Given that transposition neighbours are relatively similar in the spatial coding model, we might expect that the model should be able to capture transposed letter (TL) priming effects of the sort demonstrated by Perea and Lupker (2003) and many other researchers. Perea and Lupker found that lexical decisions to a target word like judge are faster following the TL prime jugde than a control RL (replaced letter) prime such as jupte.

To test this in the model we can switch to the pldt (primed lexical decision) trial, and specify the prime as jugde and the target as judge:

After running this trial, go to the Results tab and then the Tables top tab. The pldt.table shows the model's response (yes) and the latency to respond (in this case, 116 cycles). Now try running a trial in which the prime is the RL prime jupte. This time the response latency is 139 cycles. That is, there is a TL priming effect of 23 cycles for this target.

This difference is promising, but it's unwise to draw conclusions from a single target. Next we'll consider how to run a simulation of a larger set of trials, in order to simulate a full experiment.

Next: Loading a set of trials