The Computing Attitudes Survey (CAS) is a newly designed instrument, adapted from the Colorado Learning Attitudes about Science Survey (CLASS), for measuring novice to expert-like perceptions about computer science. In this paper we outline the iterative design process used for the adaptation and present our progress toward establishing the instrument's validity. We present results of think-aloud interviews and discuss procedures used to determine expert consensus for CAS items. We also detail results of a pilot of the instrument with 447 introductory students in Fall 2011 along with a preliminary factor analysis of this data. Findings to date show consistent interpretation of statements by faculty and students, establish expert consensus of opinion and identify eight candidate factors for further analysis.