Technical ReportPDF Available

Understanding the student with interrupted formal education (SIFE)

Authors:
Understanding the student with interrupted formal education (SIFE):
A study of SIFE skills, needs and achievement
Elaine C. Klein and Gita Martohardjono
Research Institute for the Study of Language in Urban Society
(RISLUS)
Graduate Center of the City of New York
Report to the NYC DOE Office of English Language Learners
DOE Report: Phase II: 2008-2009
July 2009
!"#$%&'()*+%,-./'%""%
%
%
0%
1. Introduction
1.1 Rationale for the SIFE Project, Phase II
1.2 Objectives of Phase II
1.3 Overview of the Report
2. Methodology
2.1 Participating Schools
2.1.1 Overview of High School A
2.1.2 Overview of High School B
2.1.3 Overview of High School C
2.1.4 Overview of High School D
2.1.5 Overview of High School E
2.2 Participants
2.2.1 Target Group: SIFE
2.2.2 High School Comparison Groups
Native English Speaker Group (NES)
English Language Learner Group (ELL)
2.3 Materials and Procedures
2.3.1 Intake Questionnaire
2.3.2 RISLUS Syntax Test (Spanish and English)
2.3.3 Oral Language Proficiency Test (Versant) (Spanish and English)
2.3.4 Working Memory Task
2.3.5 Academic Language and Literacy Diagnostic (ALLD) (Spanish and English)
2.3.5.1 ALLD Sections
2.3.6 Classroom Observation Checklist
2.3.7 Exit Questionnaire
3. Results and Discussion: Year One (2006-2007)
3.1 Participant Attrition
3.2 Intake Questionnaire
3.3 Spanish Syntax Test
!"#$%&'()*+%,-./'%""%
%
%
1%
3.4 Spanish Oral Language Proficiency Test (Versant)
3.5 Working Memory Task
3.6 Spanish ALLD Year One
3.7 Comparison of Abilities in Native Language: SIFE (Spanish ALLD) Compared to
High School Native English Speakers (English ALLD)
4. Results and Discussion: Year Two (2007-2008)
4.1 Spanish ALLD Year Two Compared to Spanish ALLD Year One
4.2 English Diagnostics
4.2.1 English Syntax Test
4.2.2 English Oral Language Proficiency Test (Versant)
4.2.3 English ALLD
SIFE (English ALLD) Compared to High School ELLs (English ALLD)
4.3 Classroom Observations
4.4 Exit Questionnaire
5. Conclusions and Recommendations
!"#$%&'()*+%,-./'%""%
%
%
2%
1. Introduction
1.1 Rationale for the SIFE Project, Phase II
Students with Interrupted Formal Education (SIFE) are a growing population in New York
City schools. Unfortunately, however, critical and necessary research for this unique and
challenging population of English language learners (ELLs) has only just begun. To date, very
few researchers or practitioners have conducted in-depth investigations of SIFE to help identify
and characterize their skills and needs. Such studies are needed to help develop research-based
instructional programs for these students.
According to the 2006-2007 Bilingual Education Student Information Survey (BESIS) from
the New York City Department of Education Office of English Language Learners, there are
about fifteen thousand SIFE in NYC schools, which make up 11% of ELLs. Most SIFE enter the
NYC school system between the eighth and tenth grades and 59% speak Spanish as their native
language. The primary criterion for SIFE identification is at least a two year gap in prior
schooling; in addition, it is thought that most SIFE enter NYC schools with little or no native
language literacy, as well as low levels of English language proficiency and academic content
knowledge.
In our pilot study, Phase I of the SIFE Project, we confirmed these assumptions among a
small group of newly arrived ninth grade SIFE (N=12), all of whom spoke Spanish as their
native language. We found that among the twelve SIFE, average reading comprehension in
Spanish was at the third grade level. The content knowledge scores of the twelve SIFE were also
very low; they scored between the third and fourth grade levels on their math skills and between
the first and second grade levels on their science and social science skills, all in the native
!"#$%&'()*+%,-./'%""%
%
%
3%
language. Thus, we concluded that unlike regular ELLs, in addition to mastery of English, SIFE
need foundational skills in their native language in order to raise their scores to the appropriate
grade level.
In Phase II of the SIFE Project, our goal was to support the findings from Phase I with a
larger number of SIFE from a variety of NYC high schools using a SIFE diagnostic instrument
that we had developed and tested during the course of the study. With this diagnostic, we
planned to track the progress of SIFE over eighteen months, investigating the academic and
social characteristics of SIFE, in addition to how well they acquired language and literacy
knowledge after being in NYC schools for one year.
1.2 Objectives of Phase II
i. Further develop and finalize a set of diagnostics to assess the skills and competencies of
SIFE, both in the native language (Spanish) and English, for future use among SIFE
entering NYC high schools;
ii. Use the diagnostics as pre-tests to characterize the backgrounds and academic skills of
incoming SIFE;
iii. Track SIFE as they progress through high school by administering diagnostics at selected
intervals;
iv. Evaluate the support provided by different types of SIFE programs and instructional
approaches by comparing the educational outcomes of SIFE under different instructional
conditions using the diagnostics as post-tests after eighteen months in school;
v. Compare SIFE to their peers:
!"#$%&'()*+%,-./'%""%
%
%
4%
High school native English speakers and ELLs. These high school comparison groups
will allow us to 1) see how SIFE differ in native language skills from native English
speakers in NYC high schools, and to 2) determine whether or not, and to what
extent, SIFE and ELLs differ in their acquisition of English after being in NYC high
schools for one year.
vi. Make recommendations for programmatic and instructional purposes.
1.3 Overview of the Report
In Section Two, we describe the methodology used in this research, including the
participating schools, the participants, the procedures, and the materials we developed and
selected. In Section Three, we describe the results and findings of our study. Section Four
provides some discussion and concludes the report with our recommendations.
2. Methodology
2.1 Participating Schools
Five New York City high schools were chosen to participate in this study by the New York
City Department of Education. The schools are geographically distributed across the city, in four
of the five boroughs: one school is located in Manhattan, one in Brooklyn, one in Queens, and
two in the Bronx. Table 1 describes the features of the five schools.
Table 1
%
!5-))6%7%
!5-))6%0%
!5-))6%1%%
!5-))6%2%
!5-))6%3%
8)5.+9):%
;*):<%
;*))=6>:%
?.:-.++.:%
;*):<%
@A'':/%
!-'6+'*'BCD):E
!-'6+'*'B%!"#$%
D):E!-'6+'*'B%
D):E!-'6+'*'B%
!-'6+'*'B%
D):E/-'6+'*'B%
!-'6+'*'B%
!9F'%)G%/5-))6%
223%%/+AB':+/%
13HH%/+AB':+/%
04HH%
/+AB':+/%
1HH%/+AB':+/%
17I7%
/+AB':+/%
!"#$%&'()*+%,-./'%""%
%
%
J%
K%)G%$88/%
7HHK%
03K%
0JK%
7HHK%
01K%
K%)G%!"#$%
DL%
7K%%
7K%%
0IMIK%%
7K%%
,*)N*.O%+>('%
$:N69/-E):6>%
9OO'*/9):%
P*.:/9+9):.6%
Q969:NA.6%
P*.:/9+9):.6%
Q969:NA.6%%
$:N69/-E):6>%
9OO'*/9):%
P*.:/9+9):.6%
Q969:NA.6%
!(.:9/-%
9:/+*A5+9):C%
/A(()*+%
#)*O.66>%:):'R%%
/)O'%87%
/5.GG)6B9:N%G)*%
5):+':+%%
!(.:9/-S%D8LT%
O.+-%
!(.:9/-S%D8LT%
O.+-%
!(.:9/-S%%D8L%%
!(.:9/-S%D8LT%
O.+-%
$:N69/-%9:/+*A5+9):C%
/A(()*+%
U):+':+%
5)A*/'/%9:%
$:N69/-%
!-'6+'*'B%$!8%
56.//'/%
$!8T%*'.B9:NT%
O.+-%
$!8%5):+':+%
56.//'/%V%0%-*/%
$!8%('*%B.>%
0%$!8%56.//'/%
('*%B.>T%7%
/59':5'%56.//%
W*.BA.+9):%*.+'/%
G)*%$88/%
0JK%%
07K%
0XK%
J7K%
1XK%%
2.2 Participants
2.2.1 Target Group: SIFE
As the majority of SIFE in New York City speak Spanish as their native language, we
selected for participation one hundred and three Spanish-speaking ELLs identified as SIFE by
their schools and/or by the New York City Department of Education. All the students were in the
ninth or tenth grades and each attended one of five New York City high schools described above.
The identification of SIFE is neither systematic nor standardized across schools. In selecting
our SIFE participants, we relied on the schools’ placement even though the schools differed in
their procedures for identifying students as SIFE. All of the schools participating in our study
used the DOE’s pre-screening SIFE identification, and four of the schools used some additional
tools in order to identify SIFE. For example, one school collected a writing sample in the
student’s native language; two of the schools gave Spanish proficiency tests to native Spanish-
speakers; and another interviewed the student and the student’s parent or guardian in order to
gather more information about the student’s educational background.
!"#$%&'()*+%,-./'%""%
%
%
I%
2.2.2 High School Comparison Groups
In order to compare SIFE with their peers, we included two participant groups from some of
the same high schools as the SIFE in our study, a native English speaker group and an English
language learner group.
Native English Speaker Group (NES)
The first comparison group consists of thirty-eight ninth graders who speak English as their
native language. The students attended three of the same high schools as the SIFE in our study.
This allowed us to control for demographics like socio-economic background, as students at a
given high school are generally drawn from the same neighborhood.
English Language Learner Group (ELL)
The second group consists of twenty-two ELLs from one of the high schools attended by the
SIFE in our study. The students had entered the New York City school system at the same time
as the SIFE group. Thus, the ELL group and the SIFE group have been in the US for roughly the
same amount of time, allowing for a controlled comparison between SIFE and ELLs in their
development of English language and literacy. This comparison is very important in determining
what makes SIFE a special subgroup of ELLs. By comparing a group of SIFE with a group of
ELLs who have had the same amount of time to learn English at the same schools, we can clarify
that the groups are different and better characterize the gap between SIFE and ELLs.
2.3 Materials and Procedures
The study was conducted over a period of eighteen months, and data were collected in two
stages, which were divided according to school year. During the students’ first year (Year One)
in the NYC school system, we administered assessments only to the SIFE group in their native
!"#$%&'()*+%,-./'%""%
%
%
X%
language (Spanish). During the students’ second year (Year Two) in the NYC school system, we
administered assessments to SIFE in both their native language and English. The NES group and
the ELL group were given assessments in English during Year Two.
All assessments were administered by graduate students trained in the procedures for giving
the tests. Spanish assessments were administered by bilingual research assistants.
2.3.1 Intake Questionnaire
Our research team developed an oral intake questionnaire in order to obtain the following
information about the students: personal and language information, family and home
background, language and literacy practices outside of school, and education history. The
questionnaire was given as an oral interview in Spanish between one interviewer and one
student, with answers audio-taped and hand-written by the interviewer1.
2.3.2 RISLUS Syntax Test (Spanish and English)
This assessment was developed by RISLUS and is designed to evaluate the acquisition of
sentence structure independently from vocabulary. In the present study, we used the Spanish
syntax test as a measure of typical native language development relevant to non-academic
functions of language2. Previous studies of children who speak one language (monolingual) have
established that the structures targeted in this test are benchmarks in native language
development, and are mastered by age ten (Sheldon 1974; Hsu, Cairns, and Fiengo 1985).
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1 The audio-recorded interviews were transcribed, a process that included checking and supplementing the responses
written by the interviewer. Responses were coded so that the results could be quantified, and coded responses were
then entered into a statistical analysis computer application (SPSS) for data analysis.
2 This is in contrast to the language skills tested in the Academic Language and Literacy Diagnostic (ALLD), to be
discussed, which are specifically relevant to the development of academic language.
!"#$%&'()*+%,-./'%""%
%
%
7H%
Besides being indicators of typical language development, these structures also occur frequently
in scholastic texts. In light of these facts, we felt that it was important to assess our participants’
comprehension of such structures.
The test combines listening comprehension and picture selection in a task that can be
administered individually or in groups. The student hears a sentence twice, and looks at three
pictures. The student is then asked to choose the picture that corresponds to the sentence. See
Appendix TKTK for samples of the sentence structures tested. Scores were calculated for each
student as percent correct.
2.3.3 Oral Language Proficiency Test (Versant) (Spanish and English)
The Versant (Pearson) is a test of oral language proficiency measuring sentence mastery,
vocabulary, fluency, and pronunciation. Each test item requires the student to understand a
spoken utterance and respond appropriately. The Versant is an automated test given over the
phone. Since the Versant is an automated test using speech recognition, all of the scores are
automatically calculated by a program accessed though the Ordinate website.
2.3.4 Working Memory Task
We administered this task because low literacy skills could (in principle) be explained by a
poor working memory. This test allowed us to control for such general problems, and attribute
any low literacy findings to factors unrelated to general cognitive development.
In order to determine the working memory of the students, we administered two tasks from
the Batería III (Riverside). Both tasks were given in Spanish on an individual basis. One task
involved word recall and the other involved reverse number recall. We administered these tasks
!"#$%&'()*+%,-./'%""%
%
%
77%
to determine if any of the students have a poor working memory. The Batería III scoring
software produced a score report for each student.
2.3.5 Academic Language and Literacy Diagnostic (ALLD) (Spanish and English)
The ALLD was developed over a period of two years through pilot testing, item analyses,
and collaboration with the NYC DOE and Pearson publishers. The Spanish ALLD used for our
research has been adapted from a US standardized Spanish language test published by Pearson:
the Aprenda Achievement Test Series, Third Edition. The English ALLD used for our research is
taken from its English equivalent, the Stanford Achievement Test Series, Tenth Edition. The
ALLD is a cumulative assessment, and contains items from a range of grade levels; it is a
multiple choice test and includes literacy and content skills, which are discussed below. (For
more detailed information on the development of the ALLD see Appendices. See Appendix X
for the criteria used for the development of the Spanish and English ALLD, and the adaptations
leading to the final versions of the diagnostics. The version of the ALLD in Spanish and English
finalized for distribution in the NYC schools is slightly different from the ALLD used for
research purposes. See Appendix Y for an outline of this diagnostic.)
2.3.5.1 ALLD Sections
i. Pre-literacy (Basic Literacy)
The basic literacy section of the assessment consists of items testing phonological and
orthographic awareness, word reading, and simple sentence comprehension.
!"#$%&'()*+%,-./'%""%
%
%
70%
ii. Reading Vocabulary
The reading vocabulary section of the ALLD assesses knowledge of synonyms3, multiple
meaning words4, and context clues5.
iii. Reading Comprehension
The reading comprehension section of the ALLD requires the students to read a passage, and
then answer multiple-choice questions about the passage. The passages are either informational
or functional. The questions assess basic understanding and some test higher level skills such as
critical analysis, strategies, and interpretation.
iv. Language
The language section of the ALLD assesses the student’s knowledge skills related to writing,
with questions of language mechanics (punctuation and grammar) and language expression (pre-
writing exercises and passage organization).
vi. Math
The Math Section of the ALLD consists of two main sections: Math Procedures and Math
Problem Solving. The Procedures section consists of math items that require number operations,
with virtually no reading. The Problem Solving section consists of word problems, and requires
the student to read in order to understand and answer the questions about math content
knowledge.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3 The synonym section requires a student to choose a synonym for a printed word.
4 The multiple meaning words section provides a target sentence, and four subsequent sentences all containing the
same word used in four different ways. The student must select the sentence in which the word is used in the same
way as in the target sentence.
5 The context clues section provides a target sentence with an underlined word; the target sentence provides a clue to
the meaning of the underlined word. The student must choose the meaning of the underlined word.
!"#$%&'()*+%,-./'%""%
%
%
71%
Since the reading vocabulary, reading comprehension, language, and math sections of the
ALLD were drawn from a range of grade levels, the scores are reported as grade level scores.
The Spanish ALLD Pre-test was administered in five forty-five minute sessions and the Post-
test and English ALLD were administered in three forty-five minute sessions. Two research
assistants administered each section of the assessment to a group of students in a classroom
setting within the students’ schools.
2.3.6 Classroom Observation Checklist
Classroom observation procedure
Supplementing the quantitative portions of the SIFE study was a qualitative portion, the
purpose of which was to observe SIFE performance in their classrooms and reflect upon
materials and instructional practices related to SIFE language and literacy learning. To collect
data to achieve this purpose, specific classes in the five schools in the study were selected, and
an observation procedure was devised to examine teaching and learning practices in each type of
class. The five types of classes were English as a second language (ESL), ESL Math, Native
Language Arts (NLA, i.e., Spanish), bilingual Mathematics, and Spanish-language Mathematics,
the academic subjects of interest in the overall study and the foci of the Academic Language and
Literacy Diagnostic (ALLD) periodically administered to the SIFE in the study. The numbers of
classroom observations in each academic subject per school are given in Table 2.
Table 2. Number of observations in SIFE schools by class subject
Subject
School
Subject totals
No. 1
No. 2
No. 3
No. 4
No. 5
ESL
2
2
1
4
3
12
!"#$%&'()*+%,-./'%""%
%
%
72%
NLA
-
2
-
1
-
3
ESL Math
1
-
2
-
-
3
Sp Math
-
1
-
-
1
2
BL Math
-
1
1
1
0
3
(School totals)
3
6
4
6
4
-
The total number of classroom observations across schools was 23. The presence of
multiple observations in some subject categories indicates that there was more than one teacher
teaching those subjects in a particular school. It can also be seen in Table 2 that there are
unequal numbers of observations across schools, reflecting the fact that not all schools had equal
numbers of all class types (e.g. ESL, NLA Spanish) and, also, that there were unanticipated
limitations in researchers' access to schools.
As for the observation procedure used, this consisted of two trained observers
simultaneously observing the same class, with each observer taking notes on separate forms. The
observers seated themselves at different points within a classroom, took their respective notes,
and then debriefed each other after the class to arrive at a consensus as to what was observed.
The observers were the consultant to the project and one of two research assistants who had
undergone several training sessions prior to the observations. At these sessions the trainees
developed focused observation and debriefing skills through repeated viewing of videotaped
high school ESL classrooms comparable to those in which the SIFE observations were to be
conducted.
There were two different observation forms used in this procedure. One form consisted
of a checklist, used to lend focus to the observations and to provide a common framework within
!"#$%&'()*+%,-./'%""%
%
%
73%
which different classroom environments could be compared. The checklist included various
aspects of the SIFE classroom lessons, for example, types of student responses to the teacher's
oral and written language, and student engagement with classroom activities. The checklist also
contained a detailed listing of aspects of literacy practices in the classroom, for example, whether
or not the literacy material used in an observed lesson was age-appropriate or grade-appropriate,
and the presence and sufficiency of native-language scaffolding. (See Appendix, "SIFE
Observation Checklist.") Each of the 30 checklist items included some kind of rating scale. For
example, in one item, "Students show interest/engagement [in classroom activities]," possible
answers were all, most, some, and few; these indicated approximately how many students seemed
to be engaged in the activity. (Each possible answer was assigned a numerical value in the data-
analysis phase of the study, so for example, all equaled 4, most = 3, some = 2, and few = 1.) The
observer answered each item by circling the appropriate answer with a pen or pencil.
A second observation form consisted of handwritten notes taken down by the consultant
during observations. These notes generally focused on the aspects of interest in the above-
mentioned checklist, but often went beyond them. When possible, these classroom notes
included notes taken during SIFE-school teacher interviews conducted after observations. The
rationale behind the use of hand-written notes was twofold. First, rich, descriptive accounts of
classroom behavior or teaching methods could be used to complement the more focused
checklist tallies with illustrative detail. Second, aspects of SIFE lessons that were not anticipated
by the range of topics represented in the checklist could conceivably be captured with a more
"wide-angled" observation form.%%%
!"#$%&'()*+%,-./'%""%
%
%
74%
2.3.7 Exit Questionnaire
The Exit Questionnaire is similar to the Intake Questionnaire, in that it is given orally on an
individual basis and asks questions related to the students' language and literacy practices. The
Exit Questionnaire also gathers information regarding the students' feelings about their
experience in school, their recommendations for improvement in their academic classes and
schools in general, and their motivation for graduating from high school and continuing on to
college.
3. Results and Discussion: Year One (2006-2007)
3.1 Participant Attrition
The number of participants reported in the results section varies by test, for example, the
number of students who completed the Spanish Literacy diagnostic at Time 1 was 98, but only
91 of those students completed the Intake Questionnaire. These varying numbers were due to
student attrition from the beginning to the end of the study. In Table 3 we have indicated the
total number of students for whom we are missing some data at Time 1, Time 2, and both Time 1
and 2 combined. The missing data is also subdivided by the three main reasons for the attrition:
(1) some students were discharged, (2) some students were unwilling to participate, and (3) some
students were frequently absent from school, or could not be located, at the time of testing.
Table 3. Number of Participants with missing data by year and in total
Year 1
Year 2
Year 1 and Year 2
191
5
24
122
13
13
12
45
55
43
53
92
1 Number includes 1 student who died during the study.
2 Number includes 1 student who was in Special Ed. prior to the onset of the study.
!"#$%&'()*+%,-./'%""%
%
%
7J%
3 Number includes 1 student who was suspended during year 2.
%
We will now report the results for Year 1, followed by Year 2.
