|
Size: 1263
Comment:
|
← Revision 16 as of 2011-08-10 17:33:11 ⇥
Size: 4868
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 6: | Line 6: |
| = For Experiment 1 = | |
| Line 8: | Line 7: |
| == Pair == w1,b2,w3&b4 all prime for gave w5,b6,w7&b8 all prime for hand w9,b10,w11&b12 all prime for pass |
= For Experiment 2 = Use tab '''Expt 2 with training''' of the excel file. <<BR>> Run the numbers with Training and No``Training combined. Then run the numbers excluding No``Training. See which one gives you better verb variation and a stronger effect. Do the better instructions do anything to our values? <<BR>> == Trial == 2_agent_gab_gauge <<BR>> 2: experiment 2 <<BR>> agent: manipulating the name of the agent <<BR>> morph: morphological overlap <<BR>> gab: First 2 letters = agent (Gabe), last letter = recipient (boy) <<BR>> gauge: the object transferred <<BR>> Codes: <<BR>> Ga = Gabe <<BR>> Ha = Hannah <<BR>> Si = Simon <<BR>> Pa = Patty <<BR>> B = boy <<BR>> W= woman <<BR>> |
| Line 14: | Line 27: |
| Controls = sentences with frequency matched words with which to compare the phon prime. Prime = the phon prime word |
For right now we are only interested in “agent” trials agent: manipulating the name of the agent Patti/Gabe/Hannah<<BR>> morph: morphological overlap <<BR>> <<BR>> |
| Line 17: | Line 32: |
| == PhonMatchIntended == * 1 = phon prime * 0 = no phon prime |
== Codes for Phonetic Matches == Phon``Match_Intended_Agent <<BR>> Phon``Match_actual_Agent <<BR>> Phon``Match_Intended_theme <<BR>> Phon``Match_actual_theme <<BR>> 0=none <<BR>> 1= initial phonetic match <<BR>> |
| Line 21: | Line 40: |
| == #Disfluencies == number of disfluencies in the sentence |
== Total #Disfluencies == Column Name = Total``Num``Disfluencies. The number of different disfluencies participants produced in their sentences. <<BR>> |
| Line 24: | Line 44: |
| == 2verbs == 1 = 2 verbs used in the sentence... throw it out! |
== Loc Disfluencies == Currently contains 6 columns: <<BR>> '''For all of these columns except 'other' a 1 indicates the presence of a disfluency in this location.''' <<BR>> <<BR>> In the case of word restarts, the '''last noun that the person chooses''' is considered the real head. For example in: ''Gabe gave the boy a fence / a cage / gate@wr.'' We count this as being before the second NP, even though we have 3 attempts at the 2nd noun phrase. <<BR>> <<BR>> In the case of word lengthenings, we assume that the reason the word is extended is because of upcoming problems. Therefore in: ''Gabe gave@ln the boy a gate.'' We count this as a disfluency '''after''' the verb and before the first noun phrase. |
| Line 27: | Line 50: |
| == DO? == 0=prepositional phrase 1= double object construction 2= neither |
|
| Line 32: | Line 51: |
| == Trial order == a= 1st block b = 2nd block of trials #position presented in the experiment... check for changes over time |
|| Column Name || Explanation || Examples || || NP-SUBJdisf || Receives a 1 for all disfluencies before the first noun.|| Yes: Um@fp Gabe gave the woman a pan . <<BR>> Yes: Gabe was going to oo sorry Gabe@sr gave the boy a harp. <<BR>> No: There is a boy wearing um@fp blue shorts. Hannah passes him the pan .|| || VERBdisf || Receives a 1 for all disfluencies between the first noun and the verb. Do NOT count verb restarts here. || Yes: Gabe / pushes over the shovel. || || POSTVB-NP1disf || Receives a 1 if there is a disfluency between the verb and the first noun phrase. || Yes: Patty handed@ln / the pan to a woman. <<BR>> Yes: Gabe pushes / the rainbow. <<BR>> YES: Gabe gave a pot / a pan@wr to the woman. || || POSTVB-NP2disf || Receives a 1 if there is a disfluency between the first noun and the second noun. || Yes: Gabe gave a heart / to a woman . <<BR>> Yes: Patty passed a boy@ln a globe . || || Other || Count of disfluencies that occur in other parts of the sentence. (Receives numbers > 0) || || |
| Line 37: | Line 58: |
| = For Experiment 2 = Participants 4&5 were improperly trained; participant 2 and 3 were overly creative and should probably be thrown out. |
Examples: <<BR>> Hannah gave / the child a uh@fp / pipe. <<BR>> Receives a 1 for POSTVB-NP1disf, and a 1 for POSTVB-NP2. Other = 0. Total Disfluencies = 3. <<BR>> <<BR>> Gabe@ln punched /// a snack // that was sitting on the table. Receives a 1 for VERBdisf, a 1 for POSTVB-NP1disf, and a 1 for Other. Total Disfluencies = 3. <<BR>> <<BR>> Ha- um@fp Patty@sr placed a ladle@ln on a table. Receives a 1 for SUBJdisf, a 1 for POSTVB-NP2, and a 1 for Other. Total Disfluencies = 3. <<BR>> <<BR>> |
| Line 40: | Line 68: |
| Participants 1,2,3,4,5, and 10 ran with after getting introduced to the picture labels. All the other participants ran in the newest version of the experiment (which includes a session of hearing the labels and seeing the pictures and then testing via seeing the pictures and being forced to choose a label). | == Subject, verb, object, recipient == Subject & Subject_intended <<BR>> Object & Object_ intended : for any analysis of the affect of the object on the sentence... make sure that Object & Object_intended are the same <<BR>> Recipient <<BR>> Verb = The verb phase. You will find things like “has passed” or “ is passing” in this column. Verb``Stem = The first active transitive verb in the sentence in singular past tense. <<BR>> <<BR>> ''Use the Ver``bStem column for any and all analysis.'' == Exclude == if != 0 then exclude <<BR>> == DO == 0 = Prepositional Phrase <<BR>> 1 = Double object <<BR>> 2 = Other (passive, incomplete, etc) <<BR>> == Strong Prime == 1 = strong <<BR>> 0 = weak <<BR>> control = filler <<BR>> == Version == No training = our original 14 subjects who learned the names of the pictures but didn't have beautiful instructional videos to watch that helped prime for pass and hand. Hopefully there is less verb variation in these people. <<BR>> These people got trained on object and people names using the same methods as the no training people, but also got an instructional video with subtle primes for pass and hand. <<BR>> |
For Experiment 2
Use tab Expt 2 with training of the excel file.
