Using Bloom's Taxonomy to Write Effective Programming Questions for Autograding Tools
Lina Battestilli, Sarah Korkes, Olivia Smith, Tiffany M. Barnes · 2019
Automated grading has become crucial in supporting large introductory Computer Science courses by assisting instructors in reducing grading time and course costs. However, novice programmers are often frustrated by auto-grading tools as they often provide minimal feedback or the questions are too complex. We propose using Bloom's Taxonomy to gradually increase the complexity of the programming question in an auto-grading tool. The easiest questions are based on Knowledge and Comprehension. Next are questions that require Application and Analysis. The most complex questions require Design and Creativity. We have developed programming questions that fit these categories. For example, for debugging the students have to understand and Analyze code. However, in order to write a program with more complex logic the students have to be Creative. The main goal is to lead novice programmers to learn more effectively and efficiently. In this poster, we present examples of the developed auto-graded programming problems based on Bloom's Taxonomy and the results of a pilot study in a CS-1 non-majors course