3.2 Intake Questionnaire
Background
Ninety-one students completed the Intake questionnaire given at Time 1. These students were
between fourteen and nineteen years old, with sixteen being the average age. Most of them
(77%) emigrated from the Dominican Republic while the other 23% emigrated from various
other countries, such as Colombia, Ecuador, El Salvador, Guatemala, Honduras, Mexico, and
Puerto Rico. The home language for 97% of the students was Spanish, with three students
reporting that they also spoke an indigenous language at home. Based on information given
about their age of arrival in US, we found that 93% of the students had been living in the US for
two or fewer years and 7% for more than two years.
The majority of students answered positively to our questions about adjusting to their new
surroundings: All reported that they liked living in NYC; 98% said that they liked their new
school, and all of them reported liking their new classes. In the following sections we report on
the most important results from the different sections of the questionnaire.
Family and Home Background
Most of the students (86%) reported living with at least one parent in the US while a smaller
percentage (14%) reported that they did not live with either parent. Of the 13 students who
reported living with neither parent, 12 reported living with other relatives. The student who did
not live with any relatives was 19 years old at the time of the interview. Among 78 students who
reported knowing the highest level of education in their households, most of them (60%)
reported high school as the highest level of education and the majority (56%) reported that
!"#$%&'()*+%,-./'%""%
%
%
7I%
schooling was conducted in Spanish and English. In addition, a majority of the students (97%)
reported having family members still living in their country of origin.
Exposure to Spanish and English
A large majority (69%) of students reported that both Spanish and English were spoken in
their neighborhoods. Out of 74 students, 78% reported interacting in English with a person in
their household. A great majority (95%) also reported being exposed to some English outside of
school in the form of watching television (30%) or via Internet access (48%), and the rest
through other means.
Reading and Writing in Spanish and English
Additionally, we asked questions about their literacy practices outside of the school and found
that 84% of the students reported that they enjoyed reading. Of these students a small majority
(52%) preferred reading in Spanish while the rest reported using some English in reading. A
majority (82%) also reported that they enjoyed writing. Of these students a large majority (77%)
preferred writing in Spanish while the rest reported using some English in writing.
Education History
Half the students reported being educated in cities while the other half reported having
some schooling in a rural school. 65% reported having no gaps in their education. Of 32
students that did report gaps in formal education, 31% reported a gap of two or more years and
69% reported a gap of less than 2 years. In addition to gaps in formal education, we also asked
students with reported gaps whether they had studied during the gap; most of them (75%)
reported they had not done so.
!"#$%&'()*+%,-./'%""%
%
%
7X%
Educational Aspirations
Ninety one students answered questions regarding their future plans and aspirations. We
divided their answers into four categories, related to what types of aspirations they had for the
future: a) Professional Aspirations (teacher, lawyer, doctor), b)- Non-Professional Aspirations
(firefighter, baseball player), c) Social Aspirations (travel, family) and d) Doesn’t Know / No
Plans. Most of the students (54%) had professional aspirations, 31% had non-professional
aspirations, 7% had social aspirations and 8% were unsure of their plans.
3.3 Spanish Syntax Test
The mean score on the Spanish syntax test was 89% correct, with a standard deviation of 12
and a range of 36% to 100%. Overall, the students do not appear to have any developmental
native language problems.
3.4 Spanish Oral Langue Proficiency Test (Versant)
The mean score on the Versant test score was 80% correct, with a standard deviation of 16
and a range of 34% to 100%6. The scoring program describes 80% correct as indicating that the
student has “fluent, smooth, intelligible speech; controls appropriate language structure for
speaking about complex material.” Again, these results suggest that the majority of these
students do not have developmental delays in their native language.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
4%Note that there were two low outliers, scoring 34% and 35%. One of these students is not
included in the further analyses as he didn’t complete many of the other tests. The other student
had likely scored very low because of the testing situation: she spoke very quietly on the phone
and was uncomfortable speaking on the phone. She has more ‘typical’ scores on other tests.%
!"#$%&'()*+%,-./'%""%
%
%
0H%
3.5 Working Memory Task
We administered the working memory task to twenty-three of the students from the SIFE
group. 78% of the students tested had scores indicating an average working memory7. Therefore,
we cannot conclude that the SIFE population is a struggling group of learners because of a
general cognitive problem of working memory.
3.6 Spanish ALLD Year One
Basic Literacy:
The mean score on the basic literacy section of the ALLD was 96% correct, with a range of
69% to 100% and a standard deviation of 4.5. Thus, the ninety-eight students in the SIFE group
have basic literacy skills and are able to decode and read simple sentences. Their high
performance on the basic literacy portion of the ALLD shows that the students do not have low-
level reading difficulties, such as dyslexia.
Reading Vocabulary:
Grade Levels
The reading vocabulary section of the ALLD Pre-test consists of items from the third grade
level through the seventh grade level. Figure 1 represents the number of students performing at
grade levels three and below, and at grade levels four through five and six through seven (see
Appendix TK for a more detailed figure, with all grade levels included).
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
7 4% scored above average and 17% scored below average. The latter group will be discussed below.
!"#$%&'()*+%,-./'%""%
%
%
07%
Figure 1. Reading Vocabulary: Percentage of Students Scoring at Grades 6-7, Grades 4-5 and Grades 3
and Below
The average grade level achieved by the students on the reading vocabulary section of the
ALLD Pre-test is fifth grade. The students are enrolled in the ninth and tenth grades, so they are
generally at least four grade levels below the expected level in their knowledge of reading
vocabulary in their native language. Forty-six percent of the students scored at the fifth grade
level or below. Fifty-four percent of the students scored at the sixth and seventh grade levels,
with twenty-two percent of the students performing at the seventh grade level, and one student
performing at ceiling (100% correct).
Subskills
Recall that the reading vocabulary section of the ALLD includes three types of items,
assessing three separate subskills: synonyms, multiple meaning words, and context clues.
Synonyms items test the ability to choose a synonym for a given word, multiple meaning words
items test the ability to recognize the multiple meanings of a given word, and context clues items
!"#$%&'()*+%,-./'%""%
%
%
00%
test the ability to use given information to determine the meaning of an unknown word. Figure 2
shows the average score of the SIFE group on these three subskills, across all grade levels.
Figure 2. Reading Vocabulary: Mean Percent Correct for Three Subskills
For the three subskills, tested at grade levels three through seven, the students are scoring
between 60 and 70 percent correct. The students are performing highest on context clues and
lowest on multiple meaning words. There is a main effect for subskill type (F(1, 97) = 4.902, p <
.01); post-hoc comparisons revealed that the only significant difference was between multiple
meaning words (M = 62%) and context clues (M = 69%, p < .01). In order to get a more detailed
description of the skill performance, we computed the subskill scores for the lower grade levels
and the higher grade levels in order to see if at the easier level or the more difficult level the
!"#$%&'()*+%,-./'%""%
%
%
01%
students were clearly scoring better on a given item type. See Figure 3 for the results with the
items divided into difficulty level.
Figure 3. Reading Vocabulary Subskills: Mean Percent Correct for Two Difficulty Levels
Note that overall the students perform better on the subskills in the third and fourth grade
levels. On items from grades three and four, the students score significantly lower on multiple
meaning words (M = 69%) than on synonyms (M = 81%, p < .001) and context clues (M = 78%,
p < .001); however, there is no statistical difference in performance on synonyms and context
clues. The students have the largest decrease in scores on the synonyms items. They appear to
know the vocabulary words generally learned in grades three and four, but do not know many of
the words generally learned in grades five, six and seven. On the items from grades five through
!"#$%&'()*+%,-./'%""%
%
%
02%
seven, however, there is no difference in performance on the three subskills (F(2, 194) = .315, p
= ns).
Reading Comprehension:
Grade Levels
The reading comprehension section of the ALLD Pre-test consists of items from second
grade through fifth grade. Figure 4 represents the number of students performing at grade levels
four and five and grade three and below (see Appendix TK for a more detailed figure, with all
grade levels included).
Figure 4: Reading Comprehension: Percentage of Students Scoring at Grades 4-5 and Grades 3
and Below.
The students’ performance on the reading comprehension section of the assessment is lower
than that of the reading vocabulary section. The majority of the students (68%) are performing at
or below the third grade level on reading comprehension in their native language. The average
grade level score on the reading comprehension section of the ALLD Pre-test is third grade.
Thus, the students are generally six grade levels below the expected level of ninth grade in
!"#$%&'()*+%,-./'%""%
%
%
03%
reading comprehension. Although twenty-three percent of the students scored at fifth grade level,
no student reached ceiling on the reading comprehension portion of the ALLD Pre-test.
Subskills
The reading comprehension section of the ALLD tests both basic understanding skills and
text level skills. Items assessing basic understanding can be answered by referring back to the
text, because the answer is explicitly stated in the text. Text level skills are higher level
comprehension skills, and require the student to think critically, make connections, and use
reading strategies. Since text level skills are higher level skills than basic understanding skills,
we would expect the students to perform better on the basic understanding skill than the text
level skill. Figure 5 represents the average scores on the two subskills on all items in the reading
comprehension portion of the Spanish ALLD Pre-test.
!"#$%&'()*+%,-./'%""%
%
%
04%
The students score significantly higher on the basic understanding skills than on the text level
skills (t(97) = 14.07; p < .001), with an average of 73% on basic understanding items and 49%
on the items testing higher level skills. The students do not appear to have strong reading
strategy skills. As with the reading vocabulary section, we also report the subskill performance
among the lower grade levels and the higher grade levels. Figure 6 represents the average score
on the subskill of the reading comprehension seciton of the diagnostic among items from the
second and third grades and from the fourth and fifth grades.
Figure 5. Reading Comprehension: Mean Percent Correct for Two Subskills
!"#$%&'()*+%,-./'%""%
%
%
0J%
Figure 6. Reading Comprehension Subskills: Percent Correct for Two Difficulty Levels
Among the lower grade levels, when the texts are easier to read and the questions are simpler
to answer, the students score quite high on both basic understanding and text level skills. We see
the difference in skill performance in the higher grade levels, with a greater decline in text level
skills. The students score 42% on text level skills while scoring 65% on basic understanding
items. We see a greater decline in the most important subskill of reading comprehension.
Language:
The language section of the ALLD consists of items from the third and fourth grade levels.
The mean grade level on the language section of the ALLD Pre-test is between third and fourth
!"#$%&'()*+%,-./'%""%
%
%
0I%
grade. The pre-writing skills measured by the language section are important as one is learning to
write. About half of the students do not evidence mastery of these skills at the fourth grade level.
While 64% of the students score at the grade four level, only two students scored at ceiling.
