View article

This paper describes and compares three sources of data used for computing instruction: real data, textbook data, and simulated data. Based on a qualitative examination of thirty data sets used in nine different courses, criteria for comparing the three sources are proposed and used to discuss the merits of each type of data. It is suggested that all three sources should be used because each has its distinct advantages. However, the argument is also given that, when appropriately created, simulated data may provide the most advantages for computing instruction. The paper concludes by making recommendations for creating well-simulated data for use in the classroom.