Run the numbers with Training and NoTraining combined. Then run the numbers excluding NoTraining. See which one gives you better verb variation and a stronger effect. Do the better instructions do anything to our values?
Trial
2_agent_gab_gauge
2: experiment 2
agent: manipulating the name of the agent
morph: morphological overlap
gab: First 2 letters = agent (Gabe), last letter = recipient (boy)
gauge: the object transferred
Codes:
Ga = Gabe
Ha = Hannah
Si = Simon
Pa = Patty
B = boy
W= woman
Trial Type
For right now we are only interested in “agent” trials agent: manipulating the name of the agent Patti/Gabe/Hannah
morph: morphological overlap
Codes for Phonetic Matches
PhonMatch_Intended_Agent
PhonMatch_actual_Agent
PhonMatch_Intended_theme
PhonMatch_actual_theme
0=none
1= initial phonetic match
Total #Disfluencies
Column Name = TotalNumDisfluencies. The number of different disfluencies participants produced in their sentences.
Loc Disfluencies
Currently contains 6 columns:
For all of these columns except 'other' a 1 indicates the presence of a disfluency in this location.
In the case of word restarts, the last noun that the person chooses is considered the real head. For example in: Gabe gave the boy a fence / a cage / gate@wr. We count this as being before the second NP, even though we have 3 attempts at the 2nd noun phrase.
In the case of word lengthenings, we assume that the reason the word is extended is because of upcoming problems. Therefore in: Gabe gave@ln the boy a gate. We count this as a disfluency after the verb and before the first noun phrase.
Column Name |
Explanation |
Examples |
NP-SUBJdisf |
Receives a 1 for all disfluencies before the first noun. |
Yes: Um@fp Gabe gave the woman a pan . |
VERBdisf |
Receives a 1 for all disfluencies between the first noun and the verb. Do NOT count verb restarts here. |
Yes: Gabe / pushes over the shovel. |
POSTVB-NP1disf |
Receives a 1 if there is a disfluency between the verb and the first noun phrase. |
Yes: Patty handed@ln / the pan to a woman. |
POSTVB-NP2disf |
Receives a 1 if there is a disfluency between the first noun and the second noun. |
Yes: Gabe gave a heart / to a woman . |
Other |
Count of disfluencies that occur in other parts of the sentence. (Receives numbers > 0) |
|
Examples:
Hannah gave / the child a uh@fp / pipe.
Receives a 1 for POSTVB-NP1disf, and a 1 for POSTVB-NP2. Other = 0. Total Disfluencies = 3.
Gabe@ln punched /// a snack // that was sitting on the table. Receives a 1 for VERBdisf, a 1 for POSTVB-NP1disf, and a 1 for Other. Total Disfluencies = 3.
Ha- um@fp Patty@sr placed a ladle@ln on a table. Receives a 1 for SUBJdisf, a 1 for POSTVB-NP2, and a 1 for Other. Total Disfluencies = 3.
Subject, verb, object, recipient
Subject & Subject_intended
Object & Object_ intended : for any analysis of the affect of the object on the sentence... make sure that Object & Object_intended are the same
Recipient
Verb = The verb phase. You will find things like “has passed” or “ is passing” in this column. VerbStem = The first active transitive verb in the sentence in singular past tense.
Use the VerbStem column for any and all analysis.
Exclude
if != 0 then exclude
DO
0 = Prepositional Phrase
1 = Double object
2 = Other (passive, incomplete, etc)
Strong Prime
1 = strong
0 = weak
control = filler
Version
No training = our original 14 subjects who learned the names of the pictures but didn't have beautiful instructional videos to watch that helped prime for pass and hand. Hopefully there is less verb variation in these people.
These people got trained on object and people names using the same methods as the no training people, but also got an instructional video with subtle primes for pass and hand.