Wednesday, October 09, 2024

Promoting authentic learning experiences: redesigning assessments to minimise student misconduct

Webinar organised by TurnitIn and hosted by ASCILITE. 

Summary of the webinar's objectives as:

 1. Authentic assessment design: Dive into the principles of authentic assessment design, which emphasises real-world relevance, meaningful tasks and opportunities for creativity and critical thinking. Explore methodologies such as project-based assessments, case studies and simulations that provide students with authentic learning experiences while reducing the incentive to cheat. 

2. Formative feedback and self-assessment: Discuss the importance of formative feedback and self-assessment in promoting academic integrity. Learn the role of continuous assessment methods, peer review and reflective exercises in fostering student engagement, metacognitive skills and ownership of learning outcomes. 

3. Assessment security measures: Explore security measures and deterrents to minimise temptations for cheating in assessments. Discuss strategies such as randomised question banks, timed assessments and proctoring technologies that mitigate opportunities for academic misconduct while preserving the integrity of the assessment process.

Notes taken:

Kwong Nui Sim representing ASCILITE welcomes participants and runs through zoom functions. Includes a Welcome to country and an overview of ACSCILTE and the annual conference this year in Melbourne in December.

Chukwudi Ogoh from Turnitin chairs the panel. Diagnostic poll starts to provide information from participants to the panel. Academic integrity still an important focus for participants. Many use portfolios or similar to reduce academic integrity.

Speaker introduced. Joon Nak Choi from Hong Kong University of Science and Technology begins with the view that there is too much focus on academic integrity with slow integration of AI into teaching and learning although students are racing ahead with using AI. Formerly, UK promoted values based education, Germans then introduce skills based education but presently knowledge based education is the main focus. However, Ai is challenging knowledge based education therefore it is important to return to value and skills based education. Therefore critical thinking, adaptability and judgement should be the fundamental objectives of learning through active learning to raise metacognitive awareness. Formative assessments are important. Essays can remain relevant and useful as they help assess critical thinking, logic and argument and application. Metacognition, adaptability, initiative and self direction can also outputs from essays. In the post-AI classroom, essays still possible by putting them through turnitin, or writing essays in class, and integration of AI into essay writing reflects workplace use.  AI assistants can be used to grade in class essays to provide timely feedback, allowing the essays to be used as a formative tool. Professor or teaching assistants can then use their time to personalise the feedback. Teach students how to use AI to brainstorm, first draft from AI with student going through to correct/triangulate, presentation on the topic required, and the final draft incorporates human feedback backed by reflection memos.

Dr. Dianne Stratton-Maher from University of Southern Queensland on embracing AI to reimagine assessment in nursing education. Teaches first year Bachelor of Nursing course on literacies and communication. Overall perspective is that AI is not a replacement for personal knowledge, but rather they are tools that can strengthen and enhance the learning journey. Undertook a qualitative descriptive design to integrate AI into the course. Case study used and students were introduced to ChatGPT to support the completion of the case study assignment. AI generated video to introduce AI and provide some AI literacy on how to use AI to support learning. A reflection was also required as to how using AI went and this was handed in as an appendix to their case study. The reflection included prompts used, responses received, evidence of evaluation of the responses etc. Shared examples of student work. 409 students submitted, 18 reported for AI but 10 were unsubstantial. Collected f2f and focus group data, using thematic analysis supported by CoPilot (with UniAQ data protection). Themes included confusion and adaption; functionality and skill development; effectiveness and reliability; academic integrity and ethical use, perceived benefits and future use; and feedback and support. Recommendations included the need to clarify guidelines and expectations; promote responsible use; providing training and support; encouraging critical evaluation; highlighting the benefits; integrate practical examples; monitor and adapt and education for staff.

Associate Professor Benito Cao from the University of Adelaide on 'don't ne sorry, just declare it'. No assessment re-design but focused on working with students on how ChatGPT is not reliable but can be useful in some aspects of academic work, Therefore, is ChatGPT used, its use is declared in an appendix. ChatGPT is not a search engine but a language model. Asked ChatCPT to generate a profile of himself and shared with students. Almost all the items were wrong! but the profile was written in a convincing manner. Summarised his university's guidelines including Deputy Vice Chancellor Academic to all students to encourage them to embrace and integrate AI but to use it with integrity. Assessment guidelines define how AI can be used - for example AI assisted ideas or AI assisted editing. Discussed the benefits, limitations and risks and risk management for Gen AI as provided but the South Australian Government. Industry guidelines also introduced, for example guidelines from Taylor and Francis allowing idea generation and exploration; language improvement; interactive online search; literature classification and coding assistance. AI must not be listed as author and its use formally acknowledged. Stressed to students the need to keep their voice and appendix recording use of prompts etc. required. 

Jane Mair provided an overview to Turnitin's approach to minimise cheating. A challenge presented by paper-based handwritten assessments. Short answers, multiple choice, sketches or labelling diagram, mathematics problems. Paper assessments persist as they align with authentic design, formative feedback can be provided that engages students and supports assessment integrity by limiting access to AI writing and other digital tools. Challenges for instructors, marking can be time-consuming, coordination and consistency for grading can be difficult, feedback inefficient, heavy reliance on multiple choice which may not assess relevant learning outcome and challenges in applying assessments for best practice. For students receive delayed feedback, some have difficulty understanding the feedback and there are increased errors and there may be need for clarification. 

Turnitin feedback now has the 'paper to digital' overview. Grading is faster using online grading, feedback is more timely, secure workflow and digital storage, and integration directly into LMS and Turnitin feedback studio. Supports long-form writing, multiple choice, short answer, math formulas etc. Demonstrated how it worked. Horizontal questions possible allowing each question to be marked across the cohort, rather than one student at a time. Similar answers can be graded at once! Feedback studio allows for an efficient workflow.

A higher education focus in all the presentations. Some good ideas if essays are still being used. Webinar is recorded and presentations will be available. 









No comments: