environment);


They still have a broad distribution of results
indicating they are not an elite group.
Some students are clearly struggling in English
at -2 and -3 Standard Deviations below the
mean.
Observations
By grade 9 this group has clearly
advanced significantly beyond the mean.
 The top end students in grade 3 have not
necessarily advanced, yet they are still in
the “excellence” category.
 It is the low end student in grade 3, the
ones who struggled in English, who have
made the most gains

Effect Size
This effect size indicates that those students who
remained in their bilingual or immersion
programs show statistically significantly greater
growth in English Language Arts than would
have been expected if they had tracked the
control (provincial census) group.
 In other words their English was better than
if they had been in an English only program
Confucius Institute in Edmonton

Although these students spend
significantly less time in English
instruction, their English results have
improved at a rate that is statistically
faster than English only program
students.
Confucius Institute in Edmonton
Current Conclusion




If we desire the strongest results in English for
the individual student, provide, early and
sustained, quality second language education.
The effect size indicates this has greater impact
than any other researched intervention.
The result will be a student with strong English
results, with bilingualism as a by product.
If you want strong English skills provide
sustained second language instruction, in
particular Mandarin.
Confucius Institute in Edmonton
Contact Information
The Confucius Institute in Edmonton
13750 Woodcroft Ave.
Edmonton, Alberta T5M 3M4
Website: www.confuciusedmonton.ca
Phone: (780) 970-5233
IMPACT OF EARLY SECOND
LANGUAGE LEARNING ON
FIRST LANGUAGE
A Study of the
Confucius Institute in Edmonton
In Partnership with
Edmonton Public Schools

similar documents