Students thought processes while engaged in computer programming

Graduation date: 1993The purpose of this qualitative study was to investigate\ud the thought processes of secondary level novice programmers\ud engaged in computer programming for the purpose of\ud generating hypotheses for consideration in future research\ud on the relationship between computer programming and problem\ud solving. A high school BASIC programming course with\ud 14 students from a single school in the tenth through the\ud twelfth grades was selected for the sample.\ud Data describing students' thought processes while programming\ud were collected during double periods in the 11th\ud and 16th weeks of the fall semester. Students worked in\ud role-assigned partnerships, wherein one student was the\ud problem solver and the other was the recorder. The problem\ud solver's task was to solve the problem using a "think\ud aloud" strategy, while the recorder took notes describing\ud the problem solver's actions to assure that audiotape recordings\ud of the problem solver's voice were maintained.\ud Following the solution of one problem, these roles were\ud switched.\ud Analysis of novice programmers' thought processes revealed\ud two categories of student problem solution strategies:\ud coded thinking and debugging. In the coded thinking\ud strategy, students approached the problems primarily\ud from the perspective of BASIC codes. This strategy was\ud similar in nature to activities involved in verbal association\ud learning, a low level thinking strategy identified by\ud Gagne (1970). Students relied on two techniques for debugging\ud syntax and logic errors. They applied a guess-and-check\ud technique to correct syntax errors or asked the\ud teacher for assistance. Similarly, when logic errors were\ud revealed, the subjects typically asked the teacher for\ud assistance and then used the guess-and-check technique to\ud correct the errors. Both techniques utilized lower level\ud thought processes than that required for problem solving\ud learning. Analysis of the subject programming processes\ud revealed that problem solving processes, as identified by\ud Polya (1988), were not involved. Future research should\ud examine students thought processes when working with a compiled\ud language such as Pascal. In addition, future research\ud should investigate the thought processes of students\ud who have had more experience than a single term of programming.\ud A case study of from two to three students explored\ud over a longer period of time may provide a clearer description\ud of student thought processes