Pair programming, as part of the Agile Development process, has noted benefits in professional software development scenarios. These successes have led to a rise in use of pair programming in educational settings, particularly in Computer Science 1 (CS1). Specifically, McDowell et al. [2006] has shown that students using pair programming in CS1 do better in a CS2 class (with solo programming) than students who don't pair in CS1. This paper seeks to address a similar question, but from a qualitative, student-focused approach. How do students define, experience, and value the pair programming experience? How do they experience and value it compared to solo programming? Does pairing in CS1 impact their confidence in their abilities? We report on semi-structured interviews with 11 subjects from two institutions where pair programming was used in CS1, and solo programming was used in the CS2. Many of the responses met our expectations; students get stuck less and explore more ideas while pairing, and believe that pair programming helped them in CS1. Other responses were more surprising. Students reported that when solo programming they were more confident and understood their programs better. Many students also said that they started work on their assignments earlier when soloing. Students also continue to use other students as resources even when working "solo."