Sharon M. Carver

Cognitive objectives in a LOGO debugging curriculum: Instruction, learning, and transfer ☆

Abstract In this paper we report two studies in which elementary-school children learned a complex computer-programming skill—how to debug LOGO graphics and list-processing programs—and then transferred the high-level goal structure of that skill to nonprogramming domains. Instruction, its assessment, and the transfer tasks were all derived from an explicit model of the debugging process, cast as a computer simulation. Debugging skills were acquired over a period of several months as part of a LOGO programming course; the transfer tasks involved correcting written instructions in a variety of domains, including arranging items, allocating resources, and following map routes. Students showed clear improvement in the transfer tasks following instruction in debugging programs, and in the second study, amount of transfer was correlated with the degree of debugging skill acquisition. Our results contrast with many earlier studies that found little transfer of problem-solving skills in general and of high-level programming skills in particular. We suggest that the key to the success of our procedure is the fact that we used an extremely precise computer simulation model of the skills required to debug LOGO graphics and list-processing programs as a concrete manifestation of the notion of “cognitive objectives”.

Assessing Children's Logo Debugging Skills with a Formal Model

In this article, we propose a new way to assess childrens acquisition of debugging skills in a LOGO environment. The assessment procedure is based on an explicit and precise model (in the form of a computer simulation) of good debugging skills. The model has four stages: 1) evaluating the programs planned and actual outcomes to determine that debugging is necessary, 2) identifying the bug by using descriptions of the discrepancy between the planned and actual outcomes to propose potential bugs, 3) locating the bug by using clues about the structure of the program to narrow the search, and 4) correcting the bug and retesting the program. We describe model-based measurements of the LOGO debugging skills actually acquired by students in a “typical” LOGO graphics course. Nine seven- to nine-year-olds were given twenty-four hours of LOGO training over a three-week period. Students learned the editing and command generation skills prerequisite to debugging but were not able to interpret commands and use clues to identify, locate, and correct bugs. We conclude by discussing objectives for teaching the models debugging skills directly.

Code comprehension problems as learning events

Code comprehension problems have been shown to be effective assessment items in computer science education. In this paper we present qualitative and quantitative results of a study evaluating the effectiveness of code comprehension questions with feedback as learning events. Students taking an introductory programming course that satisfies a university requirement interacted with an online tutoring system using code comprehension problems about simple array algorithms as a part of a homework assignment. Students answered the problems in their own words first, before selecting a multiple choice option from the system. Both the open-ended and multiple-choice responses were collected and analyzed. Results indicate that code comprehension questions with appropriate feedback can be learning events. The use of open-ended and multiple choice responses to the same question is also shown to be useful in refining distracter items for future assessment. Recommendations from this study can be applied not only to tutoring systems, but also to the type of interactions used in worked examples in class lecture and textbook production.

Using Tutors to Improve Educational Games: A Cognitive Game for Policy Argument

How might we balance assistance and penalties to intelligent tutors and educational games that increase learning and interest? We created two versions of an educational game for learning policy argumentation called Policy World. The game (only) version provided minimal feedback and penalized students for errors whereas the game+tutor version provided additional step-level teaching feedback and immediate error correction. A total of 105 university students played either the game or game+tutor version of Policy World in a randomized, controlled, two-group, between-subjects experiment, during which we measured students’ problem-solving abilities, interest in the game, self-reported competence, and pre- and posttest performance. The game+tutor version increased learning of policy analysis skills and self-reported competence. A path analysis supported the claim that greater assistance helped students to learn analysis better, which increased their feelings of competence, which increased their interest in the game. Log data of student behavior showed that debate performance improved only for students who had sufficiently mastered analysis. This study shows that we can design interesting and effective games to teach policy argumentation and how increasing tutoring and reducing penalties in educational games can increase learning without sacrificing interest.