Elicited imitation paradigm

[Evan Kidd]

Dear Liam, Brian, and list members,

I have conducted a fair bit of research using elicited imitation in first language acquisition, and have found it to be a very sensitive measure. I think it's important to use contrasting conditions controlled for word length (e.g., active versus passive) - finding (predicted) differences across conditions goes a long way to placate reviewers, since a mere "parrot effect" predicts children would perform equally across conditions. Systematic errors often show that children are comprehending the sentences themselves (e.g., I have found that children find it difficult to repeat some types of object relative clauses, but that their errors often suggest that that they interpreted the sentence correctly by assigning the correct thematic roles). 


Finally, you might like to check out an in press paper from my lab, which shows that elicited imitation predicts 4-6-year-old children's comprehension of non-canonical sentences (passives & object RCs) over and above both phonological short-term memory and complex working memory span. The ref is:


Boyle,W., Lindell, A., & Kidd, E. (in press). Investigating the role of verbal working memory in young children's sentence comprehension. Language Learning. 


It's available on Early View, but if you don't have access to the paper email me and I'll send a re-print.


Best,


Evan 

On 03/11/13, Brian MacWhinney  <macw at cmu.edu> wrote:
> 
> 
> 
> 
> Dear Liam,
>  I am referring to a study that my grad student Colleen Davy did as a part of her work on training fluency in Intermediate Spanish. There was already some commentary on this in earlier messages. As I mentioned there, questions the idea that elicited imitation involves mere parroting:
> Erlam, R. (2006). Elicited imitation as a measure of L2 implicit knowledge: An empirical validation study. Applied Linguistics, 27, 464-491. 
> 
> 
> The same issue was explored in:
> DeKeyser, R. (2003). Implicit and explicit learning. In C. Doughty & M. Long (Eds.), The Handbook of Second Language Acquisition (pp. 313-349). Oxford: Blackwell.
> 
> 
> 
> Basically, people in second language would like to be able to use elicited imitation as a measure of "implicit learning" by which they usually mean non-conscious and non-tutored learning. As in first language studies, these elicited imitation tests can include all sorts of structures, both grammatical and ungrammatical, so they are a great way of exploring this issue. However, people (i.e. reviewers) often object to the use of elicited imitation, because they think that somehow learners can just fill up their short-term memory buffer with a series of words in the second language and repeat those words without really understanding the grammatical structure at all. Never mind the fact that the same learners (both L1 and L2) can't repeat strings of 4-5 unrelated words in the second language. 
> 
> 
> To deal with this objection, Colleen contrasted learning under a variety of conditions. For the current issue, the important comparison was between basic sentence repetition (listen-repeat) in Study 1 and a method in Study 2 that required participants to produce the sentences without any immediately previous verbal stimulus. Instead, the participants (college students) learned the Spanish phrases that described certain pictures, such as el cocine la cena 'he cooks dinner' or ella sugiere 'she suggests' on the basis of clip art pictures. A combination of the two pictures would trigger the sentence "ella sugiere que el cocine la cena". In this way, practice on the sentence would occur without elicited imitation. For the purposes of the current discussion, the important result is that the two techniques led to parallel improvements in fluency. There are other results in the study in terms of the fluency-accuracy tradeoff and the effects of using small units vs. full sentences, but the important thing here is the parallel nature of the results across the methods. In the one method, it is clear that sentence planning is driven by meaningful content. Showing that the effects are the same in the other method tends to support the idea that that method also depends on processing of meaningful content. Of course, one could always say that the parallel nature of the outcome is an accident or due to other factors.
> 
> 
> Unfortunately, this study is under review, so you can't really use these results to defend the method. But you can point to Erlam, DeKeyser, and the articles from Lust et al. and McKee cited earlier.
> 
> 
> -Brian MacWhinney
> 
> 
> On Mar 10, 2013, at 3:34 PM, Liam Blything <liamblything at gmail.com> wrote:
> 
> 
> > Thank you Brian, Ana, Nan, and Roberto for your thoughts. If I follow the advice in those links, then imitation certainly seems to be worthwhile.
> > 
> > 
> > Brian, I would be interested if you are willing to talk a bit more about your forthcoming study that imitation does include "an aspect of comprehension or at least adequate parsing." This extra requirement in imitation tasks means that it might it be best to run an additional study where the comprehension aspect is removed. For example, an elicited production task with blocked presentations of each condition (e.g., in each blocked session, we emphasise through training trials that we want all sentences to use a specific connective/construction) is still designed to restrict what the child says, but has no comprehension element. I am aware of the limitations of a blocked design, but if the pattern remains the same for both paradigms then reviewers/critics would have little argument that my imitation conclusions are influenced by comprehension/parsing demands. 
> > 
> > 
> > Any thoughts are very welcome!
> > 
> > 
> > Liam Blything
> > Lancaster University PhD student.
> > 
> > 
> > On Friday, 8 March 2013 17:31:10 UTC, Liam Blything wrote:
> > > Hello all,
> > > 
> > >  
> > >  
> > > Can anyone give me some advice on what age-groups are too old for an elicited imitation paradigm? I have read that 4 years old is quite an old age to use this paradigm, but other studies have used it up to 7 years old.
> > > 
> > >  I am eliciting two clause sentences linked by a connective (before, after). When do children become at high-risk for simply 'parrotting' the target sentence?
> > > 
> > >  
> > > Many thanks for any thoughts,
> > > Liam Blything
> > > Lancaster University PhD student.
> > > 
> > > 
> > > 
> > > 
> > 
> > 
> >  -- 
> > 
> > You received this message because you are subscribed to the Google Groups "Info-CHILDES" group.
> > 
> > To unsubscribe from this group and stop receiving emails from it, send an email to info-childes+unsubscribe at googlegroups.com.
> > 
> > To post to this group, send email to info-childes at googlegroups.com.
> > 
> > To view this discussion on the web visit https://groups.google.com/d/msg/info-childes/-/Wjy_KxeNiCEJ.
> > 
> > For more options, visit https://groups.google.com/groups/opt_out.
> >  
> >  
> >  
> 
> 
> 
> 
>  
> 
> 
>  -- 
> 
> You received this message because you are subscribed to the Google Groups "Info-CHILDES" group.
> 
> To unsubscribe from this group and stop receiving emails from it, send an email to info-childes+unsubscribe at googlegroups.com.
> 
> To post to this group, send email to info-childes at googlegroups.com.
> 
> For more options, visit https://groups.google.com/groups/opt_out.
>  
>  
>  
> 

-- 
You received this message because you are subscribed to the Google Groups "Info-CHILDES" group.
To unsubscribe from this group and stop receiving emails from it, send an email to info-childes+unsubscribe at googlegroups.com.
To post to this group, send email to info-childes at googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/info-childes/attachments/20130311/47e5b53a/attachment.htm>