Summary of Reading Skills
The SIFE population tested here appears to have basic literacy skills; the students are able
to decode words and read simple sentences. However, they score lower on vocabulary and
reading comprehension skills. The average grade level performance on the vocabulary measure
was fifth grade and the average grade level performance on the reading comprehension measure
was third grade. Recall that the students are enrolled in the 9th and 10th grades, and are thus very
far behind in their academic literacy skills. The students appear to struggle with multiple-
meaning words on the vocabulary section, scoring low on this subskill in the lower and higher
grade levels tested. The students appear to have the most difficulty with text level skills in
reading comprehension, which is arguably why their reading comprehension scores are so low.
They do not have well developed higher-level reading skills, which are needed in the higher
grades the students are enrolled in.
Math:
Grade Levels
The Math section of the ALLD Pre-test includes items from third through sixth grade. Figure
7 shows the grade level performance of the SIFE during Year One, across both sections of the
test (i.e. procedures and problem solving).
!"#$%&'()*+%,-./'%""%
%
%
0X%
Figure 7. Math: Percentage of Students Scoring at Grade 6, Grades 4 and 5 and Grades 3 and Below
The mean grade level on the Math section of the ALLD Pre-test is fourth grade. Nearly half
the students (49.5%) score at or below the third grade level.
Subskills
The math section consists of two major item types: problem solving and procedures. The
problem solving items require the student to read in order to solve a math problem and the
procedures items do not require reading. We would expect that since the students do not have
high levels of literacy they would perform better on the procedures items as they do not require
reading. Figure 8 shows the results of the two subskills from all items in the math section of the
ALLD.
!"#$%&'()*+%,-./'%""%
%
%
1H%
Figure 8. Math: Mean Percent Correct for Two Subskills
Contrary to our hypothesis that the students would evidence better performance on
procedures than problem solving, the students had a mean score of 62% on the problem solving
items and a mean score of 54% on the procedures items, with a significant difference in scores
(t(100) = 4.55; p < .001). We examine the results further by reporting the subskills by level of
difficulty. Figure 9 represents the subskill results among the items from grades three and four
and the items from grades five and six.
!"#$%&'()*+%,-./'%""%
%
%
17%
Figure 9. Math Subskills: Percent Correct for Two Difficulty Levels
We see a steep decline in math skills from the lower to the higher grade levels. The
students do quite well on simple additon and subtraction of whole numbers, but they do not have
the higher level math skills required by grades five and six. The students perform equally low on
problem solving and procedures at the higher grade levels.
3.7 Comparison of Abilities in Native Language: SIFE (Spanish ALLD) Compared to High
School Native English Speakers (English ALLD)
In order to compare academic literacy skills in the native language of SIFE with the
academic literacy skills in the native language of ‘typical’ students in New York City high
!"#$%&'()*+%,-./'%""%
%
%
10%
schools, we administered the English ALLD to native English speakers attending three of the
same high schools as SIFE. Figure 10 shows the comparison of these two groups.
Figure 10. Reading Comprehension and Reading Vocabulary Grade Level Scores in Native Language:
SIFE Compared to Native English Speakers in NYC High Schools
The native English speakers scored at grade level on both reading vocabulary and reading
comprehension. The students are all enrolled in the ninth grade and score between ninth and
tenth grade on the reading vocabulary section of the ALLD and score at the ninth grade level on
reading comprehension. These results verify the grade level assessments of the ALLD, and show
how large the gap is in native language literacy skills between SIFE and their native English-
speaking peers. As reported above, in their native language, SIFE score at the fifth grade level on
reading vocabulary and at the third grade level on reading comprehension. The largest gap
between the two groups is in native language reading comprehension: SIFE generally score six
grade levels below their peers.
!"#$%&'()*+%,-./'%""%
%
%
11%
4. Results and Discussion: Year Two (2007-2008)
4.1 Spanish ALLD Year Two Compared to Spanish ALLD Year One
When reporting the Spanish ALLD Post-test results, we will focus on a comparison of the
ALLD Pre-test and Post-test among those students who completed both tests (N=48) in order to
measure gains the students made in native language academic skills in one year.
Reading Vocabulary:
In comparing the Pre-test and the Post-test we include the same number of items from each
skill and grade level, up to grade seven in reading vocabulary. We also include the results from
the higher grade levels on the Post-test (i.e. those not included in the Pre-test); that is, for reading
vocabulary we present results for grade nine. Figure 11 represents the mean score on each grade
level, on both the Pre-test and Post-test. This figure shows a comparison of the forty-eight
students who completed the reading vocabulary section of the Spanish ALLD Pre- and Post-tests
!"#$%&'()*+%,-./'%""%
%
%
12%
The scores on both tests steadily decline as the grade levels increase. However, there is a
significant increase between time one and time two. Note that the students improved in their
performance in reading vocabulary, with statistically significant gains in both sixth and seventh
grades (sixth grade: t(47) = 2.6, p = .01 and seventh grade: t(47) = 2.9, p<.01). Among the forty-
eight students who took both the Pre- and Post-test, the mean grade level score on the reading
vocabulary section of the ALLD Post-test was between sixth and seventh grade, as compared to a
mean grade level score of fifth grade on the Pre-test: this represents a gain of about 1.5 grade
levels.
Figure 11. Reading Vocabulary Results: Grade Level Performance Pre-test compared to Post-test
!"#$%&'()*+%,-./'%""%
%
%
13%
Reading Comprehension:
Again, in comparing the Pre-test and the Post-test we include the same number of items from
each skill and grade level, up to grade five in reading comprehension. We also include the results
from the higher grade levels on the Post-test (i.e. those not included in the Pre-test); that is, for
reading comprehension we present results for grades six, seven and nine. Figure 12 represents
the mean score on each grade level of the Pre- and Post-tests. This figure shows a comparison of
the scores of the forty-eight students who completed the reading comprehension section of the
native language literacy assessment in both Year One and Year Two.
The students improved their performance in reading comprehension, with significant gains at
the fifth grade level (t(47) = 2.6, p = .01). The mean grade level score of the forty-eight students
on the reading comprehension section of the ALLD Post-test was fifth grade, as compared to the
Figure 12. Reading Comprehension Results: Grade Level Performance Pre-test and Post-
test
!"#$%&'()*+%,-./'%""%
%
%
14%
mean grade level score of between third and fourth grade on the Pre-test: this represents a gain of
about 1.5 grade levels. The scores on the Post-test drop after the fifth grade level, with lower
scores on the sixth, seventh and ninth grade levels.
Language:
The language section of the ALLD Post-test consists of items from grade levels three and
four. Forty-eight students completed the language section of the native language literacy
assessment in both Year One and Year Two. The students improved in their performance in
language; this means that their writing skills in their native language increased over the course of
one year. The mean overall score was significantly better on the Post-test than the Pre-test (t(47)
= 2.6, p = .01).
Math
On the Math section of the ALLD, forty-three students completed both the Pre- and Post-
tests. The Math section of the ALLD Pre-test includes items from grades three to six and the
Post-test includes items from grades three to eight. Similarly for Math, we include the same
number of items from each skill and grade level up to grade six. We also include results from
grade eight on the Post-test; grade eight items were not included on the Pre-test. Figure 13
represents the mean score on each grade level of the Pre- and Post-tests.
!"#$%&'()*+%,-./'%""%
%
%
1J%
The students performed significantly better on the sixth grade level math items on the Post-
test than they did on the Pre-test (t(42)=5.1, p<.001). The mean grade level score for the forty-
three students on the Pre-test was between the third and fourth grade, compared to between the
fourth and fifth grade level on the Post-test: this represents a gain of approximately one grade
level. The Post-test scores at the eighth grade level are lower than the lower grade levels, with
only 40% correct.
Figure 13. Math Results: Year One Grade Level Performance compared to Year Two Grade Level
Performance
!"#$%&'()*+%,-./'%""%
%
%
1I%
Summary: Improvement (Individual Gains in Spanish T1 – T2)
Of the forty-eight SIFE who completed the ALLD Literacy in Spanish at both Year One
and Year Two, most of them, 79%, did at least as well or better overall (this includes the
Vocabulary, Reading Comprehension and Language Sections). A paired-samples T-test shows
that the difference in their overall Literacy scores from Year One to Year Two is significant.
When just looking at the Spanish Vocabulary Section, again most SIFE, 73%, performed as well
or better on this section of the ALLD. The difference in Vocabulary scores from Year One to
Year Two is also significant. On the Spanish Reading Comprehension section, slightly less than
half of SIFE, 44%, performed as well or better in Year Two as they did in Year One, however,
the difference in Reading Comprehension scores from Year One to Year Two is not significant.
For the Spanish Language Section, the majority of SIFE, 73%, achieved scores at Year Two that
were either the same or better as they had achieved at Year One. The difference in the scores on
the Language Section of the ALLD is significant.
Improvement (Individual Gains in Math T1 – T2)
Of the forty-three SIFE that completed the Spanish Math ALLD sections at both Year
One and Year Two, the majority of students, 67%, did as well or better at Year Two on Spanish
ALLD Math. The difference in their scores from Year One to Year Two is also significant.
4.2 English Diagnostics
We administered English diagnostics to SIFE in Year Two only, as in Year One they had
(presumably) only been in the NYC school system for less than one year and would likely not
have English skills sufficient for completing diagnostics in English. In our pilot study, we found
that our twelve participants who had also been in NYC schools for less than one year were not
!"#$%&'()*+%,-./'%""%
%
%
1X%
able to understand and complete English diagnostics. Recall that in addition to our target group
of SIFE, we include comparison groups who also took the English ALLD.
4.2.1 English Syntax Test
The mean percent correct on the English syntax test was 63%, with a standard deviation of
17 and a range of 32% to 90%. The students are still in the beginning stages of acquiring
complex English syntactic structures.
4.2.2 English Oral Language Proficiency Test (Versant)
The mean percent correct overall on the English Versant test was 34%, with a standard
deviation of 9 and a range of 25% to 54%. The scoring program describes this score as indicating
that the student “can manage some slow, short, isolated utterances, or spoken formulas, but has
difficulty following any native conversation; test-taker may often pause to search for words and
may be difficult to understand.” Just as with the syntax test, the Versant results indicate that the
students are still in the beginning stages of acquiring English.
The above results regarding the English Versant and the English Syntax test show that our
SIFE are still in the process of acquiring the English language.
4.2.3 English ALLD
We administered the English ALLD to the SIFE students in order to measure their
acquisition of English academic literacy after being in NYC schools for almost two years. Figure
14 reports the grade level scores on the vocabulary section of the English ALLD.
!"#$%&'()*+%,-./'%""%
%
%
2H%
Figure 14. Vocabulary: Percentage of Students Scoring at Grades 9-11, Grades 5-7 and Grades 3 and
Below
Although the students have been in a NYC high school for almost two years, their English
academic skills are still very low. On the reading vocabulary section of the ALLD the majority
of students score at a level below high school, with 19% of the students scoring at the high
school grade levels (9-11). And 73% of the students score at the elementary school level, with
58% of the students scoring at or below the third grade level.
