Abstract
Recommendations of the ACM/IEEE Joint Curriculum Task Force on the Core of Computer Science and the dissatisfaction with the poor performance of too many CS1 students at Middle Tennessee State University combined to motivate the Computer Science Department to compare the use of closed labs and open labs. A controlled experiment was conducted in the Fall 1992 and Spring 1993 semester of CS1 classes. Some of these classes were chosen for closed lab and others for open lab experiences. ANOVAs of various comparison groups consistently showed significant differences on comprehensive CS1 exams in favor of the students in closed labs.