Tuesday, June 08, 2010

Back to Maltese...

Work in Jerusalem continues apace. Priming experiment items are almost selected, and remain to be recorded, spliced, and programmed into DMDX. The two word familiarity experiments are coming along well too - items are being vetted by expert teams of native speakers for both Hebrew and Arabic. But today's potentially very exciting news concerns Maltese. A hugely important warning, though: the data you are about to see described have not been through any process of outlier removal, and no statistical analyses have been run. Translation: take this with a grain of salt.

First off, we only have data processed for our auditory priming experiments, but not our subliminal priming experiments. In the auditory experiments, subjects heard prime-target pairs. In the root priming experiment, there are three conditions: identity (prime and target are identical words), root-related (prime and target share a consonantal root but are in different binyanim), and unrelated (prime and target share neither a root nor a binyan). For pairs of real-word primes and real-word targets, here's our preliminary data:

In the binyan priming experiment, everything is the same except that in the related condition, the prime and the target share the same binyan. Here's that graph:

So, at the risk of overinterpreting, it appears that when a prime and a target share a consonantal root (but not a binyan), there is as much facilitation as when the prime and target are identical, but when a prime and a target share a binyan (but not a consonantal root) there is no such facilitation.

I'm not going to say anything more committal, since until we run the appropriate statistical analyses these are simply pretty pictures. But there we have it, our first look at the data.


Post a Comment

<< Home