Thanks for that, Yisrael. That sounds like a very interesting approach. Still, I was hoping for more of a cognitive psychology/psycholinguistic method.

Cross-modal priming could be used: reading in one language, and when the eyes land on the critical word (eye tracking) or when the critical word is revealed (self-paced reading), an audio stimulus in the other language is played for yes/no lexical decision. If the written word is priming the auditory word in the other language, the RT should be faster (and more accurate) than an unrelated word.

It is possible to use Stroop task, using words from L1 as primes, and asking participants to do the picture naming task in L2 (targets). This is of course modified Stroop task that includes picture naming.

You can manipulate either orthographic or or semantic features of the word in order to get facilitation or inhibition on picture naming.

Another way is to use cognates.

ERP's can indeed be useful. We used both behavioral (masked priming paradigm combined with reading aloud and reaction time measurement) and electrophysiological measures (ERP's) to show cross-language priming (see work by my student Kalinka Timmer; e.g. Timmer & Schiller, BRES, 2012; Timmer et al., LCP, in press). I could imagine masked priming (whether or not in combination with EEG/ERP) may be a valid method for your purposes.

Think-aloud:

It has been demonstrated that the think-aloud procedure (Van Den Haak et al., 2003; Leow & Morgan-Short, 2004; Albrechtson et al., 2008), while increasing task time, does not react with cognitive tasks and is likely to generate useful qualitative data about what the respondent is thinking while making responses. If an associative response has been primed in some way then it would be possible to work out what kind of priming is involved based on such concurrent verbalisations. The downside to this method is that as some associations are made subconsciously the respondent may not be aware of how or why he or she made a particular response...in such cases there will be nothing to verbalise.

do you include stroop and picture/word interference tasks in the "lexical decision" task?Also, implicit memory tests might be helpful here

John,

visual search tasks may be quite useful for your purpose.

You could also use eye-tracking experiments to devise priming effects.

We should not forget an obvious fact that priming( positive or negative) is a psycholinguistic method that usually targets the learner's subconscious in order to create a targeted behavior. Seen from this perspective there are hundreds of techniques and hundreds of ways to calculate results. A rule of thumb is that the teacher is the best person who knows his students and consequently knows how to choose/create his/her own techniques.