See Figure 15 for the Reading Comprehension grade level results of the English ALLD.
Figure 15: Reading Comprehension: Percentage of Students Scoring at Grades 7-11, Grades 4-6
and Grades 3 and Below.
!"#$%&'()*+%,-./'%""%
%
%
27%
The reading comprehension skills of the SIFE are also very low, with the majority of students
(77%) scoring at or below the third grade level. Only 4% of the students score at a high school
level on the reading comprehension section of the English ALLD, and 8% of the students scored
at a middle school level in English reading comprehension. See Figures 16 and 17 below for the
scores on the subskills of reading comprehension: basic understandig and text level skills.
!"#$%&'()*+%,-./'%""%
%
%
20%
Figure 16: Basic Understanding Subskill of Reading Comprehension
Figure 17: Text Level Subskill of Reading Comprehension
The students score above 60% on both item types at the second and third grade levels.
We see a steady decline in text level skills, with the students scoring 31% correct on the text
!"#$%&'()*+%,-./'%""%
%
%
21%
level items from the fourth and fifth grades; this low level holds for the remainder of the grade
levels.
Math Results
%Figure 18 shows the math grade levels for the students on the English ALLD.
Figure 18. English Math Results: Percentage of SIFE Scoring at Each Grade Level
On the math section of the English ALLD, the majority of the students (53%) scored at or
below the third grade level. None of the students scored at a high school level on the math
section of the English ALLD. In order compare the procedures and problem solving sections, we
computed the scores for these two subskills. Figure 19 represents the mean percent correct on the
two subskills of the math section.
!"#$%&'()*+%,-./'%""%
%
%
22%
Figure 19: English Math Results by Subskill
As expected, the students score higher on the procedures section than the problem solving
section, since the problem solving items require reading in English. See Figures 20 and 21 for the
percent correct on the two subskills across the grade levels.
!"#$%&'()*+%,-./'%""%
%
%
23%
Figure 20: Procedures Items: Percent Correct by Grade Level
Figure 21: Problem Solving Items: Percent Correct by Grade Level
!"#$%&'()*+%,-./'%""%
%
%
24%
The students score 89% correct on the procedures items from grade three. This shows
that the students have mastered the procedures items in grade three; these items involve addition
and subtraction of whole numbers. On other procedures items across the higher grade levels, the
students are scoring much lower, below 50% beginning with grades 6 and 7. The students also
score much lower on the problem solving items at the third grade level than the procedures
items. This result confirms that reading math problems in English results in lower scores on the
third grade items.
Performance in L2: SIFE (English ALLD) Compared to High School ELLs (English ALLD)
We administered the English ALLD to other ELLs in order to compare the results of the
SIFE group with other ELLs who have been enrolled in NYC high schools for the same amount
of time. Figure 22 shows the comparison of the two groups on their performance on the Reading
Vocabulary and Reading Comprehension sections of the English ALLD.
Figure 1. SIFE Compared to ELLs
%
!"#$%&'()*+%,-./'%""%
%
%
2J%
Although the ELLs are also not performing at the expected grade level (now tenth grade)
in English reading vocabulary or reading comprehension, they are performing at a higher grade
level than the SIFE on both those sections. The ELLs have an average reading vocabulary grade
level between eighth and ninth grade, whereas the SIFE have an average reading vocabulary
grade level of fourth grade. The ELLs have an average reading comprehension grade level of
sixth grade, whereas the SIFE have an average reading comprehension grade level of third grade.
Reading comprehension is well below expected grade level for both groups. The ELLs have
higher vocabulary scores, but still need instruction focused on higher level skills involved in
reading comprehension. These scores show that SIFE are struggling more than other ELLs in
acquiring English academic literacy skills, but both groups need instruction in vocabulary.
4.3 Classroom Observations
Summary report of SIFE classroom observations
ESL classes. Analysis of the information gathered from the observations revealed a
number of features, some of which appeared to positively enhance the academic literacy
experience for SIFE in ESL classes, and others that had a less effective impact. Positive aspects
of ESL instruction could be found in the areas of teaching materials and teaching activities.
Some of these activities did not deal with academic literacy per se, but were basic ESL activities
focusing on the four skills (i.e. listening, speaking, reading and writing) with continual visual-
text support on the board. Examples of such activities and materials included the use of arts and
crafts (e.g., Halloween mask-making) to focus attention on descriptive adjectives in a whole-
class paragraph-writing exercise, and continual focus on phonics and vocabulary in meaningful
contexts (e.g., breakfast foods in the U.S. and in the students’ countries of origin). In such cases
the activities provided vocabulary as a foundation for a culminating activity involving writing.
!"#$%&'()*+%,-./'%""%
%
%
2I%
Moreover, in three-quarters of the observations, the students engaged in activities that supported
literacy – for example, vocabulary practice and phonics practice. Also, in slightly over half of
the classes observed, academic content was provided, most of which was appropriate for the
level of the students in the class, whether sheltered or non-sheltered. However, in nearly half of
the observed cases, the texts were not age-appropriate, even though student comprehension of
written input from the teacher on the board, in texts, or on handouts, was generally good, with
scaffolding. (This is in contrast to student comprehension of teacher oral input; see below.)
Another positive aspect of ESL instruction concerned creative use of some of the so-
called “SIFE solutions” packages (Read 180, Rigor, Language!, and Achieve3000). It was
observed that those schools possessing these literacy materials found that they could not
implement them in their entirety, as suggested, due to the low proficiency of the SIFE students,
so either modified them or used components from them piecemeal. These components were
sometimes integrated in lesson plans focused on content. For example, one Rigor reader dealing
with a natural-science content area (glaciers) was integrated into a lesson whose activities
reinforced one another, so that the class was at once a lesson in cartographical conventions,
geographic knowledge, and vocabulary reinforcement (place names, colors), and including an
exercise in accessing written text (scanning for specific information) followed by some content-
focused writing (information transfer).
Differentiated instruction emerged as one promising approach, though there were few
cases of this. Where this was used, however, it was well designed. For example, one teacher
modified the standard use of the Read 180 program (i.e., rotating three separate groups of
learners through three respective work stations) to accommodate the SIFE students: various
!"#$%&'()*+%,-./'%""%
%
%
2X%
computer exercises dealing with phonics for (relatively) higher-proficiency students; story
reading and answering comprehension questions in writing for mid-proficiency students; and
book-sorting and listing activity for very low-proficiency students. Unfortunately, however,
non-differentiation of materials and instruction was the rule across most of the classes observed.
Despite the positive aspects of ESL instruction noted above, there were other factors that
seemed to limit the potential of the some of the above-mentioned materials and teaching
strategies, and thus possibly inhibiting significant progress in English literacy development
among the SIFE students. One factor may be the amount of total class time devoted to literacy
activities. In the ESL classes, this ranged from zero time to 100%, with the average being
61.25%, the bulk of this devoted to low-level reading skills such as phonics and vocabulary.
Qualitative analyses revealed some potential causes for low levels of involvement and learning,
including poor classroom management, poorly planned or implemented group activities, and the
use of too much “teacher talk.” In many classes, teaching effectiveness was limited when the
teacher gave unclear directions for activities, and/or not modifying their speech to accommodate
students’ low aural comprehension levels: Some teachers addressed students as though they
were native speakers of English, even in cases where teachers were aware of students’ SIFE
status. (In this connection, it should be noted that in three of the five schools observed, teachers
were unaware of the students' SIFE status.) Despite the fact that in all ESL classes there was a
wide range of ongoing assessment strategies used e.g., asking questions, making direct or
indirect corrections of errors, recasts, and circulation among workgroups and pairs in some
ESL classes, teachers were not seemingly aware of students’ comprehension or non-
comprehension of the material being taught or practiced (for example, from students’ facial
!"#$%&'()*+%,-./'%""%
%
%
3H%
expressions or body language), with missed opportunities for using informal assessment
strategies.
Transitions between activities of a lesson were sometimes abrupt and confusing for many
SIFE students. Classroom management was also occasionally problematic due to selection of
content. For example, during one class, the topic (Columbus) was observed not to engage
several of the students from the Dominican Republic, who complained among themselves (in
Spanish) that they had already been exposed to the material in their home country. Finally, and
perhaps crucially, the effectiveness of scaffolding in English (i.e. using various materials to aid
in students’ comprehension) was uneven across the classes observed. Scaffolding of content in
Spanish was for the most part either absent from the ESL classes observed, or was poorly used,
especially for the lower-literacy students in the classes – in these cases the teacher's scaffolding
seemed targeted, consciously or not, at the higher-proficiency students in the class.
NLA and Spanish/bilingual content (mathematics) classes. Regarding Spanish NLA,
bilingual, and Spanish-language content instruction, certain factors appeared to promote Spanish-
language learning. In contrast with the ESL classes, the amount of total class time spent on
literacy activities and skills ranged from 25% to 100%, with an average of 75%, the bulk of this
devoted to low-level skills such as vocabulary. As with the ESL classes, there were very few
instances of differentiated instruction; one example of such instruction in a mixed class was
when the SIFE group sequenced pictures of a story and discussed the story with a student
teacher, while the mainstream ELLs re-sequenced pictures illustrating a story and wrote short
sentences describing the pictured story scenes.
However, more prevalent aspects of instruction that appeared facilitative of Spanish
literacy learning, and of content learning through Spanish, included good classroom management
!"#$%&'()*+%,-./'%""%
%
%
37%
through clear directions for learning activities, whether in NLA or Math. There were virtually no
classroom management issues in the Spanish-language classes observed, only very transient and
manageable ones. In addition, classroom materials in content classes were generally suited to
SIFE-student age as well as proficiency level. As for use of group work, grouping SIFE with
mainstream ELLs after teacher-fronted presentation also seemed to be facilitative of content
learning in some bilingual Math and Spanish Math classes. Finally, ongoing assessment by the
teacher, consisting of circulation and monitoring of individuals, pairs, or workgroups, with
teacher recasting of earlier-presented material and directions, seemed to facilitate learning of
content in Spanish.
Certain other activities seemed helpful for students in addressing NLA texts, for example,
whole-class pre-reading activities in which vocabulary, key ideas, or (in some cases) narrative
structure is pre-taught. Second, during-reading activities, in which the teacher has the class
pause during oral, whole-class readings of text and asks them focused questions about content,
often constituted scaffolding for acquisition of vocabulary, in that unfamiliar words in the story
can be discussed in relation to the narrative context. Third, silent paired reading and whole-class
reading activities allowed for practice and discussion of vocabulary and meaning. Fourth,
vocabulary work and repeated listening of audiotaped NLA text readings appeared to help
students make associations between sound and symbol. Finally and significantly, sufficiency of
scaffolding of language and content was nearly omnipresent in the Spanish classes observed,
e.g., teacher-fronted presentation followed by group work, in turn followed by whole-class
sharing-out with student and teacher feedback in Spanish, as in the bilingual Math class. In this
connection, student comprehension of Spanish written input by the teacher, as on the blackboard,
texts, or handouts, was generally good.
