De-identifying name-spelling activities in CHAT

[Sarah Surrain]

Hi all,

I'm working on a new corpus of child language data drawn from randomly selected 5-minute clips of daylong home recordings. We're transcribing in ELAN using CHAT conventions. Our de-identification protocol replaces the child's spoken name with "Child_name" throughout. I have two related questions that I don't think the CHAT manual addresses directly, and I'd love the community's input.

1. De-identifying a name-spelling activity: In this clip, a father and daughter dyad engage in a name-spelling activity. He guides her through the letters and sounds of her name as she writes the letters. Transcribing this verbatim would effectively reveal the child's name through the sequence of letters and sounds, even if we never write the name itself.

Here is a pseudonymized version of the exchange (the child's name has been replaced with "Vanessa"). MA1 = Male Adult 1 (the father):

Original Spanish:
*MA1: v@l.
*MA1: /va/.
*MA1: /ne/.
*MA1: tan rápido?
*MA1: /s/.
*CHI: es s@l.
*MA1: s@l Vanes +...
*MA1: s@l.
*MA1: a@l, Vanessa.

English Translation:
*MA1: v@l.
*MA1: /va/.
*MA1: /ne/.
*MA1: so fast?
*MA1: /s/.
*CHI: it's s@l.
*MA1: s@l, Vanes...
*MA1: s@l.
*MA1: a@l, Vanessa.

I've considered replacing the actual letters and sounds with the corresponding letters/sounds of a pseudonym, or just using a placeholder like [Child_name spelled] or &=spells_child_name.

2. Transcribing deliberate, isolated letter sounds and syllables. This exchange also raises another transcription question. The father produces isolated phonemes (/s/) and syllables (/va/, /ne/) as deliberate pedagogical acts — he is teaching his daughter how to spell her name by modeling its component sounds. This is distinct from letter naming, which CHAT handles with @l. Is there a CHAT convention for this? (Because it is in Spanish, we have the added complication that sometimes the letter name and letter sound are the same, as in the final A of Vanessa.)

Any guidance on either question would be much appreciated!

Thanks,
Sarah Surrain

[Janet Bang]

Hi Sarah, 

De-identifying a name-spelling activity - This has come up for us too before. The decision for us came down to how we were going to analyze the data. We ultimately decided we wanted to capture the letter-usage, and we were ok with using a pseudonym or replacement letters. For example, for us with a name like 'Hannah', we decided to keep the repetition and use a nonsense set of letters (e.g., abccba - using @l with each letter, so it could retain the number of letters and any repeated letters). Each time they used the child's name we replaced it with something like "child_name".

For any syllable teaching or teaching with a partial word, we marked them as a cloze utterance with our own postcode convention (below). This did mean that the partial words did not count for our purposes when we analyzed tokens/types, but it was a small portion of the data so ultimately we let that go but retained it in the transcript and had the postcode in case we wanted to go back to it.

*MOT:       &+tur? [+ cloze]
*CHI:        &+tle. [+ cloze]

When things were in Spanish, we also added the precode when the whole utterance was in Spanish, or we added the @s for each word. 

Our conventions are here: https://talkbank.org/childes/access/Eng-NA/FMB_home.html.

Feel free to email me separately if you want me to share our internal lab manual. 

Janet

--
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+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/chibolts/d895d935-a90e-45c3-ba1e-f1f26d703c12n%40googlegroups.com.

--

Janet Y. Bang, Ph.D (she/her/hers)

Assistant Professor

Child and Adolescent Development

Lurie College of Education, San José State University

janet...@sjsu.edu | 408-924-3714

https://www.sjsu.edu/education/faculty/janet-bang.php

[Sarah Surrain]

Thanks, Janet! We will follow your lead and use a pseudonym with the same number of letters to capture the letter usage (but use Child_name when the full name is spoken). 

I like your postcode convention for cloze prompts, but I think what is going on here is different. The dad is producing isolated syllables from the child's name to scaffold her writing, not to prompt her to complete the word. I wonder if the best fit for this is from p. 49 of the current CHAT manual: "In Japanese, many letters refer to whole syllables or “kana” such as ro or ka. To represent this as well as strings of letters in English, use the @k symbol, as in ka@k or jklmn@k." 

If we did that, the pseudonimized version would be as follows:

*MA1: v@l.
*MA1: va@k.
*MA1: ne@k.
*MA1: tan rápido?
*MA1: &=s_sound.
*CHI: es s@l.
*MA1: s@l vanes@k +...
*MA1: s@l.
*MA1: a@l, Child_name.

Any thoughts on whether this is how the @k is intended to be used?  Or a better way to represent the dad pronouncing the /s/ sound to prompt child to write the letter S?

-Sarah Surrain

Assistant Professor of Early Childhood Education
Early Childhood, Multilingual, and Special Education
University of Nevada, Las Vegas