Inquiry about punctuation tags

[N AG]

Respected Sir,

I hope you will be fine. As you know I am working with motivational speeches. I want to hide punctuation marks in my keyword analysis because comma, question marks appear at the top of my keyword list. While I am working on linguistic items not on punctuation.

AI mentioned I can hide them from punctuation token feature in tagg option by opening global setting of antconc. But I could not find that option. Can I hide punctuation or I have to copy me data and then exclude the punctuation items from my keyword data? Kindly guide me. Thanks.

Nadia

[N AG]

Respected sir,

My second inquiry is about Effect size setting. You told me how i can do my data analysis in antconc by generating word lists and then keyword analysis of Male and female motivational speeches. You said I should use default p value loglikelihood setting And there is no need to change it. But I did not ask about Effect size, it's default setting is Dice. So, I discussed with gemini AI. It suggested log ratio (how many times more or less) or DRF (specify differnce with percentage) are best in my case. Do I need log ratio in my Study or should I go for Dice default setting of Effect size option like i did with other settings.

Minimum word frequency should be Select 5 or it also should be the by default 1.

I have data more than 1lac and 40 thousand in each male and female data sets and total 2lac 80thousand plus words. So normalized frequency will Be set on 10000 words.

1000 or million will over or under eggagerate the Data findings. So i read different articles and found 10thousand is suitable. What is your opinion sir? I need your guidence for these 3 inquiries. Thanks in advance.

Nadia

[Laurence Anthony]

Hi, 

Normally, punctuation is not considered as part of words so it shouldn't appear in your keywords list. But, from your direct messages to me, it seems that you are tagging your data at the outset. If so, you should load in your data into AntConc using the simple word, pos, headword indexer in the corpus manager. Once you do that, all the punctuation etc. will be handled automatically and the results should be much better.

I hope that helps!

Laurence.

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

--
You received this message because you are subscribed to the Google Groups "AntConc-Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to antconc+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/antconc/CAEjwtaZR%2BR0TroACBgev%2BmN2OcTGcO8UnzxqB9dnYB%2B5r2ubjw%40mail.gmail.com.

[Laurence Anthony]

Hi,

1) Each effect size measure has its pros and cons. So, I recommend not blindly following what an AI engine says. The standard settings in AntConc are fine, and you have several options there to choose from.

2) The normalized frequency can be set at 10,000 or 100,000 or 1,000,000 depending on the size of your data. The only important thing is that you state what the normalization setting was when you explain your results.

I hope that helps!

Laurence.

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

--

You received this message because you are subscribed to the Google Groups "AntConc-Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to antconc+u...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAEjwtaYobMGAu_kRq_C-tfhQaJaNhA9g31mfP%2BQ7nQPqrxBM7g%40mail.gmail.com.

[N AG]

Thankyou So much Sir. I saw a paper published in 2022.they have used antconc 3.5.8 version and in their keyword Analysis they showed +sign in front of loglikelihood keyness results. They used brown corpus as refrence Corpus In their word.

I have my own male/female two datasets, I never see those + signs in my keyword analysis. Do you have excluded this keyness feature in latest 4 versions or it is still availble sir?

Nadia. 

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAL6Fgv10FQ7y%2BHGsbN-W02yuL5za8HxmdSbWXAe3Y%2BtYkc%2BgGA%40mail.gmail.com.

[N AG]

Respected Sir,

Thanks for this suggestion. I uploaded my data by following your youtube videos. Let me do it again and let you know either it is working or not.

Sir I read antconc Group conversation for compound Nouns or hyphen Words. Where i can see them in antconc? They appear in list but not in concordance. Do i have to use any n gram feature or  what? Because I understood through Chat that we can see those black-wheel in antconc but how don't know.

Yes sir, I did not follow AI blindly. I have not any option so I just discussed but did not take its answer. 

Sir I have so much function words in my frequency and keyword lists while I just want to see content Words. 

Is it OK if I paste my data on spread sheet and just keep content words and remove all function words? Or there is no need to delete them I can simple ignore them and work with content Words wherever they are coming in the list. Thanks. 

Nadia

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAL6Fgv1neNpTmvkgsr6K%3Dg_TuN1KfqvbGwZRmt_K_5eToTmjoA%40mail.gmail.com.

[Laurence Anthony]

+ just means positive keywords.

Laurence

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAEjwtaZzuKifTJP7z8YAA5%2BjVUrwfH-sr0F-hQNUQX%3DKQWdh%2BQ%40mail.gmail.com.

[Laurence Anthony]

Keywords might generate function words if they are key in your target corpus. I suggest you analyze the data that you have in front of you without delete words just because they are functions words.

I'm not sure what you mean by "They appear in list but not in concordance".

I hope that helps!

Laurence.

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAEjwtaZBBxOAfDnGNdprfPBUDBUmqR885Y7CXLad3aKUxTLhjg%40mail.gmail.com.

[N AG]

Respected Sir,

Respected sir,

You are right + means positive but I never see this sign in keyness column, as mentioned by them in their published paper, when I run keyword anysis. I have tried to upload my data again like I did earlier by seeing your videos. The punctuation is still there. 

I want to see linguistic items (noun, Verbs, adjectives and modal Verbs) not the functional Words Or any type of punctuation. As you suggested just see what is coming from your data. They will object on this thing during defence.

Am I doing it in a correct way or something wrong? 

Shall i clean my data or share the files so you can see what is the problem.

You asked what words appeared But when double clicked antconc show warning for No hits found. I have attached the screenshot of put-down word. I have many words but do not know how to see them in antconc.

Secondly, you said antconc different statistical tools have pros and cons

Log ratio tells us how many times word is occurred and DRF tells the thing in percentage. That was the thing mentioned by AI. Now i should for dice or Log ratio or just use antconc statistical tools' BY default setting?

Ginalthing you said you could not understand what I meant when say appear in list but not in cwoncordance. As I Tap the word 'put-down' antconc gives warning 'No hits found'. How i can see them in antconc suggest me. Thanks for your time and patience sir you invested on us who are trying to understanding with lots of questions. But your guidence always makes things clear for me. Stay blessed. 

Nadia

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAL6Fgv2YO-w5qA7BasNODmkcHNEgLLf0W%3D%2BJ4G6T3dzLWU7xHw%40mail.gmail.com.

[Laurence Anthony]

Hi Nadia,

You are starting to do what you did in the past, asking multiple questions interspersed with comments. Also, the English you are using is very difficult to parse as it contains numerous grammatical errors and typos. Can you carefully check your language, and the questions you want to ask. List them up clearly, one by one. 

Also, please remember that this is a discussion forum about AntConc. It's not a place to ask general questions about basic corpus methods, which should find from watching tutorials and reading books. If you are studying a course in corpus linguistics, most of these questions are probably best directed to your supervisor.

I hope that helps!

Laurence.

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAEjwtaYcBRCLxndqFGor7xe4YqT181oP86y-z5qm8Awvgwr20Q%40mail.gmail.com.

[N AG]

Hi Respected Sir,

Accept my apologies for this inconvenience. Now I have two questions. Sir, while generating frequency for wordlists and keyword lists, I checked the regex box and now I can see all hyphenated words too. I checked and unchecked the regex box to see any difference in the total number of tokens, but I didn't notice any difference. So, is it okay if I check the regex box? Is this the correct way? Can I continue with this? I read this thing in a book

"Corpus Linguistics and the Description of English by HANS LINDQUIST, published in 2023. 

Gemini AI suggested me to type this type of tags if you want to see your desired word category only. 

"I typed the *_VV* verb tag and I can see all types of verbs. I repeated this with other adjective, modal verb tags, and I can see all types with a single tag because I want to see the top 50 verbs, nouns, adverbs, adjectives, and modal verbs for hedges and boosters. This is working for me, and I'm not facing any issues. Again, is this the correct way, and can I stick to this method to find the words I want to see? Yes or no?

BY using these two ways I am not facing any issue or not a single token is missing in list and concordance...like previously I see "no hits found" when I try to see the compound Words or hyphenated words. 

My supervisor is not a corpus linguist and I do not know anyone from teachers' category who can discuss or guide me about Corpus Or antconc. So i read and try to Understand or ask from you. Thanks in advance. Stay blessed.

Nadia

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAL6Fgv0p%2Btntkij_ypRrY3GYsX%3D%2BMMGCUhfPWu6SZcw6zY%3Ds-g%40mail.gmail.com.

[Laurence Anthony]

Hi Nadia,

It's great to hear that you are making progress. In the word/keyword tool, the search function is only for searching within the list, so the total number of words in the list will never change.

For your searches like "*_VV*" you don't need to turn the regex option on, because these are just standard wildcard searches. I recommend that you only turn on the regex option when you want to do regex searches. Otherwise, the results can be unexpected.

In the future, when you ask questions here in the discussion group, avoid wording like the following:

" So, is it okay if I check the regex box? Is this the correct way? Can I continue with this?"

You are asking 3 questions when 1 is sufficient.

=> "Is this the correct way to do my search?"

Similarly, you use 3 questions below:

"  Again, is this the correct way, and can I stick to this method to find the words I want to see? Yes or no?"

=> "Is this correct way to find the words I want to see."

I hope that helps!

Laurence.

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAEjwtaaqw_j7_mgVRLeMufW2SbEJSrmLAKGwMaDgPgCVQyyKkg%40mail.gmail.com.

[N AG]

Respected sir, 

Yes you are right sir. I jst want to that one question mention by you. Thank you for your patience, guidence and teaching me how to communicate properly. I will take care these things. By the way sir, after typing my text I asked meta to check it and meta said. No need to change or correction😁 everything is fine. Thanks.

Nadia. 

To view this discussion visit https://groups.google.com/d/msgid/antconc/CAL6Fgv0LOXSAFMohwuUkxGTU6m7WxDYQikzMorWP6ZFf10AutQ%40mail.gmail.com.