Objective
Framed by the current socio-political context of school
reform in the USA, particularly in Massachusetts, this pilot
project consisted of the collection of multimedia longitudinal
records of texts produced by English Language Learners (ELLs)
in Amy Rivera's fourth grade class. Amy Rivera received
her Master’s Degree in Education through ACCELA and
has participated in a number of studies of her classroom
practices (see ACCELA research). The data that comprise
this corpus were collected over the course of an academic
year as a way of documenting how ELLs’ abilities to
produce academic language change over time in a more fine-grained
manner than other measures of student learning typically
afford (e.g., MCAS data).
Corpus Description
This small-specialized multimedia corpus encompasses the
texts produced by five ELLs during three curriculum units
in their English Language Arts class. Instruction focused
on the reading and writing of personal narratives over the
course of an academic year. It is small corpus because it
consists of the texts produced by five students at three
times of data collection. It is specialized because it focuses
on one genre, personal written narratives. It is multimedia
because it includes the scanned and transcribed versions,
as well as the sound or video recorded readings, of the
students’ original texts. It is longitudinal because
it tries to capture the development of the five focal ELL
students’ language use over time. In addition to the
students’ texts, all contextual information regarding
the teaching of this genre and the students’ processes
of interpretation, production, and distribution of the texts,
was ethnographically analyzed through the use of class video
recordings, field notes, and teacher and students’
interviews.
In building The ACCELA 4GWNEL Learner Corpus, I collected,
transcribed and digitally scanned approximately 20 ELLs’
handwritten final texts, at three different times (October,
February and May) over one academic year. For the transcription
of the texts, I took into consideration some of the recommendations
made by the researchers in the Lancaster-Leverhulme Corpus
of Children’s Writing (LCCW) (external link to http://bowland-files.lancs.ac.uk/monkey/lever/lccw.htm).
Besides the scanned and transcribed versions of the texts,
I also used sound or video recordings of students’
readings of their texts, thus making this one of the first
multimedia corpus of elementary school-aged English Language
Learners compiled.
Multimodal Features
• Written: a raw longitudinal collection of the transcribed
version of final drafts of students written narratives
• Visual: The scanned images of original handwritten
texts
• Spoken: The transcribed versions of the texts read
aloud by the children authors
• Audio: The recordings of the texts being read by
the children authors
• Audiovisual: video clips of contexts of production,
interpretation and distribution of texts within the students’
literacy practices
Explicit Design Criteria and Documentation
Granger, a leading expert in the field of computer learner
corpora, maintains that “learner corpora should be
compiled according to strict design criteria, some of which
are the same as for native corpora (see Atkins & Clear,
1992), while others, relating to both the learner and the
task, are specific to learner corpora” (Granger, 2002,
p.9). Drawing on Granger’s learner corpus design criteria,
I documented the following variables in my design of this
corpus:
Learner Variables
• Age
• Learning context: 4th Grade English language Arts
• Mother tongue: Spanish (Puerto Rico)
Task Variables
• Writing type: Free writing
• Time limit: 1-4 weeks projects
• Use of reference tools: dictionaries, published
written stories, locally produced stories
• Audience: teacher, peers, parents
I completed the text preparation phase of the project and
am currently analyzing and assembling case study data into
hypermedia (HTML) documents (see below). Once this analysis
is completed, the corpus and derived materials will be available
to the ACCELA teachers, students, administrators and researchers
as well as to a wider research community via the Internet.
Please visit this site to see forthcoming updates on this
project.
Further information
For more information on the ACCELA 4GNEL Learner Corpus,
including details of availability, please contact Pablo
Jimenez-Caicedo (Jimenez@educ.umass.edu) at The University
of Massachusetts, Amherst.
|
