Skip to main content Skip to navigation

Integrating Natural Language Prompting Tasks in Introductory Programming Courses

Project Overview

The document explores the role of generative AI in education, particularly through the implementation of Natural Language Prompting tasks in introductory programming courses. It underscores how these tasks can transform the learning experience by shifting the focus from rote syntax mastery to a more engaging problem-solving approach, thereby making programming accessible to a broader and more diverse range of students. The findings reveal that students who typically face challenges in traditional programming assessments tend to perform better with natural language tasks, suggesting a significant pedagogical shift is necessary to adapt to the evolving educational landscape influenced by AI technologies. This evidence highlights the potential of generative AI not only to enhance learning outcomes but also to foster inclusivity and engagement among students who may otherwise struggle with conventional programming methodologies. Overall, the document advocates for a reevaluation of teaching strategies in the context of AI advancements to better support student learning and success.

Key Applications

Natural Language Prompting Tasks for coding

Context: Introductory programming course for engineering students with varying levels of prior experience.

Implementation: Incorporated two types of prompt-focused activities into lab sessions: EiPE (Explain in Plain English) and Prompt Problems.

Outcomes: Students reported improved understanding of code logic and communication about programming tasks. Performance on natural language tasks was less correlated with self-reported difficulty, suggesting they assess different skills compared to traditional assessments.

Challenges: Students initially found crafting effective prompts challenging. Some perceived the tasks as gimmicky.

Implementation Barriers

Skill barrier

Students struggled to form successful prompts and understand generated code, which hinders effective interaction with LLMs.

Proposed Solutions: Explicit teaching of prompt engineering and iterative learning strategies.

Perception barrier

Some students viewed natural language tasks as gimmicky and less effective than traditional programming tasks.

Proposed Solutions: Positive feedback from other students suggests reinforcing the pedagogical value of these tasks.

Project Team

Chris Kerslake

Researcher

Paul Denny

Researcher

David H Smith IV

Researcher

James Prather

Researcher

Juho Leinonen

Researcher

Andrew Luxton-Reilly

Researcher

Stephen MacNeil

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Chris Kerslake, Paul Denny, David H Smith IV, James Prather, Juho Leinonen, Andrew Luxton-Reilly, Stephen MacNeil

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies