Measures in TAALED

47 views
Skip to first unread message

Matthew C. Watterson/교양과

unread,
Apr 3, 2023, 9:43:09 PM4/3/23
to Suite of automatic linguistic analysis tools

Hello everyone,


Excuse my very basic question. 


I am using TAALED to estimate lexical diversity in some short spoken texts, using MTLD Original index as my method. The output csw includes two lexical density results: 


lexical_density_types


lexical_density_tokens

Which of these two measures should I be reporting as the measure of lexicat diversity?


Best regards


Matthew


Kristopher Kyle

unread,
Apr 3, 2023, 10:41:46 PM4/3/23
to Matthew C. Watterson/교양과, Suite of automatic linguistic analysis tools
Hi Matthew,

There isn't a clear answer here (that I am aware of). Most research has indicated that while lexical density is helpful for distinguishing between registers and modes, it isn't a helpful indicator of proficiency in the studies I have read. It is included in TAALED mostly for historical purposes (or for those who wish to explore genre differences).

Hope that helps! Anyone who uses lexical density indices on a regular basis feel free to chime in!

Kris

--
You received this message because you are subscribed to the Google Groups "Suite of automatic linguistic analysis tools" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linguistic-analysi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/linguistic-analysis-tools/1ebe528b-64c2-4a5f-a423-13b8abd100b3n%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


--
Kristopher Kyle
Associate Professor
Department of Linguistics
University of Oregon
Message has been deleted
Message has been deleted

Kristopher Kyle

unread,
Apr 4, 2023, 6:07:12 PM4/4/23
to Matthew C. Watterson/교양과, Suite of automatic linguistic analysis tools
Hi Mathew

I think I understand the confusion now.

Neither of those indices are for MTLD. You should be reporting "mtld_original_aw".

Best,

Kris

On Tue, Apr 4, 2023 at 11:44 AM Matthew C. Watterson/교양과 <m...@hongik.ac.kr> wrote:
Thank you, Kris.

I suppose my question is, if I do decide I want to report 'lexical diversity' (via MTLD) for my data, which number from the TAALED csv output should I be reporting: "the lexical_density_types" one or "the lexical_density_tokens", or something else?.

The studies I've read seem to report just one number for MTLD. Would this be one of these? Or something else?

Best Matthew

2023년 4월 4일 화요일 오전 11시 41분 46초 UTC+9에 kristop...@gmail.com님이 작성:

For more options, visit https://groups.google.com/d/optout.

Matthew C. Watterson/교양과

unread,
Apr 4, 2023, 7:42:56 PM4/4/23
to Suite of automatic linguistic analysis tools
Thank you, Kris.

Yes, for some reason the first time I ran TAALED that column was missing, must have not clicked the right boxes. Ran it again just now, and that "mtld_...aw" data is there now. 

Problem solved!

Thanks again.

Matthew

2023년 4월 5일 수요일 오전 7시 7분 12초 UTC+9에 kristop...@gmail.com님이 작성:
Reply all
Reply to author
Forward
0 new messages