Tutorial 9: Final Checks Before Submission

Version 2.0

Last Updated: May 22, 2020

To suggest revisions or receive notifications when changes are made to the tutorials, contact ACLEW-DAS@babylanguagelab.org

Check 1. File playback

Listen through the recording one more time while keeping an eye on your annotations:

Are there any missing utterances?

Any problems with onset and offset boundaries?

Any utterances accidentally on the wrong speaker tier?

Any other problems you notice?

Watch closely and fix where needed. Your careful reviewing is super important for preventing confusing errors later on.

Check 2. Transcription formatting

Go to the Grid view of your file, select a speaker’s transcription tier from the dropdown menu, and look through their utterances for violations of the minCHAT transcription formatting. Repeat until you’ve checked all speakers.

​

1. Go into Grid mode

2. Pick a speaker

3. Review the transcriptions for formatting errors

Check 2. Transcription formatting

Some common errors to look for include:

• Capitalization of words that aren’t proper nouns (especially at the beginning of an utterance).
• Missing terminal markers.
• Missing letters after symbols (eg. @__ instead of @c).
• Obvious spelling errors or typos.

​

Check 3. Annotation checks

Staying on the Grid view of your file, select each speaker’s xds tier and check that (a) there are no missing annotations and (b) that each “C” or “T” utterance is transcribed.

Then, also in Grid view, switch to the target child’s tiers and make sure there are no missing annotations:

• 0-7 months:
• vcm is completely filled in
• 8+ months:
• vcm is completely filled in,
• lex is filled in for utterances where vcm is “C”, and
• mwu is filled in for utterances where lex is “W”

​

Check 4. Speaker metadata

Sometimes you’ll have additional information about the speakers, like who each individual is in relation to the child. Add this information by going to Tier > Change Tier Attributes… and then going to the transcription tier of the speaker you want to update (in this example, it’s FA1), and add your notes to the “Participant” box. Then click Change. Add participant information to other speakers as needed and, when done, click Close.

​

1. Pick a transcription tier to update

2. Add your speaker notes

3. Confirm the change

Note:

For now, each lab may keep their own notation system for speaker metadata, but we may want to create a shared system later on.

Check 5. Test your file on the minCHAT checker

Upload your file to the minCHAT checker tool, which automatically checks for errors you may have missed.

Instructions for how to upload are here.

The minCHAT checker is here.

Fix any remaining errors before proceeding.

Check 6. Check filename, safely save, and submit

Make sure that your .eaf file is securely saved in the right with the correct name. If you are unsure on saving protocols, consult with your lab’s PI.

Submit it through the normal system and update any lab-internal to-do lists and/or spreadsheets with your progress.

Done!

And onto the next one...

Acknowledgements

This tutorial was made possible by the significant contributions of:

Dr. Marisa Casillas

​

Additional support provided by:

Maartje Weenik

Annick Bosch

​