Position paper: Assessing knowledge in blocks-based and text-based programming languages

Assessing student knowledge of programming concepts is a long studied, but many would say unsolved, problem. From the initial Empirical Studies of Programmers series [1] to McCracken's study [2] to Elliott-Tew's validated assessment of fundamental CS knowledge [3] many have attempted to precisely measure the knowledge and skills of students learning to program. All of the previous efforts have been centered on text based languages. Now we have another element to add to the mix - block-based programming languages. Can any of the existing assessment instruments work for both text and block based programming languages? Do we believe they will measure the same knowledge? Can existing assessments be modified and tailored for block-based languages? Are the two approaches equivalent, and would we even want to assess the same knowledge outcomes for both? These are but a few of the questions I will address in this position paper.