MATTR

Wed Aug 19 23:24:54 UTC 2015

Dear Nan,
    I think 10 is a reasonable window for the average.  Taking too small a window could lead to instability in the repeated samples, because there would not be enough diversity.  A larger window would be okay, but only if you have a large sample size.  The important idea underlying MATTR is that you are using a standard window throughout in order to get around the sample size problem.  In Covington’s terms, you are “cutting the Gordian knot”.

—Brian MacWhinney
  
> On Aug 19, 2015, at 6:52 PM, Nan Bernstein Ratner <nratner at umd.edu> wrote:
> 
> I just tried MATTR and it does output in spreadsheet; you simply use the +d2 option in addition to the +bN option.
> My question is about the value to use for N; you use 10, but is there a rhyme or reason to picking a window size?
> N
>  
>  
> Nan Bernstein Ratner, Fellow, ASHA; Fellow, AAAS
> Board Certified Specialist-Child Language
> Professor
> Department of Hearing and Speech Sciences
> University of Maryland
> 0100 Lefrak Hall
> College Park, MD 20742
> 301-405-4217, 301-314-2023 (fax)
> nratner at umd.edu <mailto:nratner at umd.edu>
>  
> http://hesp.umd.edu/facultyprofile/Bernstein%20Ratner/Nan <http://hesp.umd.edu/facultyprofile/Bernstein%20Ratner/Nan>
>  
> Director, University of Maryland Autism Research Consortium (UMARC)
> Participating Faculty, Neuroscience and Cognitive Science (NACS)
> Research Faculty, Maryland Language Science Center
> Participating Faculty, Developmental Science
>  
> From: chibolts at googlegroups.com <mailto:chibolts at googlegroups.com> [mailto:chibolts at googlegroups.com <mailto:chibolts at googlegroups.com>] On Behalf Of Amber
> Sent: Wednesday, August 19, 2015 5:23 PM
> To: chibolts <chibolts at googlegroups.com <mailto:chibolts at googlegroups.com>>
> Subject: Re: VOCD manual
>  
> I've done it and it is working - thank you. One more question - I wondered if it is possible to have the MATTR output formatted like KIDEVAL is, in a spreadsheet with each transcript listed in a row and the MATTR in the column next to it. This would make it so much easier to copy + paste into my stats software. 
> 
> Thanks, 
> Amber 
> 
> On Wednesday, 19 August 2015 19:08:15 UTC+1, Amber wrote:
> Thanks Leonid! Going to download a newer version. 
> 
> Best
> Amber 
> 
> On Wednesday, 19 August 2015 18:56:02 UTC+1, Spektor, Leonid: CMU wrote:
> Amber,
>  
>     CHECK does not look at %mor tier. KIDEVAL will check word's one to one correspondence between %mor tier and main speaker tier, But, if you have already ran KIDEVAL with your current version of CLAN without any errors, then the latest version of CLAN will not find errors either. You do not need to re-MOR your data to use MATTR analyses, because FREQ is only looking at main speaker tiers and is ignoring %mor tiers. But, if you  want to get more accurate results from KIDEVAL, for example, then you should get new MOR grammar and re-MOR your data. In fact you should do that even if you do not update your CLAN, because we constantly improve MOR grammar and this will give you more accurate results.
> 
> Leonid.
>  
> On Aug 19, 2015, at 13:14, Amber <ambal... at gmail.com <mailto:ambal... at gmail.com>> wrote:
>  
> Dear Brian
> 
> Thank you. When I have gone to run the +bN command, it is not working. I am thinking that I need to download a newer version of CLAN (I can see from previous posts that MATTR has only recently been implemented). However, I want to avoid having to re-MOR and check all my transcripts to ensure their compatibility with the newer version of the software. Is there anyway to update my current CLAN software (without downloading the newer version) so that will allow me to run +bN? 
> 
> Best wishes
> Amber 
> 
> On Wednesday, 19 August 2015 17:22:53 UTC+1, Brian MacWhinney wrote:
> Dear Amber,
>  
>     Thanks for noting that old link in the manual.  I have removed it and replaced it with a pointer to 
