About choices of TACCO features

[mayu...@sina.com]

I am going to do a research on research articles with TACCO. Now I have three questions:

1. The tool is trained on the newspaper section of COCA, can I use the tool to analyze academic writing? if I can, is there any proof for my justification?

2. The indices are over 100, so it will be a bit exhaustive to use all of them for the research. Can I group them or choose some of them? Can anyone share any related reseach?

3. Can I do a comparison between the indices of TACCO and Coh-metrix? I am not sure if the indices of coherence are comparable. 

Thank you very much!

[Scott Crossley]

Short answers below

1. The tool is trained on the newspaper section of COCA, can I use the tool to analyze academic writing? if I can, is there any proof for my justification?

Sure. We have used the tool successfully for many academic analyses.

 

2. The indices are over 100, so it will be a bit exhaustive to use all of them for the research. Can I group them or choose some of them? Can anyone share any related reseach?

See a lot of our earlier work. Consider the following

1. Control for multicollinearity

2. Remove variables that do not show at least a small effect with your outcome variables

3. Check for normal distribution

4. Use dimensionality reduction

 

3. Can I do a comparison between the indices of TACCO and Coh-metrix? I am not sure if the indices of coherence are comparable. 

Some of them are (like connectives) but many variables are new (paragraph overlap and source overlap measures). You will have to make the links between the two tools.

Scott

 

Thank you very much!

--
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/a3cce2c5-007f-4e4c-b347-608e2eb9ac9fn%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Scott Crossley

Professor, Department of Special Education

Data Science Institute | Peabody College

Vanderbilt University

https://www.linguisticanalysistools.org/

[mayu...@sina.com]

Thank you for your kind reply!