!"#$%&'()*+%,-./'%""%
%
%
30%
Tentative recommendations: In considering the following recommendations, it should
be noted, as mentioned above, that this portion of the study was non-experimental; that is, not all
schools had equal numbers of all class or program types (e.g., ESL, NLA Spanish, bilingual,
content area, sheltered, non-sheltered.). Again, this was partly due to human-resource
constraints and partly due to problems with school access for local administrative reasons. Thus,
given that the observations were made of a select sample of classes, and since this qualitative
portion of the study could not control for most of the numerous aspects of the teaching
approaches or materials that were of interest, therefore any conclusions derived from the data
should be made with some caution.
However, the observation data do suggest some tentative recommendations regarding
SIFE instructional practice, which may be considered also candidates for research hypotheses in
follow-up studies involving SIFE. Chief among these is the need for detailed teacher awareness
of SIFE student vocabulary and aural and oral proficiency in English, which some ESL teachers
seemed not to fully appreciate, apparently treating SIFE in the manner of mainstream ELLs.
Such awareness by teachers has implications for a more measured rate of teacher classroom
speech, grammatical complexity of teacher speech, teacher use of oral and written vocabulary,
effective scaffolding of content as well as of language, clearly-indicated transitions between
classroom activities, classroom management generally, and overall effectiveness of literacy
activities. Related to this is effective teacher use of the informal assessment strategy of using
facial expressions and body language as cues to signal a need for recasts and/or a reduction of
rate of teacher speech. Finally, differentiated instruction is an approach that should be addressed
in the areas of materials development and in-class implementation, not only in non-sheltered
!"#$%&'()*+%,-./'%""%
%
%
31%
programs in which SIFE are mixed with mainstream ELLs, but potentially in sheltered programs
as well, since grade-level proficiency will differ among those categorized as SIFE.
4.4 Exit Questionnaire
At the end of Year Two, forty-four participants completed the Exit Questionnaire. The
majority of the students (57%) report being enrolled in the tenth grade; the rest were dispersed
from ninth through eleventh grades and one student reported being in the twelfth grade. In the
sections below we have summarized the results, and – where indicated – we have included some
correlations between their recorded responses on the Questionnaire and their results of the
ALLD.
Attitudes toward education
A large majority of students (80-100%) had very positive attitudes toward education.
Almost all responded that they believe that graduating from high school is important (98%) and
that their families believe graduating from high school is important (95%). Similarly, All SIFE
believe that it is important to graduate from college; almost all students (98%) report that this is
also important to their families. The majority of the students believe they will graduate from high
school (82%) and college (80%). The majority of SIFE report (68%) that they like going to
school both in their country of origin and in the United States.
Attitudes toward ESL classes
In general attitudes were very positive toward their ESL classes. Students generally
report understanding their teachers and finding their readings interesting. Only 40% of SIFE
report that they generally do not understand their ESL teachers when the teachers are speaking
English, nor do these students understand the readings in their ESL classes. The majority of SIFE
!"#$%&'()*+%,-./'%""%
%
%
32%
felt that their ESL classes helped them read (77%) and write (84%) better in English, and most
(61%) felt that they read better in English now than they did last year.
Attitudes toward English and Spanish
In general, attitudes were positive toward English. All SIFE believe it is important to
read and write well in English and most report that their families also think it is important to
learn English. Most SIFE think that it is important to spend time with friends that only speak
English.
Attitudes toward Spanish were also positive, although not as positive as for English.
Although 84% of SIFE believe it is important to read and write well in Spanish, only 30% report
that their families strongly believe that they should have classes in Spanish while learning
English. While most SIFE (61%) feel that Spanish is important to them, only 52% think it is
important to spend time with friends who speak Spanish.
Language Exposure: Correlations with the ALLD
Based on how students responded to questions on the amount of exposure they have had
to English, we arrived at a language experience index where the higher the index the student
received, the more English exposure the student has had. The total language exposure index
included all of the factors listed below.
Listening and Speaking: Amount of English the student hears and speaks in the home and
in the neighborhood.
Reading and Writing: Amount of English the student uses in reading and writing.
Because language input or exposure is considered so critical to language and literacy
development, we examined correlations between students’ Language Exposure indices and their
performance on the ALLD in both Spanish and English, with the following results:
!"#$%&'()*+%,-./'%""%
%
%
33%
We found a positive correlation between the Language Exposure index in total (i.e.
reading, writing, listening and speaking in English) and the English ALLD. That is, students that
reported using more English overall also performed better on the English ALLD. There was also
positive correlation between the Reading and Writing index and the Spanish ALLD. This means
that the students with a higher Reading and Writing index -- those who reported reading and
writing most often in English -- also achieved higher scores on the Spanish ALLD. This
suggests that literacy in the native language is related to literacy exposure and use in the second
language.
Aside from Language Exposure, we also compared ALLD results with attitudes toward
education, self-efficacy (whether the student believes s/he will graduate from high school and/or
college), and attitudes toward English and Spanish. Of all these comparisons, only attitudes
toward English were significantly correlated with the ALLD. For both the Time 2 Spanish Total
Literacy, attitudes toward English had a negative correlation with the Spanish ALLD score. This
means that the more positive the students’ attitudes were toward English, the lower their scores
were on the Time 2 Spanish ALLD (Pearson correlation: Spanish Literacy, r = -.31; Spanish
Vocabulary, r = -.34, p < .05).
Feelings about the Home Country: Correlations with the ALLD
Of the forty-four SIFE who participated in the Exit Questionnaire, the largest group (48%) is
ambivalent about whether they want to return to live in their home countries. When comparing
these students’ performance on the ALLD, we found that students who reported being
ambivalent about returning home did significantly better on English Reading Comprehension
than those who intend to return to their home countries (p = 0.006). That is, students who plan to
leave the US did not perform as well on the English ALLD, presumably because there was no
!"#$%&'()*+%,-./'%""%
%
%
34%
strong need to maintain English when they return home. On the other hand, those who thought
they might remain here had potentially greater reason for working to improve their English
skills, supporting the role that motivation plays in literacy development.
Native Language Arts Classes: Correlations with Spanish language gains
Four of the five schools included in the study offered some sort of native language
support to the SIFE students. Three schools offered Native Language Arts classes specifically
for SIFE. One of the schools that did not offer sheltered Native Language Arts classes for SIFE
offered an after-school class to help students prepare for the Spanish regents. Of the forty-four
SIFE that were administered the exit questionnaire, thirty-three report taking Spanish Language
Arts classes.
We compared individual gains in Spanish for students who have had Spanish Native
Language Arts classes (33 students) to those who have not (11 students) and found that there
were no significant differences between the groups in Spanish language gains. Furthermore we
found no significant differences in individual gains between students who were enrolled in
sheltered (native language arts) programs and those who were not in such sheltered programs,
where there could potentially have been more focused attention on their Spanish language needs.
Since this was not the case, we presume that improvement in Spanish was unrelated to their
school experiences.
Future Plans and Aspirations: Correlations with the ALLD
The forty-four SIFE that completed both the Intake and Exit questionnaires also answered
questions regarding their future plans and aspirations. We divided their answers into four
categories, related to what types of aspirations they had for the future: a) Professional
!"#$%&'()*+%,-./'%""%
%
%
3J%
Aspirations (teacher, lawyer, doctor), b) Non-Professional Aspirations (firefighter, baseball
player), c) Social Aspirations (travel, family) and d) Doesn’t Know / No Plans.
From the Intake Questionnaire, there were twenty-five students that reported having
professional aspirations; however, a year later only fourteen of those reported having
professional aspirations at the time of the Exit Questionnaire.
We compared the performance on the ALLD of the fourteen SIFE that maintained
professional aspirations to those who either didn’t maintain professional aspirations (or never
had professional aspirations) to see whether there was any relationship between their aspirations
and their academic performance. We noticed that there was a tendency for the professional
aspiration group to outperform the remaining twenty-seven on all sections of the ALLD in both
Spanish and English. Strikingly there was a tendency for this group to have higher scores on all
sections of the ALLD, although we only found the difference to be significant between the
groups on the Time 2 Spanish Literacy overall (p <.05) and the Time 2 English Math section (p
<.05). It is possible that other scores would have reached significance if we had larger groups,
further supporting the role that motivation and expectations have on literacy development.
Success in the United States
When SIFE were asked if they thought it would be easy to be successful in the United
States, 25% said “no”, 39% said “maybe” and 36% said “yes”. This means that 75% thought it
would be possible to be successful in the United States. The reasons they gave were that they
believed that if they worked hard, studied and were motivated there would be an abundance of
opportunities to succeed. The 25% who thought it would not be easy to be successful cited the
need for more skills in English as the primary obstacle in achieving success. There were no
!"#$%&'()*+%,-./'%""%
%
%
3I%
significant differences in scores at time 2 for the SIFE that thought it would be difficult to be
successful in the US compared to those who thought it would or may be easy to be successful.
5. Conclusions and Recommendations
Our two-year study of 98 SIFE revealed the following: While these students have typically
developing oral language skills in their native language, they are, in that same language, severely
under-developed in academic skills, in particular academic vocabulary and reading
comprehension. The skills that are in most need of reinforcement are academic vocabulary
beyond the 5th grade, and reading comprehension beyond the 3rd grade. Furthermore, this group
needs to develop higher-level text comprehension skills, such as inferencing and critical thinking
skills.
Math skills, as tested in the native language, also need to be strengthened but do not seem to be
dependent on language ability, as procedural and problem-solving skills were equally wanting in
this group. On average, they were 6 years behind expected grade level (grade 9 and 10).
After 2 years of English instruction, SIFE were still far behind regular ELLs receiving the same
type of instruction. This is most likely a consequence of SIFE (but not ELLs) lacking literacy
skills in their native language. SIFE thus have ‘triple the work’ needed for academic success:
1. Unlike other ELLs, they would benefit from furthering their native language literacy skills to
help develop L2 literacy; 2. Unlike other ELLs, they need to develop the background knowledge
prerequisites for learning grade level academic content 3. Like other ELLs, they must acquire L2
English language and literacy skills.
!"#$%&'()*+%,-./'%""%
%
%
3X%
Finally, school experience and academic/background knowledge and skills are severely limited
for these students. Based on the above, we make the following recommendations:
1. Provide an additional year of schooling, prior to secondary school, to “frontload” as many
skills and as much knowledge as possible in preparation for entrance into mainstream secondary
school classes.
2. Use native oral language skills to build academic language in the second language (L2)
English.
