A quick update on morphological family size statistics: additional regression analyses seem to continue supporting a facilitatory role for two factors in Hebrew lexical access: related family size, and word frequency, though apparently word frequency is only significantly facilitatory when taken as a linear measure rather than a log measure. Interesting. Additionally, unrelated family size has an inhibitory effect. This pretty much replicates the results from Moscoso del Prado-Martin's 2005 study but no one's ever done it in the auditory domain before our experiment. Also, there's never been a word frequency effect documented for Hebrew in the auditory domain, and interestingly, the frequency measures are based on a printed corpus and still have a significant effect in spoken word recognition. Neat!
The current goal is to figure out how to model this using mixed-effects modeling, which requires learning some more about R. To this end, Mike Hammond, Andy Wedel, and I plan to begin a reading group this semester to focus on learning how to use R. Based on a couple hours of work with the software yesterday, I clearly need such a group: I can't even get R to read my data file at the moment! There's nothing like being humbled by a command-line interface, that's for sure. Sigh.