Automated style feedback for advanced beginner Java programmers

We created an Eclipse plug-in called FrenchPress that offers students feedback on their Java programming style. It is designed not for novices but for students taking their second or third Java course. Advanced beginner students know enough to produce a program with the desired input/output behavior...

Full description

Saved in:
Bibliographic Details
Published in2016 IEEE Frontiers in Education Conference (FIE) pp. 1 - 9
Main Authors Blau, Hannah, Kolovson, Samantha, Adrion, W. Richards, Moll, Robert
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2016
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:We created an Eclipse plug-in called FrenchPress that offers students feedback on their Java programming style. It is designed not for novices but for students taking their second or third Java course. Advanced beginner students know enough to produce a program with the desired input/output behavior, but fail to understand it could still be poorly written. Large class sizes in introductory courses make it difficult for instructors to give their students individualized attention. FrenchPress automates a small subset of the feedback students might have received from educators. The system diagnoses issues characteristic of programmers who have not yet assimilated the object-oriented paradigm, such as misuse of the public modifier, fields that should have been local variables, and instance variables that should have been class constants. We conducted a classroom trial of the plug-in covering four assignments in data structures and algorithms. Among students whose code triggered one or more diagnostic rules, the percentage who modified their program in response to FrenchPress feedback varied from a high of 59% on the first project to a low of 23% on the second and fourth projects. On the user surveys following each assignment, 56-66% of respondents said they were satisfied with FrenchPress performance.
DOI:10.1109/FIE.2016.7757728