Transcript PowerPoint - Lancaster University

Tasks – with and without
corrective feedback.
Rosemary Erlam
The University of Auckland
[email protected]
Shawn Loewen
Michigan State University
acknowledging my co-researcher
• Shawn
Loewen
• Michigan
State
University
The night/date of your dreams
• Who did you go
out with?
• Where did you
go?
• What did you
do?
• What time did
you come home?
Etc etc
feedback
•
•
•
•
S: then we fall in love
T: we fell in love?
S: we fell in love
[implicit feedback]
• S: I go out with Clive
Owen
• T: I go out? I went
out
• S: I went out with
Clive Owen
• [explicit feedback]
vs no feedback
What do you predict?
What does the research literature
say?
• Three recent meta-analyses
• 1. Russell & Spada (2006)
• 15 studies investigating the effectiveness of oral and
written feedback
• effect sizes large, although smaller for oral than written
feedback
• 2/ Li (in press)
• 33 studies examining the effectiveness of corrective
feedback (following errors in oral production) in second
language learning
• medium effect for corrective feedback maintained over
time
• lab-based studies show a greater effect than classroombased ones
• shorter treatments generated a greater effect than longer
3. Mackey & Goo (2007)
• meta analysis of research on interaction
• 16 examined effectiveness of corrective
feedback given to learners during oral
interaction
• large effect sizes on all post tests
• need studies with delayed post tests
• interaction with feedback may not be more
effective than interaction alone
More research needed . . .
• effectiveness of feedback needs to be
investigated in relation to different target
structures (Ellis, 2007)
• range of measures of learning need to be
used– measures of implicit as well as
explicit knowledge (Ellis, 2007)
• studies that include delayed post tests
(Mackey & Goo, 2007)
Research questions
• Do learners completing tasks make gains
in implicit language knowledge when they
are given feedback targeting specific
language errors?
• Do they make greater gains than students
who complete the same tasks but get no
feedback?
Research questions
• Do learners completing tasks make gains
in explicit language knowledge when they
are given feedback targeting specific
language errors?
• Do they make greater gains than students
who complete the same tasks but get no
feedback?
Participants
• 50 students of L2 French from an
American university
• 32 in Year 2, 18 in Year 3
• average age 20
• 40 female, 10 male!!
• all but one had English as L1
• 60% of Year 2 & 80% of Year 3 had spent
time in a French speaking country –
average 5 months
Research design
• Pre-test
• Participants completed 8 tasks designed
to elicit the target structures
• 2 sessions – half an hour targeting each
target structure – 2 hours in total
• 40 students in feedback group, 10 in no
feedback
• Posttest 1 – 1 day later
• Posttest 2 – 3 weeks later
noun adjective agreement
• Un arbre (masculine)
• Un arbre vert
• Une voiture (feminine)
• Une voiture verte
• low perceptual salience &
low communicative value
• unacquired by classroom
learners despite
frequency in the input
(Harley, 1989)
Use of être with intransitive verbs in
the passé composé
• passé composé –
auxiliary + verb
• for most verbs auxiliary is
avoir
• for reflexive verbs & small
no of intransitive verbs
auxiliary is être
– J’ai fait du cheval
– Je suis monté sur l’échelle
• differs in non obvious
ways from L1
• does not carry a heavy
communicative load
Research design cont.
• Year 2 students (n = 22)
• Worked at tasks eliciting
noun/adj agreement and
use of être in passé
composé (2 hours)
• Year 3 students ( n = 18)
• Worked at tasks eliciting
noun/adj agreement (1
hour)
• Both received feedback
• No feedback – Yr 2
students (n = 10)
• Worked at tasks designed
to elicit noun/adjective
agreement and use of
être in passé composé (2
hours)
• Received no feedback
Tasks . . .
•
•
•
•
•
•
Les personnages de tele
Comment est-elle/il?
âge
taille
physique
caractère
•
•
•
•
•
La soirée de vos rêves?
sortis avec qui?
ou allé?
fait quoi?
rentré à quelle heure etc?
Feedback
• Implicit
• S: je pense elle n’est pas
intelligent parce qu’elle
• R: elle n’est pas intelligente?
• S: elle n’est pas intelligente
• S: je ne sais à quelle heure
nous avons rentrés
• R: nous sommes rentrés?
• S: nous sommes rentrés parce
que Espagne est un autre
continent
• Explicit
• S: elle est heureux
• R: elle est heureux? Elle
est heureuse
• S: heureuse
• S: ils ont allé
• R: ils ont allé? Ils sont
allés
• S: ils sont allés, oui, ils
sont allés au café
Feedback . . .
• directed at individual students but tasks
designed to optimize likelihood that all
students attend to corrective episodes
• Groups received average of
19 instances of feedback for noun/adj
agreement (range 8 – 32)
10 instances of feedback for être in
passé composé (range 3 – 16)
instruments
• Implicit language
knowledge
• Elicited imitation test
(Erlam, 2006; Ellis, 2005)
Les petites filles rêvent de se marier
en robe blanche.
