The really interesting thing is that parameter estimates strongly support obligatory encoding of the last item during sequential presentation of 3 difficult-to-name visual characters. Participants only seem to encode more than 1 item when a non-final list item is prioritized.