Preference for debugging strategies and debugging tools and their relationship with course achievement: Preliminary results of a study involving novice programmers.

Laura Melissa Cruz Castro; Jenny Patricia Quintana-Cifuentes; Akash Kumar

Download Paper | Permalink

Conference: 2023 ASEE Annual Conference & Exposition
Location: Baltimore , Maryland
Publication Date: June 25, 2023
Start Date: June 25, 2023
End Date: June 28, 2023
Conference Session: Computing and Information Technology Division (CIT) Technical Session 1
Tagged Division: Computing and Information Technology Division (CIT)
Page Count: 18
DOI: 10.18260/1-2--43938
Permanent URL: https://peer.asee.org/43938
Download Count: 165

Request a correction

Paper Authors

biography

Laura Melissa Cruz Castro University of Florida

visit author page

Laura Melissa Cruz Castro is an instructional assistant professor in the engineering education department at the University of Florida.

visit author page

biography

Jenny Patricia Quintana-Cifuentes University of Louisiana at Monroe

visit author page

Jenny Quintana is an assistant professor at the University of Louisiana Monroe. Dr, Quintana has a Ph.D. in Engineering Education from Purdue University and two master's degrees from Purdue University. One is in Technology leadership and innovation, and the other is environmental and ecology engineering. Dr. Quintana completed her undergraduate studies in Technological Design in Bogota, Colombia.

visit author page

author page

Akash Kumar

Download Paper | Permalink

Abstract

More than 50% of the time spent in building solutions is spent debugging software. Despite the importance of debugging in software development, how students debug, and the strategies and reasoning students use for debugging software are still not clear. This work in progress will present preliminary results on students’ preferred debugging strategies and compare them with their learning gains during a programming course. We focus on answering the following questions: a) “Is there a difference between students’ preference for debugging strategies and their course achievements?”; b) “Is there a change in debugging strategies preferences across time for novice programmers?”; c) “Is there a relationship between software debugging tools and the conceptual understanding of debugging strategies?”

This study is conducted during the Fall of 2022 in a 16-week programming fundamentals II course at a large public southwestern university. This semester, 328 students enrolled from a variety of engineering and computer science majors. Our data was gathered from a debugging assignment, which is an open-ended questionnaire. The questionnaire’s open-ended items aim to uncover students’ thought processes when helping others to debug their code and students’ strategies when debugging their own code. A coding book was developed to capture students' utterances and classify them into Metzger’s debugging strategies: a) incremental development, b) program slicing, c) sanity checks, d) error variables for controlling behavior, e) cause elimination methods, f) turning debugging code on and off and g) traceback. A logistic regression model was conducted to identify which of the debugging strategies relate the most to course achievements. The response variable for the logistic regression corresponds to whether the students’ final grades are A or not. After obtaining the model, the goodness of fit tests will be performed, as well as an analysis of the coefficients corresponding to the strategies.

Through our analysis of the open-ended responses, our study sought to identify the debugging strategies that students use in relation to their learning gains. Through the information collected some strategies emerged as more aligned with the student's expertise. Furthermore, this study helps uncover the students’ awareness of the different existing debugging strategies. Future analysis can illuminate the relationships between software debugging tools and students’ understanding of the underlying concepts and processes behind debugging strategies. Study design, data collection, and preliminary data are presented.

Citation
Format

Cruz Castro, L. M., & Quintana-Cifuentes, J. P., & Kumar, A. (2023, June), Preference for debugging strategies and debugging tools and their relationship with course achievement: Preliminary results of a study involving novice programmers. Paper presented at 2023 ASEE Annual Conference & Exposition, Baltimore , Maryland. 10.18260/1-2--43938

TY - CPAPER
AB - More than 50% of the time spent in building solutions is spent debugging software. Despite the importance of debugging in software development, how students debug, and the strategies and reasoning students use for debugging software are still not clear. This work in progress will present preliminary results on students’ preferred debugging strategies and compare them with their learning gains during a programming course. We focus on answering the following questions: a) “Is there a difference between students’ preference for debugging strategies and their course achievements?”; b) “Is there a change in debugging strategies preferences across time for novice programmers?”; c) “Is there a relationship between software debugging tools and the conceptual understanding of debugging strategies?”

This study is conducted during the Fall of 2022 in a 16-week programming fundamentals II course at a large public southwestern university. This semester, 328 students enrolled from a variety of engineering and computer science majors. Our data was gathered from a debugging assignment, which is an open-ended questionnaire. The questionnaire’s open-ended items aim to uncover students’ thought processes when helping others to debug their code and students’ strategies when debugging their own code. A coding book was developed to capture students&#39; utterances and classify them into Metzger’s debugging strategies: a) incremental development, b) program slicing, c) sanity checks, d) error variables for controlling behavior, e) cause elimination methods, f) turning debugging code on and off and g) traceback. A logistic regression model was conducted to identify which of the debugging strategies relate the most to course achievements. The response variable for the logistic regression corresponds to whether the students’ final grades are A or not. After obtaining the model, the goodness of fit tests will be performed, as well as an analysis of the coefficients corresponding to the strategies.

Through our analysis of the open-ended responses, our study sought to identify the debugging strategies that students use in relation to their learning gains. Through the information collected some strategies emerged as more aligned with the student&#39;s expertise. Furthermore, this study helps uncover the students’ awareness of the different existing debugging strategies. Future analysis can illuminate the relationships between software debugging tools and students’ understanding of the underlying concepts and processes behind debugging strategies. Study design, data collection, and preliminary data are presented.
AU - Laura Melissa Cruz Castro
AU - Jenny Patricia Quintana-Cifuentes
AU - Akash Kumar
CY - Baltimore , Maryland
DA - 2023/06/25
PB - ASEE Conferences
TI - Preference for debugging strategies and debugging tools and their relationship with course achievement: Preliminary results of a study involving novice programmers.
UR - https://peer.asee.org/43938
DO - 10.18260/1-2--43938
ER -