Effect of talker variability on spelling.
Ashley's file was built to do the following:
- allow for sign in
- Pre-test people to determine a list of vocabulary terms they could not correctly identify
- Recognize the person when he/she signs in the next time
- Take those items and randomize them to high or low variability conditions
- train the person using the appropriate high/low conditions
- Post-test the person using the identification task
- post-test the person using the ratings task
- create an output file with the data we need that shows up on the desktop
We want Bonnie/Rebekkah's file to do somethign similiar
-
allow for sign in
-
Pre-test people to determine a list of vocabulary terms they could not correctly identify DIFFERENCES: instead of displaying a field of 4 images, people will see a picture and hear an audio file. They must type in the correct spelling for the word.
ALSO: The items are linked in terms of orthographic transparency, so the computer needs to keep track of PAIRS of items that are incorrect -
Recognize the person when he/she signs in the next time
-
Take those items and randomize them to high or low variability conditions
-
train the person using the appropriate high/low conditions DIFFERENCES: Training is presentation of a single image per word with different talkers (HIGH) or the single image with a single talker (LOW)
-
Post-test the person using the identification task DIFFERENCES: Each word will be spelled once. Image will be presented, talker will be novel. Answer will be typed in.
-
post-test the person using the PERCEPTION task DIFFERENCES: Instead of ratings, we'll use perception. People will see the image/hear a novel talker name the word. They will see 6 versions of the word, in a random order.
- Correct Spelling 2) Implausible Foil 3-6) Plausible Foils. NOTE: One foil will be the person's own mispelling of the word. Thus, the computer needs to recognize if the person got the word wrong, and use that spelling. If they get it right at post-test, then we use their original misspelling. We also need to make sure their misspelling is different from the other foils.
- create an output file with the data we need that shows up on the desktop DIFFERENCES: We want to see what was trained, what the answers were, and then have a summary file that organizes the information into easy columns. (Examples to come)