Investigating the characteristics of effective recasts in the ESL classroom

1
Investigating the characteristics of effective
recasts in the ESL classroom

• Shawn Loewen Jenefer Philp
• TBLT Conference

2
Recasts

• Target-like reformulations following a learners
  non-TL utterance.
• Maintains the central meaning, while changing
  elements of the form (lexical, morphological,
  syntactic, phonological components). (Long, 1996)

3
Example of Recast

• S when I was soldier I used to wear the
  balaclava
• T and why did you wear it S for protection
  from the cold or for another reason
• S just wind uh protection to wind and cold
• T protection from
• S uh from wind and cold
• T right, okay not for a disguise

4
Potential benefits of recasts

• Provide implicit negative feedback
• Juxtapose the learners incorrect utterance with
  the target-like version
• Contingent on the learners production
• Incidental

5
Potential limitations of recasts

• Ambiguous
• Learners are not pushed in their output
• Differentially beneficial depending on form
• Other forms of feedback may be more beneficial

6
Research rationale

• Studies of recasts have not always worked from
  the same definition.
• Characteristics of recasts may vary, and these
  variations may influence their effectiveness.

7
Research Questions

• What are the characteristics of recasts in young
  adult ESL classes?
• Are particular characteristics of recasts
  associated with successful uptake/accurate test
  scores?

8
Participants

• Private language school in Auckland, New Zealand
• 12 Teachers
• Native speakers of English
• 8 male, 4 female
• 118 Students
• Intermediate English proficiency
• Primarily from Korea, China and Japan

9
Procedure

• 17 hours of audio-recorded classroom observations
• FFEs and Recasts identified and coded
• Individualised tests designed and administered
  (Immediate and Delayed)

10
Characteristics of recasts

• FFEs with recasts were coded for the following
  features

11
Linguistic Focus

• Lexical
• Morphological/Syntactic
• Phonological
• Combination

12
Prosodic emphasis

• Stressed or Unstressed
• H some people have racism
• T some people ARE racist
• H are racist

13
Intonation

• Declarative or Interrogative
• S somebody steal my paper (.) stolen
• T someone stole your paper?

14
Length of Recast

• More or less than 5 morphemes
• S my brain will s- I seem will seem to be
  explode
• T my my brain seems to be exploding

15
Segmentation

• Segmented or Whole
• S why he why should he release?
• T be released

16
Number of Changes

• 1 Change or 2 Changes
• S damag-ed (.) wall damag-ed
• T the hotel wall was damaged

17
Number of Feedback Moves

• 1 or 2
• S the title of the story is girl had blood in
  her scalp
• T blood?
• S bloot
• T bullet bullet
• Sbullet bullet in her scalp

18
Successful Uptake

• Students incorporate the linguistic form into
  their own production
• S maybe everyday have a lot of people die by the
  cigarette
• T die from cigarettes
• S die from cigarettes, so maybe

19
Testing Procedure

• Students were tested orally on the linguistic
  forms targeted in FFEs in which students
  themselves were involved.

20
TEST Correction Example

• The following sentences are incorrect or
  inappropriate. Please listen and tell me how you
  could make the sentences better.
•  
• 1. I used to wear the balaclava for protection to
  wind and cold.

21
Test Pronunciation

• Learners were asked to first read aloud a
  sentence containing the targeted word/phrase and
  then to read aloud the target word/phrase in
  isolation.

22
Test Scores

• Responses to test items were scored as correct or
  incorrect according to the response provided in
  the recast/FFE

23
Results

• 465 FFEs
• 228 Recasts

24
Linguistic Focus
25
Length of Recast
26
Prosodic Emphasis
27
Segmentation
28
Number of Changes
29
Number of Feedback Moves
30
Intonation
31
Predictors of Successful uptake
32
Predictors of Successful Uptake

• Stressed recasts
• Declarative recasts
• Recast with only one change
• FFEs with more than one feedback move

33
Predictors of Accurate Test Scores
34
Predictors of Accurate Test Scores

• Interrogative Recasts
• Recasts with fewer than 5 morphemes
• Recasts with only one change

35
Summary

• Recasts in these classes were generally short,
  stressed, segmented, declarative with only one
  change.
• The same characteristics did not predict both
  successful uptake and correct test scores.