3. Focus on the learning of English language and literacy skills in all content area classes through
the use of native language support and specialized and differentiated instruction geared to meet
the needs of a very diverse student group.
4. Develop and implement a curriculum to build academic and literacy skills and background
knowledge and accelerate the learning needed for upper level school readiness; included in this
framework is a focus on critical thinking skills and the development of good academic and social
habits to help in the school and cultural adjustment process.
Given that even with one year of regular instruction, SIFE showed an average gain of 1.5 grades
in native language vocabulary and reading comprehension, and an average gain of 1 grade in
Math, we are confident that even greater gains can be made with the type of enhanced and
accelerated instruction proposed here.
!"#$%&'()*+%,-./'%""%
%
%
4H%
Bibliography
August, D., & Shanahan, T. (2006). Developing Literacy in Second Language Learners.
Mahwah: Lawrence Erlbaum Associates.
Chamot, A. U., Keatley, C., & J. Schiavone. (2000). Literacy Development in Adolescent
English Language Learners. Report to the Field-Initiated Studees Education Research
Grant Program, National Institute on the Education of At-Risk Students, US Department of
Education.
Cummins, J. (1981). The role of primary language development in promoting educational
success for language minority students. Los Angeles: California State Department of
Education, Office of Bilingual, Bicultural Education.
DeCapua, A., Smathers, W., & FrankTang, L. (2007). Schooling, Interrupted. Educational
Leadership , 40-46.
Genesee, F, Lindholm-Leary, K., Saunders, W. and D. Christian (2006). Educating English
Language Learners. New York: Cambridge University Press.
Goldenberg, C. (2008). Teaching English language learners: What the research does – and does
not –say. American Educator 32.2, Summer 2008, 8-43.
Kamil, M. (2003). Adolescents and literacy: Reading for the 21st century. Washington, DC:
Alliance for Excellent Education.
Klein, E. C. & Martohardjono, G. (2006). Understanding the student with interrupted formal
education (SIFE): A study of SIFE skills, needs and achievement. Report to the Office of
English Language Learners, New York City Department of Education (May, 2006)
Morse, A. (2005). A Look at Immigrant Youth: Prospects and Promising Practices. National
Conference of State Legislatures: Children's Policy Initiative.
National Center for Education Statistics. (2005). The Condition of Education. Washington, DC:
U.S. Government Printing Office.
New York City Department of Education. (2007). New York City’s English Language Learners:
Demographics and Performance. New York: New York City Department of Education.
Ruiz-de-Velasco, J., & Fix, M. (2000). Overlooked and Underserved: Immigrant Students in
U.S. Secondary Schools. Urban Institute.
Short, D., & Fitzsimmons, S. (2007). Double the Work. New York: Carnegie Corporation.
!"#$%&'()*+%,-./'%""%
%
%
47%
Short, D., Boyson, B., & Coltrane, B. (2003). Final National Conference for Educators of
Newcomer Students and Pilot Study on Newcomer Program Literacy and Assessment
Practices. Washington DC: Center for Applied Linguistics.
Slabin , R. & Cheung, A. (2005). A synthesis of research on language of reading instruction for
English Language Learners. Review of Educational Research 75, 247-281.
Toukamaa, T., & Skutnabb-Kangas, T. (1977). The intensive teaching of the mother tongue to
migrant children at pre-school age. Tampere, Finland: Department of Sociology,
University of Tampere.
!"#$%&'()*+%,-./'%""%
%
%
40%
Appendices
Appendix 1.: Development of ALLD
Appendix 2: Outline of ALLD to be distributed to NYC schools
Appendix 3: RISLUS Syntax Test
Appendix 4: Spanish ALLD Pre-test Grade Level Charts for SIFE
Appendix 5: English ALLD Grade Level Charts for SIFE
!"#$%&'()*+%,-./'%""%
%
%
41%
Appendix 1: Development of ALLD
As of the start of this study, there were no appropriate assessments for adolescent ELLs with
low native language literacy. We began our pilot study by consulting a number of existing
assessments. We paid close attention to the following concepts: 1) inclusion of assessments of
pre-literacy and literacy, and content areas of math, science and social science; 2) assessment of
skills from a range of grade levels; 8 3) equivalence of versions in Spanish and English (since we
planned on giving an English version of the diagnostic); 4) group administration, so that the New
York City school system could implement the assessment successfully with enough time and
resources for the testing of all SIFE entering the school system.
None met all of the above requirements. During the pilot, we administered the Batería III
(Riverside), an individually administered assessment which met all of the requirements except
for being group administered. We piloted this assessment with twelve SIFE students, and based
on their results, we reconsidered the other assessments we had consulted, looking for a group
assessment. We adapted a US standardized Spanish language test published by Harcourt, the
Aprenda Achievement Test Series, Third Edition (from this point forward, referred to as the
Aprenda III), into an appropriate assessment for SIFE. The Aprenda III met all of the original
requirements, except that it is not a cumulative assessment. The Aprenda III consists of a group
of tests designed for students in Kindergarten through grade twelve, with separate test booklets
for each grade level. In order to meet the criteria of a cumulative assessment, we took items from
the Aprenda III test booklets from grade levels one through seven and included them in the
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
I%;'5.A/'%)A*%*'/'.*5-%N).6%9/%+)%B'+'*O9:'%+-'%5-.*.5+'*9/+95/%)G%!"#$T%.BO9:9/+'*9:N%.:%.//'//O':+%G*)O%):'%
N*.B'%6'Y'6%Z)A6B%:)+%N9Y'%A/%+-'%9:G)*O.+9):%Z'%:''B%.Q)A+%!"#$M%"G%Z'%.BO9:9/+'*'B%.%:9:+-%N*.B'%6'Y'6%
.//'//O':+%.:B%+-'%/+AB':+/%B9B%:)+%(.//T%Z'%Z)A6B%):6>%=:)Z%+-'>%.*'%:)+%.+%+-'%:9:+-%N*.B'%6'Y'6T%QA+%Z'%Z)A6B%
:)+%=:)Z%.+%Z-95-%6'Y'6%+-'>%.*'%[9M'M%9+%5)A6B%Q'%'9N-+-%N*.B'%)*%/'5):B%N*.B'\M%
!"#$%&'()*+%,-./'%""%
%
%
42%
ALLD Pre-test. Our selection of grade levels to include was based on the results of our pilot
study.
During Year Two of the study the students were given the Spanish ALLD Post-test in order
to measure gains in native language skills over one year. We modified the ALLD Pre-test after
performing item analyses, omitting the items with the least efficacy. The Spanish ALLD Post-
test consists of many of the same test items as the pre-test, but includes fewer items from each
grade level and is reordered. The Spanish Post-test ALLD also has additional grade levels in the
literacy section. For the Post-test we did not administer the basic literacy section, but did
administer the reading vocabulary, reading comprehension, language and math sections.
The English ALLD is taken from the Stanford Achievement Test Series, Tenth Edition, which is
the English equivalent of the Aprenda III. As we administered the English ALLD to our
comparison groups whom we expected to perform well on the assessment, we included up to
grade eleven in the English ALLD.
!"#$%&'()*+%,-./'%""%
%
%
43%
Appendix 2: Outline of ALLD to be distributed to NYC schools
The version of the ALLD to be distributed to NYC schools is slightly different than the
versions of the ALLD we report on in this report. The ALLD to be distributed went through
more edits, with feedback from both the DOE and the publisher of the standardized tests which
the ALLD was drawn from (Harcourt, which is now Pearson). The outline below has all changes
from the original in brackets, and describes the English and Spanish ALLD.
Description of English and Spanish ALLD to be distributed to NYC schools:
Book One: Pre-literacy Section
[Similar to the Pre-literacy section of the Spanish ALLD Pre-test, but with fewer items.]
Book Two: Literacy and Math Sections:
Word Study: Grade level TK [Added in the Spanish ALLD; items written by our research
team]
Reading Vocabulary: Grade levels three through eleven [Added higher grade levels]
Reading Comprehension: Grade levels two through eleven [Added higher grade levels]
Language: Grade levels three and four
Math: Grade levels three through ten [Added higher grade levels]
!"#$%&'()*+%,-./'%""%
%
%
44%
Appendix 3: Syntax Test
a) Subject coordination: The bear and the dog chase the cat.
b) Relative clause: The bear who touches the dog dances.
c) Temporal adverbial clauses: After jumping, the dog kicks the cat.
d) Subjectless subordinate clauses: The cat tells the bear to jump.
!"#$%&'()*+%,-./'%""%
%
%
4J%
Appendix 4: ALLD Spanish Pre-test Grade Level Charts
Vocabulary Results: Percentage of Students Scoring at Each Grade Level
Reading Comprehension Results: Percentage of Students Scoring at Each Grade Level
!"#$%&'()*+%,-./'%""%
%
%
4I%
Language Results: Percentage of Students Scoring at Each Grade Level
Math Results: Percentage of Students Scoring at Each Grade Level
!"#$%&'()*+%,-./'%""%
%
%
4X%
Appendix 5: English ALLD Grade Level Charts for SIFE
English Reading Vocabulary Results: Percentage of SIFE Scoring at Each Grade Level
!"#$%&'()*+%,-./'%""%
%
%
JH%
English Reading Comprehension Results: Percentage of SIFE Scoring at Each Grade Level
... The use of ERP with auditorily presented stimuli is especially well suited for studying processing in the first-learned language of heritage speakers, as subjects are not required to provide any conscious response to the stimuli, which avoids possible confounds related to literacy and education. Heritage speakers typically do not receive formal schooling in the heritage language and often lack age-appropriate literacy skills in the heritage language (Beaudrie and Fairclough 2012; City of New York Department of Education 2011; Menken and Kleyn 2010;Valdés et al. 2006), which often results in poor performance on experimental tasks that require metalinguistic judgments about acceptability or grammaticality (Klein and Martohardjono 2009). ERP measures voltage changes at the scalp associated with the synchronous firing of brain cell populations in response to a stimulus and the ERP technique has been used successfully to study language processing in monolingual (Friederici 2002;Kutas and Hillyard 1980;Osterhout 1994), bilingual (Kotz 2009;Moreno, Rodríguez-Fornells, and Laine 2008), and learner populations (Osterhout et al. 2008). ...
... SD=6.82); were 18-50 years old at the time of testing (M=26.61, SD=7.30); and were raised speaking primarily Spanish until at least age 10 by caregivers who met the late bilingual criteria described above. 1 To ensure adequate levels of Spanish proficiency for our heritage speakers, who are typically English dominant, we tested their auditory comprehension of complex sentences using a computerized picture-pointing task: the RISLUS Multilingual Syntax Test (RMST) (Klein and Martohardjono 2009). All heritage speakers performed at ceiling on this test, indicating fluency in Spanish. ...