> Malvern, D., Richards, B., Chipere, N., & Purán, P. (2004). Lexical diversity and language development. New York: Palgrave Macmillan.
>  
> In regards to sample size, VOCD requires 100 utterances.  For smaller samples and more generally, you may wish to consider using MATTR, as described in my previous ChiBolts message on this topic.  There I gave a reference to 
>  
> Covington, M. A., & McFall, J. D. (2010). Cutting the Gordian knot: The moving-average type–token ratio (MATTR). Journal of Quantitative Linguistics, 17(2), 94-100. 
>  
> and you may wish to read the recent comparison of VOCD, TTR, and MATTR
>  
> Fergadiotis, G., Wright, H., & Green, S. (2015). Psychometric evaluation of lexical diversity indices: Assessing length effects. Journal of Speech, Language, and Hearing Research, 58, 840. 
>  
> My take-home from this is that people should only use TTR when comparing across samples of the same length and even then, VOCD or MATTR would be better.  In general, researchers should prefer MATTR to VOCD.  In CLAN, you run MATTR using this option in FREQ
> 
> +bN This option calculates the lexical diversity of a sample using the Moving Average Type-Token Ratio (MATTR). This index is based on a moving window that computes TTRs for each successive window of fixed length (N). Initially, a window length is selected (e.g., 10 words) and the TTR for words 1- 10 is estimated. Then, the TTR is estimated for words 2-11, then 3-12, and so on to the end of the text. For the final score, the estimated TTRs are averaged.
> 
> 
> —Brian MacWhinney
>  
>  
> -- 
> You received this message because you are subscribed to the Google Groups "chibolts" group.
> To unsubscribe from this group and stop receiving emails from it, send an email tochibolts+u... at googlegroups.com <mailto:chibolts+u... at googlegroups.com>.
> To post to this group, send email to chib... at googlegroups.com <mailto:chib... at googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/2bef66d3-ac3e-468b-b15a-f3624b731eb0%40googlegroups.com <https://groups.google.com/d/msgid/chibolts/2bef66d3-ac3e-468b-b15a-f3624b731eb0%40googlegroups.com?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>  
> -- 
> You received this message because you are subscribed to the Google Groups "chibolts" group.
> To unsubscribe from this group and stop receiving emails from it, send an email tochibolts+unsubscribe at googlegroups.com <mailto:chibolts+unsubscribe at googlegroups.com>.
> To post to this group, send email to chibolts at googlegroups.com <mailto:chibolts at googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/d9633c76-3ba0-476d-9980-6fdd1a93ec79%40googlegroups.com <https://groups.google.com/d/msgid/chibolts/d9633c76-3ba0-476d-9980-6fdd1a93ec79%40googlegroups.com?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
> 
> -- 
> You received this message because you are subscribed to the Google Groups "chibolts" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to chibolts+unsubscribe at googlegroups.com <mailto:chibolts+unsubscribe at googlegroups.com>.
> To post to this group, send email to chibolts at googlegroups.com <mailto:chibolts at googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/78F7051232E584458D81A07B6C78AF7D013FCE3C%40OITMX1008.AD.UMD.EDU <https://groups.google.com/d/msgid/chibolts/78F7051232E584458D81A07B6C78AF7D013FCE3C%40OITMX1008.AD.UMD.EDU?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.

-- 
You received this message because you are subscribed to the Google Groups "chibolts" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chibolts+unsubscribe at googlegroups.com.
To post to this group, send email to chibolts at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/A606E23A-32E8-4009-9176-A00D772DA046%40cmu.edu.
For more options, visit https://groups.google.com/d/optout.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/chibolts/attachments/20150819/4639a926/attachment.htm>