• Spontaneous
production test
Décrivez la Princesse Diana et la
Mère Thérèse. Vous avez la
possibilité de passer une soirée
avec l’une d’elles. Laquelle
choisissez-vous? Pourquoi?
• Explicit language
knowledge
• Untimed grammaticality
judgment test
• Ungrammatical
sentences only (Ellis,
2004; 2005)
C’est une idée faux.
Research questions
• Do learners completing tasks make gains
in implicit language knowledge when they
are given feedback targeting specific
language errors?
• Do they make greater gains than students
who complete the same tasks but get no
feedback?
Elicited imitation test
noun adjective agreement
1
0.9
0.8
0.7
0.6
feedback
0.5
no feedback
0.4
0.3
0.2
0.1
0
1
2
3
Elicited imitation test
verb etre
1
0.9
0.8
0.7
0.6
feedback
0.5
no feedback
0.4
0.3
0.2
0.1
0
1
2
3
Spontaneous production test
noun adjective accuracy
100
90
80
70
60
feedback
50
no feedback
40
30
20
10
0
1
2
3
Spontaneous production test
verb etre accuracy
100
90
80
70
60
feedback
50
no feedback
40
30
20
10
0
1
2
3
Research questions
• Do learners completing tasks make gains
in explicit language knowledge when they
are given feedback targeting specific
language errors?
• Do they make greater gains than students
who complete the same tasks but get no
feedback?
Grammaticality judgment test
noun adjective agreement
1
0.9
0.8
0.7
0.6
feedback
0.5
no feedback
0.4
0.3
0.2
0.1
0
1
2
3
Grammaticality judgment test
verb etre
1
0.9
0.8
0.7
0.6
feedback
0.5
no feedback
0.4
0.3
0.2
0.1
0
1
2
3
Research questions
• Do learners completing tasks make gains
in implicit language knowledge when they
are given feedback targeting specific
language errors? Yes/No
• Do they make greater gains than students
who complete the same tasks but get no
feedback? No
Research questions cont
• Do learners completing tasks make gains
in explicit language knowledge when they
are given feedback targeting specific
language errors? Yes
• Do they make greater gains than students
who complete the same tasks but get no
feedback? Yes – for noun/adjective
agreement, No for être
Conclusions . . .explanations
• feedback facilitated learning
• But tasks (designed to elicit target structures)
also resulted in learning
Why?
• may have focused learner’s attention briefly on
form whilst engaged in communication of
meaning
• may have noticed gaps between their own
interlanguage resources & language they
needed
Vocabulary prompt
• sortir
• aller
• rentrer
• Etc
• [use of verb être in the passé composé is
a rule that is easy to apply
• allows for item learning rather than system
learning]
Reasons cont.
• No Feedback group reported high
awareness of target structures
• opportunity to engage in a different type of
instruction may have motivated them more
to attend to the content of the activities
(Lyster & Mori, 2006; Yang & Lyster,
forthcoming)
Awareness of target structure . . .
Noun/adj
être
n
%
n
%
feedback
40
64
22
38
no
feedback
10
44
10
67
What is missing?
• A control group that completed
the tests only
• [no tasks]
Control group – EI test
1
0.9
0.8
0.7
0.6
noun adj agreement
0.5
verb etre
0.4
0.3
0.2
0.1
0
1
2
3
Control group – GJT test
1
0.9
0.8
0.7
0.6
noun adjective agreement
0.5
verb etre
0.4
0.3
0.2
0.1
0
1
2
3
references
•
•
•
•
•
•
•
Erlam, R. (2006). Elicited imitation as a measure of L2 implicit knowledge: An
empirical validation study. Applied Linguistics, 27(3), 464-491.
Ellis, R. (2007). The differential effects of corrective feedback on two grammatical
structures. In A. Mackey (Ed.), Conversational interaction in second
language acquisition (pp. 339-361). Oxford: Oxford University Press.
Harley, B. (1989). Functional grammar in French immersion: A classroom experiment. Applied
Linguistics, 10, 331-359.
Li, S. (in press). The effectiveness of corrective feedback in SLA: A meta-analysis. Language
Learning.
Lyster, R., & Mori, H. (2006). Interactional feedback and instructional
counterbalance. Studies in Second Language Acquisition, 28, 269-300.
Mackey, A., & Goo, J. (2007). Interaction research in SLA: A meta-analysis and
research synthesis. In A. Mackey (Ed.), Conversational interaction in second
language acquisition (pp. 407-453). Oxford: Oxford University Press.
Russell, J., & Spada, N. (2006). The effectiveness of corrective feedback for the
acquisition of L2 grammar. In J. M. Norris & L. Ortega (Eds.), Synthesizing
research on language learning and teaching (pp. 133-164 ). Amsterdam: John Benjamins.