A comparison of the experiments

Old

Expt 2

Expt 1

Length

96 trials

126 trials

96 trials

Communication task?

yes, with Katrina as partner

no

no

Memory task?

recall task of 40 items

pictures -> labels

pictures -> labels

Training?

see list of pictures with labels for 2 minutes

40 objects, 4 names repeated twice
Part 1: See picture, hear name twice
Part 2: Test for label, hear name again

40 objects at a time, 80 total
Part 1: See picture, hear name twice
Part 2: Test for label, hear name again

#Experimental Trials

55

58

72

#Fillers

42

72

16

Nature of Exptl Trials

4 gi-primes, 2 ga-primes, 6 ha-primes plus their frequency matched words to use as a comparison. I was pretty new at this at the time.

3x3 design
Patti, Hannah, Gabe X hand give or pass X objects beginning with ha, ga, pa, or something else entirely

2x design
every verb pair has 4 sets of objects... used transitive verbs that could be described in at least 2 ways. 2 verbs prime for the one interpretation, 2 for the other, and the other 4 objects are frequency matched objects to give a base rate. Includes 12 hgp trials, 4 toss trials and 6 other verb pairs

Nature of Fillers

toss, throw, show, and offer trials with a balanced number of verbs and characters

18 super-strong phonological priming trials for transitive verbs, 56 transitive verbs hum-hum among 12 different verbs! most trials include Simon as the agent

4 transitive verbs... Simon kicks, pulls, pushes, or touches a woman or a boy

Results

10 subjects
showed preliminary priming effects from the object to the verb, about 25% of the hgp trials were 'hand' or 'pass', also showed the 'gi' primed 'gave' better than 'ga' did.

12 subjects, preliminary feedforward phonological priming, also effects of repetition on fluency and ditransitive noun phrase ordering

not yet analyzed

Problems

imbalanced number of trials per experimental condition! fillers often repeated the same people and objects, work on finding feedforward evidence first, too many experimental trials?

no variation of verbs

not yet analyzed

Breaking News

I ran a subject in experiment 2 today, but this time made it a communicative task (with me) and reinstated the context of the recall experiment. Between the two changes, subject 14p changed verb usage pretty regularly (I won't have a real idea until I look at the data), but he did not get stuck saying 'gave' the whole time.

I would like to try a version where I am not the other subject... perhaps I could have a confederate to press the space button and turn the task back into a communicative one. The memory task portion seemed to trip subject 14p up. When asked he said, "I wanted to make sure that I was labeling the objects correctly. I thought that one of the items was a handle at first, but I couldn't remember if it might have had a different label." Either way, I'm excited and I hope that this keeps happening. I'm running two more people in the old experiment tomorrow. -Katrina Housel 7/31/08

ProjectsPhonPrimingOldVnew (last edited 2011-08-10 15:00:31 by echidna)

MoinMoin Appliance - Powered by TurnKey Linux