... Twenty of the participants were labeled SIFE; the remaining forty-two were ELLs. The New York City Department of Education (NYCDOE) designates students as SIFE if they had missed at least two years of formal schooling prior to arrival in the U.S. Due to missed schooling, SIFE have low levels of L1 academic skills (Klein and Martohardjono 2006). Newly arrived ELLs who were not labeled SIFE were also selected for the study; as these students had consistently attended school in their home countries, they were expected to have higher levels of L1 literacy than SIFE. ...
... This was an important step, as research on SIFE and ELLs has suggested that the designation based on missed schooling may not be sufficient to predict ageappropriate literacy skills. ELLs, who can also be under-schooled for their age, but who are not categorized as SIFE as they do not have interruptions in their schooling can also exhibit low literacy in the home language when appropriately tested (Klein and Martohardjono 2006). ...
Article
Full-text available
Previous research on the development of second language (L2) reading comprehension among adolescents and adults has looked at the relative importance of students’ second language (L2) linguistic skills and home language (L1) reading comprehension skills in the development of L2 reading comprehension, with many studies finding a more important role for L2 language proficiency. However, the previous research has been done largely in foreign language settings with students who have age-appropriate reading skills in the L1. We extend the question of the relative importance of L1 reading skills and L2 language proficiency to L2 reading comprehension to Spanish-speaking adolescent newcomer emergent bilinguals in New York City with low levels of L1 reading, as research on these students is lacking and little is known about their development of L2 academic skills. We find that for this population L1 reading comprehension plays a significant role in L2 reading comprehension equal to, if not greater than, L2 language proficiency. These findings suggest the need to include continued support of L1 reading skills in order for reading in the L2 to develop, at least for newcomer students with low literacy skills.
... Most studies on the Spanish of second-generation bilinguals in the U.S. use explicit and often written metalinguistic tasks, which introduce potential confounds since for the vast majority of this population, the language of instruction and literacy is English from elementary school through college (Beaudrie and Fairclough 2012, City of New York Department of Education 2011, Menken and Kleyn 2010, Valdés et al. 2006. Spanish-language instruments that require metalinguistic judgments of grammaticality, felicity, and logic thus confront this population with a highly unfamiliar, if not downright unnatural, task (Bialystok 1986, Bialystok and Ryan 1985, Birdsong 1989, and secondgeneration bilinguals tend to show hesitancy and indetermination when asked to make metalinguistic judgments about Spanish even when they show high proficiency and oral fluency in that language (Klein and Martohardjono 2009). ...
... SD=6.82); were 18-50 years old at the time of testing (M=26.61, SD=7.30); and were raised speaking primarily Spanish until at least age 10 by first-generation caregivers. 1 To ensure adequate levels of Spanish proficiency for our second-generation bilinguals, who are often English dominant, we tested their auditory comprehension of complex sentences using the RISLUS Multilingual Syntax Test (RMST): a computerized picture-pointing task that assesses comprehension of auditorily-presented complex sentences, which has been used in previous research on bilingual populations with limited Spanish literacy (Klein and Martohardjono 2009). We administered this assessment rather than standardized language proficiency tests because it does not rely on advanced reading or writing skills. ...
Article
Full-text available
Two types of Spanish-English bilinguals were tested in an event-related potential (ERP) experiment on a contrast in the two languages exemplified in (1) and (2) in order to investigate linguistic permeability during processing of Spanish (1a and 2a). In Spanish, but not English, absence of the complementizer que is ungrammatical. (1) a. Qué hermana confesó Inés que había comido la tarta? b. *What sister did Inés confess that had eaten the cake? (2) a. *Qué hermana confesó Inés Ø había comido la tarta? b. What sister did Inés confess Ø had eaten the cake? In a first analysis, we grouped subjects by generation and compared ERP responses to que-less vs. que-full sentences. A significant N400 effect was found for first-, but not second-generation, suggesting reduced sensitivity to missing que for the latter. However, a second analysis, using linear mixed modeling to test predictiveness of individual speaker variables revealed generation to be non-predictive of N400 amplitude. Instead, current language use, cumulative exposure to English, and socioeconomic status (SES) were significant predictors for all subjects: increased English use, exposure, and SES resulted in smaller N400 amplitude to the anomaly in Spanish shown in (2a). Our results show that a priori classification of bilinguals masks gradient cross-linguistic effects, and processing is permeable in all bilinguals depending on amount of language use. Results from an acceptability judgment task administered to the same subjects using a subset of the same stimuli show that both subject groups judge que-less and que-full to be equally natural. These results suggest that behavioral measures that rely on metalinguistic judgments may not be good indicators of processing, and that having to appeal to metalinguistic knowledge may mask intrinsic knowledge.
... Upon entering our schools, these newly arrived students bring their personal experiences, knowledge of their home languages and their personal abilities. They often have high aspirations to secure professional jobs, go to college and provide for their families (Klein & Martohardjono, 2006). In order to achieve these goals, they must negotiate life in an unfamiliar place, and compete with academic rigors such as high-stakes standardized tests, advanced-level homework assignments requiring academic language, all while acquiring English and gaining new content knowledge through second language instruction 1 (Menken, 2013). ...
... HS literature has noted that the L1 acquisition of HSs may be incomplete (Benmamoun et al., 2010;Scontras et al., 2015) or may undergo attrition due to disuse and lack of sustained exposure (Montrul, 2004;Polinsky, 2011). This bilingual population is also noted for its linguistic insecurity (Klein and Martohardjono, 2009). While a variety of studies exist that explore HS behavioral outcomes, virtually nothing is known about their psycholinguistic processing (as commented in Bolger and Zapata, 2011). ...
Thesis
Full-text available
Traditionally, heritage speakers are recognized as a heterogeneous group whose skills in their heritage language are unlike those of monolinguals or L2 learners of that language. Indeed, much evidence confirms the cognitive and linguistic uniqueness of this population. However, highly proficient heritage speakers may pattern more similarly to another bilingual population than typically assumed: first-generation late bilinguals. The present study examines group-level processing differences between Spanish heritage speakers and Spanish-English late bilinguals in English, the second-learned and current societal majority language of these populations. Dominance is also analyzed as a possible effect of group processing differences, since traditionally and definitionally it is a main factor in distinguishing these two bilingual groups. The robustly-attested processing asymmetry between subject-relative clauses and object-relative clauses is examined utilizing eye-tracking in the visual world paradigm to analyze bilingual processing in this study. Processing is measured through gaze fixation, and language dominance is operationalized using a novel relative fluency (RF) index. Results support that (1) These two populations do not pattern differently from each other in their accuracy, response time, or target fixation proportions throughout either RC condition, (2) Both groups are only showing mild evidence for a subject/object processing asymmetry, and (3) There are group differences in language dominance, but dominance has no main effect on gaze fixation.
... This degree of academic under-preparedness is similar to other groups of students such as SIFE (Students with Interrupted Formal Education). Even though participants in this study are not classified by the Department of Education as SIFE (SIFE are defined as having a documented gap of at least two years in their education [64]). So, while many participants in this study reported having a gap in their education of two years or longer, it was rarely officially documented, they share many characteristics of SIFE. ...
Article
Full-text available
Article
Full-text available
Muchos programas, en España o Europa, entienden al alumnado inmigrante como un colectivo homogéneo, aplicando soluciones iguales a todos, sin lograr promover su inclusión socio-educativa. Sin embargo, algunos países anglosajones observan diferencias en este grupo que influyen en su éxito educativo e inclusión social y cuyo conocimiento puede ayudar a promover actuaciones que atiendan mejor a esta población. El artículo describe y analiza el concepto de estudiante con escolaridad formal limitada o interrumpida y discute los factores que la escuela tiene en cuenta para fomentar su inclusión socio-educativa, abogando por una intervención holística que tenga en cuenta este alumnado necesita mucho más que centrar su aprendizaje únicamente el nuevo idioma, por ejemplo el apoyo social para superar los problemas socioemocionales que les acompañan. Resulta especialmente relevante para España y Europa, pues recientemente acuden cada vez más alumnos con estas características a nuestras escuelas.
Article
This article offers a critical review of research about emergent bilingual students in secondary school, where the academic demands placed upon them are great, and where instruction typically remains steadfast in its monolingualism. I focus on recent scholarship about the diversity within this student population, and center on ‘students with interrupted formal education’ (SIFE, new arrivals who have no home language literacy skills or are at the beginning stages of literacy learning) and ‘long-term English language learners’ (LTELLs, primarily educated in their receiving country yet still eligible for language support services). Little has been published about these students, making this a significant area of inquiry. Moreover, both groups are characterized by poor performance and together illustrate the characteristics of secondary students at various points along an academic language and literacy continuum. While existing research provides important information to help us improve secondary schooling for emergent bilinguals, it has also perpetuated deficit views of these students by focusing solely on their perceived academic shortcomings. Grounded in a new body of research in applied linguistics that examines the students’ complex, creative, and dynamic language and literacy practices, I apply a translanguaging lens to critique the positioning of such students as deficient, with implications for research and practice.
Article
Full-text available
This article reviews experimental studies comparing bilingual and English-only reading programs for English language learners. The review method is best-evidence synthesis, which uses a systematic literature search, quantification of outcomes as effect sizes, and extensive discussion of individual studies that meet inclusion standards. A total of 17 studies met the inclusion standards. Among 13 studies focusing on elementary reading for Spanish-dominant students, 9 favored bilingual approaches on English reading measures, and 4 found no differences, for a median effect size of +0.45. Weighted by sample size, an effect size of +0.33 was computed, which is significantly different from zero (p < .05). One of two studies of heritage languages (French and Choctaw) and two secondary studies favored bilingual approaches. The review concludes that although the number of high-quality studies is small, existing evidence favors bilingual approaches, especially paired bilingual strategies that teach reading in the native language and English at different times each day. However, further research using longitudinal, randomized designs is needed to determine how best to ensure reading success for all English language learners.
Book
Cambridge Core - Applied Linguistics - Educating English Language Learners - by Fred Genesee
  • A Decapua
  • W Smathers
  • L Franktang
DeCapua, A., Smathers, W., & FrankTang, L. (2007). Schooling, Interrupted. Educational Leadership, 40-46.
Understanding the student with interrupted formal education (SIFE): A study of SIFE skills, needs and achievement. Report to the Office of English Language Learners
  • E C Klein
  • G Martohardjono
Klein, E. C. & Martohardjono, G. (2006). Understanding the student with interrupted formal education (SIFE): A study of SIFE skills, needs and achievement. Report to the Office of English Language Learners, New York City Department of Education (May, 2006)
New York City's English Language Learners: Demographics and Performance
New York City Department of Education. (2007). New York City's English Language Learners: Demographics and Performance. New York: New York City Department of Education.
A Look at Immigrant Youth: Prospects and Promising Practices
  • A Morse
Morse, A. (2005). A Look at Immigrant Youth: Prospects and Promising Practices. National Conference of State Legislatures: Children's Policy Initiative.