Objective Assessment in Java Programming Language Using Rubrics

Abstract

Aim/Purpose: This paper focuses on designing and implementing the rubric for objective JAVA programming assessments. An unsupervised learning approach was used to group learners based on their performance in the results obtained from the rubric, reflecting their learning ability. Background: Students' learning outcomes have been evaluated subjectively using a rubric for years. Subjective assessments are simple to construct yet inconsistent and biased to evaluate. Objective assessments are stable, reliable, and easy to conduct. However, they usually lack rubrics. Methodology: In this study, a Top-Down assessment approach is followed, i.e., a rubric focused on the learning outcome of the subject is designed, and the proficiency of learners is judged by their performance in conducting the task given. A JAVA rubric is proposed based on the learning outcomes like syntactical, logical, conceptual, and advanced JAVA skills. A JAVA objective quiz (with multiple correct options) is prepared based on the rubric criteria, comprising five questions per criterion. The examination was conducted for 209 students (100 from the MCA course and 109 from B.Tech. course). The suggested rubric was used to compute the results. K-means clustering was applied to the results to classify the students according to their learning preferences and abilities. Contribution: This work contributes to the field of rubric designing by creating an objective programming assessment and analyzing the learners’ performance using machine learning techniques. It also facilitates a reliable feedback approach offering various possibilities in student learning analytics. Findings: The designed rubric, partial scoring, and cluster analysis of the results help us to provide individual feedback and also, group the students based on their learning skills. Like on average, learners are good at remembering the syntax and concepts, mediocre in logical and critical thinking, and need more practice in code optimization and designing applications. Recommendations for Practitioners: The practical implications of this work include rubric designing for objective assessments and building an informative feedback process. Faculty can use this approach as an alternative assessment measure. They are the strong pillars of e-assessments and virtual learning platforms. Recommendation for Researchers: This research presents a novel approach to rubric-based objective assessments. Thus, it provides a fresh perspective to the researchers promising enough opportunities in the current era of digital education. Impact on Society: In order to accomplish the shared objective of reflective learning, the grading rubric and its accompanying analysis can be utilized by both instructors and students. As an instructional assessment tool, the rubric helps instructors to align their pedagogies with the students’ learning levels and assists students in updating their learning paths based on the informative topic-wise scores generated with the help of the rubric. Future Research: The designed rubric in this study can be extended to other programming languages and subjects. Further, an adaptable weighted rubric can be created to execute a flexible and reflective learning process. In addition, outcome-based learning can be achieved by measuring and analyzing student improvements after rubric evaluation.

Student Outcome Assessment on Structured Query Language using Rubrics and Automated Feedback Generation

“…The criteria that have been chosen are grounded in Bloom's Taxonomy of Learning Domains. The student learning outcome can be assessed in to six levels such as, remember, understand, apply, analyze, and evaluate [29] [30].…”

Identification of Students with Similar Performances in Micro-Learning Programming Courses with Automatically Evaluated Student Assignments

“…The following papers used clustering analysis in close connection with programming. Gupta et al [16] proposed a rubric based on the learning outcomes like syntactical, logical, conceptual, and advanced JAVA skills. K-Means clustering was applied to the results to classify the students according to their learning preferences and abilities.…”