2024-07-02 15:00:19
The use of artificial intelligence in education to mark exams represents an important advance in the way teachers handle student papers. AI will improve the efficiency, fairness, and speed of remediation, but this may raise questions regarding its impact and limitations.
Artificial Intelligence: Improving Efficiency
AI is able to analyze large amounts of data in record time and, based on machine learning algorithms and examples of transcripts corrected by teachers, will be able to identify patterns and structures in student responses. This will allow you to correct not only closed answers, but also essays and theses by assessing the coherence, argumentation and quality of the language used.
Texas Example
Texas recently implemented a new scoring system for its annual exam (STAAR) that incorporates AI to help with exam grading. The state increased the number of open questions on the exam, which typically requires more manpower and higher costs. To overcome this, the state set up a hybrid system where the AI is designed to correct all open questions, then one-quarter of the corrections are reviewed by human teachers, and copies that present uncertainty or questionable results are also checked by teachers.
Using AI to reduce the number of proofreaders from 6,000 to 2,000 would save $15 million to $20 million per year, but in some cases AI may not be able to properly evaluate certain copies, especially in creative writing, and subsequently receive better ratings than those of human proofreaders.
A fairer assessment?
The main advantage of AI is its objectivity. In fact, unlike human graders, AI is not affected by fatigue, unconscious bias or mood swings, which can affect teachers’ decisions on how to grade papers. AI can provide uniform assessments and, only if well-programmed, can reduce inequalities between students.
Challenges and limitations
Despite its many benefits, elements such as nuance and creativity in student responses may escape the algorithm in some cases. In addition, ethical issues may arise in terms of transparency of grading criteria and protection of student data, and it is necessary to regularly update AI systems so that they meet educational standards and respect the cultural and linguistic diversity of students.
Future Outlook
AI will not replace teachers, but will support them, by automating assessments for the first time, allowing teachers to focus on the qualitative aspects of proofreading. Incorporating AI into proofreading can also transform assessment systems, making them more interactive and adaptive, thereby enhancing students’ analytical and critical skills.
This hybrid approach, combining machine rigor and human intelligence, may become the ideal correction model in the coming years. Therefore, AI applied to exam marking will allow for fairer, faster, and more targeted assessments, but implementation must be balanced to maximize its benefits and minimize the risk of error to ensure that the assessment is complete and fair.
1720030758
#Artificial #Intelligence #Test #Grading #Potential #Breakthrough #Digital #Economy #Blog
