research proposal
ybilmz1314
EkiertdiGennaro2016AAAL.pdf
Focused Written Corrective Feedback:
What a Replication Study Reveals
About Linguistic Target Mastery Monika Ekiert, LaGuardia CC, City University of New York
Kristen di Gennaro, Pace University
The Debate
Truscott (1996). The case against grammar correction in
L2 writing classes.
Argued that corrective feedback regarding students’ grammar on writing
assignments was not only ineffective but potentially harmful.
Ferris (1999). The case for grammar correction in L2
writing classes: A response to Truscott.
Strongly objected to Truscott’s claims, stating that such claims are more
harmful to students than error correction.
The Debate
Truscott (1996). The case against grammar correction in
L2 writing classes.
Argued that corrective feedback regarding students’ grammar on writing
assignments was not only ineffective but potentially harmful.
Ferris (1999). The case for grammar correction in L2
writing classes: A response to Truscott.
Strongly objected to Truscott’s claims, stating that such claims are more
harmful to students than error correction.
The Debate
Corrective feedback (CF) remains the most contentious
issue in second language (L2) writing research.
Over 300 published papers have been produced on this
topic.
Research Perspectives
Writing researchers motivated by practical pedagogical
concerns If WCF is not effective (Truscott, 1996, 2007), then why should
teachers dedicate so many hours providing WCF to their
students?
If WCF is effective (Ferris, 1999, 2004), what are its effects?
Which is the most effective type of WCF?
Research Perspectives
Writing researchers motivated by practical pedagogical
concerns If WCF is not effective (Truscott, 1996, 2007), then why should
teachers dedicate so many hours providing WCF to their
students?
If WCF is effective (Ferris, 1999, 2004), what are its effects?
Which is the most effective type of WCF?
Researchers in the instructed SLA strand drawn to WCF
for its researchability (Ellis, 2010) CF is an area where theory and practice interface
WCF can be observed, measured, and controlled
The “article” studies
Effectiveness of WCF on accuracy of article usage
(Bitchner & Knoch, 2010; Ellis et al., 2008; Sheen, 2007)
Why articles?
Unavoidable
Noted difficulty across proficiency levels
Rule-governed uses
Referential indefinite a for first mentions
Referential definite the for subsequent mentions
Teachable
Observable
Measurable
Findings from “Article” Studies
In all “article” studies, treatment groups outperformed
the control groups -- evidence in favor of WCF.
Results suggest that WCF has a positive effect on
learners’ accuracy in using articles to express first
mention (a) and subsequent mention (the).
Results appear to contradict Truscott’s (2007) meta-
analysis finding that WCF has no effect, or a slightly
negative effect on learners’ accuracy.
Unresolved Problems: Linguistic Target
“Because there are occasions when the definite article is
required for referring to something for the first time … or for
referring to mass nouns, WCF was not provided on such
occasions” (Bitchener & Knoch, 2010, p. 202).
Unresolved Problems: Linguistic Target
“Because there are occasions when the definite article is
required for referring to something for the first time … or for
referring to mass nouns, WCF was not provided on such
occasions” (Bitchener & Knoch, 2010, p. 202).
There are exceptions to the “rule” students were learning.
Unresolved Problems: Linguistic Target
“… the use of obligatory occasion analysis … meant that the
students were not required to delete articles. … [O]ne effect
of the correction might have been to signal to learners that
they needed to use articles a lot and may have led to errors of
overuse. … It is possible that the correction led to overuse of
articles in contexts that were not the focus of this study and
that did not require the use of an article but this remains an
issue for further study” (Ellis et al., 2008, p. 369, footnote).
Unresolved Problems: Linguistic Target
“… the use of obligatory occasion analysis … meant that the
students were not required to delete articles. … [O]ne effect
of the correction might have been to signal to learners that
they needed to use articles a lot and may have led to errors of
overuse. … It is possible that the correction led to overuse of
articles in contexts that were not the focus of this study and
that did not require the use of an article but this remains an
issue for further study” (Ellis et al., 2008, p. 369, footnote).
Ignored overuse
Further research needed
The Current Study
Aims to fill this gap identified, but underreported, by
previous researchers
Accuracy, in our study, is defined …
in terms of how well an L2 user has learned to use an article
with regard to where it is and it is not required.
Research Questions
1. What is the impact of WCF selectively focused on two
article functions on learners’ accuracy with articles in
other contexts?
2. Do these effects change depending on the type of WCF?
Method
Quasi-experimental design (intact classes)
pre-test → immediate post-test→ delayed post-test
3 groups:
direct feedback group
direct feedback + metalinguistic explanation group
control group
2 types of instruments
Free production and controlled production
Design
Week 1 Weeks 3-5 Week 5 Week 11
Pre-test Treatment Immediate Delayed
x 3 Post-test Post-test
Participants
63 ESL students enrolled in a college-based, academic ESL
program (level low intermediate to intermediate)
3 intact writing classes (the same instructor)
Multiple L1s (Spanish, Chinese, Bengali, Tibetan, Nepali,
Urdu, Hindi, Greek, Creole, Korean, Polish, Arabic, Turkish,
Burmese, Pashtoo)
Group 1: Direct error correction on articles (n=22)
Group 2: Direct error correction and metalinguistic
explanation on articles (n=23)
Group 3: Control; received no corrective feedback on
article errors (n=18)
Focus of WCF
First and subsequent mentions requiring a and the
Jane bought a ring and a necklace for her mother’s
birthday.
Her mother liked the ring, but hated the necklace.
Treatment for Group 1 DF
Direct written error correction:
- incorrect uses with “a” or “the” were corrected
above each error
- “a” or “the” were inserted where they were omitted
but required
Treatment for Group 2 DF + ME
Written meta-linguistic explanation
Students received the following explanation attached
to their piece of writing:
Use “a” when referring to something for the first time.
Use “the” when referring to something that has already been
mentioned.
Illustration of the rule taken from each writing task
A man and a woman went to a restaurant for dinner. The man ordered a bottle of wine and the woman drank the wine.
Group 3 Control
Students received summary end notes on the overall
quality of their writing (Ferris, 2004, 2006)
No in-text corrections provided
No reference to article use made
Instruments
Designed to meet the following criteria:
written mode
narrative genre
connected discourse
Two types of written tests:
Picture description → free production
Missing word → controlled production
Instruments: Picture description
The accompanying narrative
story was handed to the
students with instructions to
read it silently.
Written stimulus (of
approximately 300 - 400
words) was replaced with the
pictorial stimulus and the
students were asked to write
the story themselves.
Participants were given 30
minutes.
6 forms developed.
Instruments: Missing word
Each narrative (200-300 words
long) was based on an adapted
Aesop fable.
Items were embedded in
sentences forming a coherent
text.
No blanks were provided.
Participants were instructed
to read the fable and insert
missing words wherever they
deemed it necessary–a task
resembling error correction.
Participants were given 20
minutes.
3 forms developed.
Procedures
Pre-test: picture description + missing word
Treatment:
Students received feedback on the picture description
narratives on three occasions (separated by a week)
Immediate post-test: picture description + missing word
Delayed post-test: picture description + missing word
Analysis of WCF
First and subsequent mentions of referents (treatment
focus)
first mentions requiring a (referential indefinites)
subsequent mentions requiring the (anaphoric definites)
BUT ALSO
Analysis of WCF
First and subsequent mentions of referents (treatment
focus)
first mentions requiring a (referential indefinites)
subsequent mentions requiring the (anaphoric definites)
BUT ALSO
First mention definites
Situational
Ex: Pass me the salt.
Cataphoric
Ex: The shade on this
lamp is really ugly.
Nonreferential indefinites
Ex: John is a plumber.
Idiomatic uses of
indefinites and definites
Ex: In a few minutes
Ex: In the meantime
Data Analysis
Omission and misuse were identified
All articles produced by students were coded by
article type (to identify article usage beyond the
treated articles)
Data Analysis
Accuracy Calculated by means of obligatory occasion analysis (the total
number of correctly supplied articles divided by the total number of obligatory occasions and expressed as proportions of 1).
Overuse Calculated by means of overuse occasion analysis (the total
number of overused articles divided by the total number of obligatory occasions and expressed as proportions of 1).
Scores on both accuracy and overuse analyzed with a series of mixed ANOVAs and post-hoc tests.
Overall impact of WCF on all articles
For all articles, there was a significant change over time
averaged across all groups.
Also, the effect of time varied among the groups
significantly. In other words, different groups developed
differentially over time.
At the immediate post-test, DF+ME and Control differed
significantly from each other.
Overall impact of WCF on all articles
Scores on all articles by group and time* (group x time*)
*
*
Impact on ‘treated’ vs. ‘untreated’ articles
For ‘treated’ articles, there was a significant change over
time averaged across all groups.
For ‘untreated’ articles, the effect of time varied among
the groups significantly. In other words, the three groups
developed differentially over time.
At the immediate post-test, DF+ME different significantly
from Control and DF groups on ‘untreated’ articles.
Impact on ‘treated’ vs. ‘untreated’ articles
Scores on “treated”
articles (time*)
*
Impact on ‘treated’ vs. ‘untreated’ articles
Scores on “treated”
articles (time*)
Scores on “untreated”
articles (group x time*)
*
Results on accuracy for each group
Direct Feedback overtime
Results on accuracy for each group
Direct Feedback + Metalinguistic Explanation overtime
Results on accuracy for each group
Control overtime
Results on article overuse
Scores by group and time
Summary of Results
The control group outperformed or matched the two experimental groups for accuracy on all articles, both first- and subsequent-mention uses and other article uses.
WCF focusing on only two functions of the article system inadvertently impacted the remaining functions of the system.
The impact appears to be negative in that, while improving on the “treated” features, the L2 learners experienced loss of accuracy on the “untreated” target features.
The provision of partial metalinguistic information may lead to overuse of a given structure.
Discussion
In the instructed SLA research, target feature selection
deserves an open and honest discussion.
Target feature meanings in relation to their learnability
are rarely considered.; “researchability” is not helpful
here.
Select feature uses, often driven by simple rules, are
targeted by WCF and FonF studies limiting the findings’
generalizability.
Discussion
On a positive note …
Having students engage in writing tasks in which certain
grammatical structures arise naturally and frequently may
be both necessary and sufficient to improve L2 learners’
performance with those structures.
Discussion
Is Truscott right?
How beneficial is WCF if it leads to greater accuracy is
some areas, but greater inaccuracy in other areas?
Writing instructors may need to adjust their expectations
regarding students’ improvement in grammatical accuracy,
including forms that have been corrected and taught; they
may need to be alert to potential overgeneralizations.
Q & A
Thank you!
