1 Ahadi and Lister (2013) found that many of their introductory programming students had fallen behind as early as week 3 of semester, and those students often then stayed behind. Our later work (Ahadi, Lister and Teague 2014) supported that finding, for students at another institution. In this paper, we go one step further than those earlier studies by observing a number of students as they complete programming tasks while thinking aloud. We describe the types of inconsistencies students manifest, which are often not evident on analysis of conventional written tests. We again interpret our findings using neoPiagetian theory. We conclude with some thoughts on the pedagogical implications of our research results.