How idiosyncratic are 2-3 year olds' vocabularies?

[Brian MacWhinney]

Dear Barbara,

    What a shame.  Betty had written me, very much out of the blue, less than a year ago saying that she would be sending the data soon.  If the files are still available, it would seem to be our duty to do whatever we can to preserve the data and to anonymize it, in accord with what she wanted.  

Best regards,

-- Brian MacWhinney

On Nov 20, 2012, at 9:39 PM, Barbara Pearson <bpearson at research.umass.edu> wrote:

> Dear Brian,
> 
> I read last month that Betty Hart passed away.
> 
> If you'll recall, the data were on 5 1/2 inch floppies.  *IF* someone from her lab could make them available, could anyone read them?
> The hold-up was anonymizing them.  I don't think she ever got to it.
> 
> Would have been nice.  I'm happy at least that she put as much information as she did in the appendices of her 1999 book--but it's not transcripts!
> 
> Cheers,
> Barbara
> 
> On Nov 20, 2012, at 3:15 PM, Brian MacWhinney wrote:
> 
>> Dear Info-CHILDES,
>> 
>> Linda is right.  You need to compare apples with apples.  You can basically do the first part of this already in CLAN.  Using the  /lib/ne32 folder for the example, the command is
>> 
>> freq +d2 +t*CHI *.cha
>> 
>> This then produces stat.frq.xls as an Excel spreadsheet output.  The next step would be to write a program in either Excel or better R to do all the comparisons and statistics across the children.  
>> 
>> We don't really want to do this in a pre-canned fashion, because people are going to have very different ideas about the shape of the comparison and the comparison set.  However, if people are interested and can tightly specify the shape of the analysis, then I think John Kowalski, who works with me on SLA stuff could probably write the relevant R script.  It is possible that the version Ping is specifying might be the best, but the output from that might be more than people really want.  
>> 
>> By the way,  every few years I get a note from Betty Hart saying that she will soon send me the transcripts from her study.  It would be a lovely corpus to have.
>> 
>> -- Brian MacWhinney
>> 
>> On Nov 20, 2012, at 3:58 PM, Linda Smith <smith4 at indiana.edu> wrote:
>> 
>>> Hi,
>>> 
>>> This is a very interesting question.  In an unpublished study (for reasons noted below), Shohei Hidaka (a former post doc) and I looked at the MCDI productive vocabularies for over 100 children from 18 to 30 months.  We asked how well any one word predicted knowing another word --across children or groups of children --at any point in time.  It is like doing a big Chi square.  Does, for example,  --across children --knowing "milk" predicts already knowing "dog"  (not because they are semantically related but because of regularities in the acquisition and ordering of words).  There were differences among groups of children (slower versus more rapid learners), and also hints of changes with age --converging to being all statistically alike with age and more different young. 
>>> 
>>> BUT the problem is the measure.  The MCDI was specifically made --and constrained to --the words most children acquired (had to be produced by 50% of children at 30 months in original normative sample to be included). This was done so that it would be useful as a relative measure, but is minimizes the idioscyncracies and make make the hard to find.  Diary studies or recordings might be better.  
>>> 
>>> Anyway, I think this is something the field really needs to answer.
>>> 
>>> Linda Smith 
>>> 
>>> On Monday, November 19, 2012 6:47:46 PM UTC-5, Elena Nicoladis wrote:
>>> Dear colleagues,
>>> I'm trying to find some sort of quantification of how idiosyncratic children aged 2-3 years are in terms of vocabulary. Has anyone, for example, attempted something as foolhardy as calculating the average percentage of words shared by all/most/some preschoolers?? 
>>> 
>>> Any leads would be much appreciated.
>>> Elena
>>> 
>>> -- 
>>> Elena Nicoladis
>>> 
>>> "Since all the sciences, and especially psychology, are still immersed in such tremendous realms of the uncertain and the unknown, the best that any individual scientist, especially any psychologist, can do seems to be to follow his own gleam and his own bent, however inadequate they may be. In fact, I suppose that actually this is what we all do. In the end, the only sure criterion is to have fun."
>>> Edward Tolman
>>> 
>>> 
>>> -- 
>>> You received this message because you are subscribed to the Google Groups "Info-CHILDES" group.
>>> To post to this group, send email to info-childes at googlegroups.com.
>>> To unsubscribe from this group, send email to info-childes+unsubscribe at googlegroups.com.
>>> To view this discussion on the web visit https://groups.google.com/d/msg/info-childes/-/LLTUPfu4380J.
>>> 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 post to this group, send email to info-childes at googlegroups.com.
>> To unsubscribe from this group, send email to info-childes+unsubscribe 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 post to this group, send email to info-childes at googlegroups.com.
> To unsubscribe from this group, send email to info-childes+unsubscribe 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 post to this group, send email to info-childes at googlegroups.com.
To unsubscribe from this group, send email to info-childes+unsubscribe 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/20121120/85540b94/attachment.htm>