sanz morgan short04 by 0L4UyZ65


									                                     Sanz, C., & Morgan-Short, K. (2004).
              Positive Evidence Versus Explicit Rule Presentation and Explicit Negative Feedback:
                          A Computer-Assisted Study. Language Learning, 54(1), 35-78.

                        SLARG Week 26 (27/05/04) – Summary prepared by Karen Roehr

o It is not clear whether exposure to input is sufficient for successful SLA. Researchers, teachers and learners
   generally assume that some type of additional form-focused instruction will be beneficial.
o In the majority of cases, findings from laboratory-based and classroom-based studies comparing different
   instructional conditions suggest that both explicit rule explanation and explicit feedback have facilitative
o Conversely, studies using the paradigm of processing instruction (PI – VanPatten) indicate that structured
   input activities play a major role and may be sufficient for learning; explicit information may not be needed.
o Thus, practice activities alone may result in acquisition, provided that attending to the targeted linguistic
   feature is made an essential component of the tasks students carry out.
o Few studies to date have tried to disentangle the effects of explicit information given before practice (a priori
   rule explanation) versus explicit information given during practice (explicit negative feedback).

Null hypothesis
o Providing learners with explicit information about the targeted L2 feature either before or during exposure to
   input-based practice or both will not affect learners' ability to interpret and produce L2 sentences containing
   the targeted L2 feature.

Design of the study
o 69 participants; L1 English learners of L2 Spanish at a U.S. university
o Targeted feature: word order of Spanish sentences containing preverbal direct object pronouns
o a typical learner mistake is to interpret sentences according to the first-noun strategy (SVO) and to ignore
   more reliable morphological markers which indicate that the first NP in the sentence is not the subject
o e.g. Lo besa la chica ('Him kisses the girl') is incorrectly interpreted as 'He kisses the girl'
o Pre-test and immediate post-test consisting of three measures:
o sentence interpretation: Which sentence matches the picture? (2-way multiple-choice)
o sentence completion: Complete the sentence so it matches the picture sequence.
o production of a written text: Re-tell a silent video story in as much detail as possible.
o Experimental treatment that follows PI guidelines:
o exposure to oral and written input at sentence and text level by means of structured practice activities
o computer-based administration of 56 practice items which require learners to react to the input (duration: 1
o computer program gives implicit feedback to all learners: correct responses are accepted; incorrect responses
   require the learner to try again
o manipulation through presence or absence of metalinguistic information given prior to the task (explanation)
   or during task completion (feedback)
o explanation (E) = explicit grammar instruction and information on the correct strategy to apply the grammar
   rule as well as the result of applying the wrong strategy
o feedback (F) = given immediately after a learner makes a mistake during practice; focused on the targeted
 four experimental conditions: + E / + F; + E / - F; - E / + F; - E / - F

Data analysis and results
o ANOVAs with 3 independent variables (explanation, feedback, time) and 3 dependent variables
   (performance on each of the three test components)
o significant differences between pre-test and post-test performances are found for all experimental conditions
o the explanation and feedback conditions had no significant effects, either individually or in combination
o thus, it made no difference whether explicit information was provided once, twice, or not at all, or whether it
   was provided before (E) or during practice (F)
o all significant effects can be attributed to time, i.e. exposure to the structured input activities alone
o the null hypothesis is supported

o Explicit information about the targeted linguistic feature did not enhance acquisition.
o This seems to contradict previous research which found beneficial effects for explicit instruction.
o The key may be the input provided: In the current study, participants had to respond to each item by
   processing the targeted form for meaning.
o Linking form and meaning was therefore task-essential, i.e. an inbuilt characteristic of the practice activities.
o In case of structured input and task-essential practice, supplementary information about the target form does
   not appear to improve acquisition and may therefore be unnecessary.
o Enhanced positive evidence which makes the targeted feature more frequent and salient appears to be
o However, as implicit learning is slow, a minimum number of practice items may be needed (56 in the current
   study vs. 18 in Rosa (1999), which resulted in different findings – p.70)

Limitations / improvements:
o It is not clear if practice alone was responsible for acquisition because the role of implicit feedback was not
   studied independently.
o In the interpretation task in particular, the number of possible responses was highly constrained, which may
   have facilitated learning.
o No delayed post-test was given, so no information is available about long-term retention.
o Verbal protocols would be useful to investigate learners' focus of attention and their thought processes.

To top