Selcuk Acar, PhD
University of North Texas
Since the release of ChatGPT and Google Bard (now Gemini), it has become difficult to dismiss the news, debates, and controversies surrounding AI and its significant impact on many aspects of our lives. AI is poised to continue growing in influence as its power and areas of application expand. Gifted education is not exempt from the changes introduced by AI. In this brief commentary, I will discuss major areas in gifted education likely to be impacted by greater use of AI. Broadly speaking, I will address them in two areas: a) gifted identification, and b) instruction, programming, and development.
Gifted Identification
Gifted identification has long been questioned due to practices resulting in the underrepresentation of culturally, linguistically, and economically diverse students. These practices include the sole reliance on IQ tests, the gatekeeping role of teachers who may not reflect the diversity of the student body, bias in the tests used for identification, and a lack of relevance to programming and instruction. Proposed solutions to address these issues include universal screening, consideration of local norms, the use of multiple measures, and the employment of authentic and alternative assessment tools. However, these solutions also present challenges, such as the costs associated with administering the tests.
This is where AI can provide some solutions. The research we have conducted over the last several years around creativity assessment scoring has shown very promising results. Creativity tests, in general, are less prone to show bias against culturally, linguistically, and economically diverse students compared to tests of intelligence or academic achievement, as they partly rely on what students know rather than what they do not know. The major challenge of creativity test scoring lies in the scoring process itself, as these tests are typically open-ended with no correct responses and no answer key. Instead, there are some scoring guidelines aimed at increasing objectivity. However, scoring these tests requires paid training to become certified and manual time and labor costs after certification. The costs of testing are proportional to the scale of testing, making universal screening/consideration a major financial challenge for schools.
We have demonstrated in our research (Acar et al., 2024; Organisciak et al., 2023) that AI can mimic human ratings and score creativity tests in an automated manner. This means that tests could be reliably scored by AI instead of humans. With the significant reduction in scoring costs enabled by AI, the total cost of universal screening/consideration efforts is becoming more manageable, particularly in the realm of creativity assessment. This exemplifies how AI can positively replace human involvement by freeing up resources for instruction and programming rather than burdening teachers with the monotonous task of test grading.
Instruction, Programming, and Development
Developments in testing have also shown potential in using AI for programming and intervention. Let’s revisit the example of creativity. In a recent study (de Chantal & Organisciak, 2023), researchers utilized automated scoring provided by AI to offer instant feedback on participants’ performance on a creativity test. With knowledge of their performance, students receive task-specific feedback to enhance their performance. This concept extends beyond testing to everyday AI-guided tutorials, where students receive feedback based on their mistakes and potential ways to strengthen their projects or assignments. Schools are already implementing computerized tutorials, but with AI, customization and individualization are becoming more prevalent.
Another significant challenge in gifted education is the effectiveness and feasibility of differentiation. While differentiation has always received attention in our gifted education textbooks and has been successfully implemented by some teachers in certain schools and districts, consideration of individual differences, interests, and needs of a diverse student body poses a serious challenge for teachers aiming for more scalable implementation. Unfortunately, many teachers lack the time and/or training to prepare differentiated instruction. AI can assist with this challenge by providing teachers with differentiated lesson plans that consider student interests, backgrounds, and readiness levels. New tools such as GPT4Teachers (https://gpt4teachers.com/#/) process teacher input to obtain lesson plans that can be adjusted based on diverse student interests, needs, and readiness. This is just the beginning, and tools specific to gifted education specialists, coordinators, and teachers may be developed in the future.
Issues with AI Use
While the work I have summarized so far is exciting and promising, the use of AI requires close attention to problems with ethical use. Regarding assessment and identification, the human judges involved in scoring the tasks, which are then used to train the AI algorithm, are critical. If these human judges are not sufficiently representative and diverse, the AI scores may reflect biases that they (or society) may hold. On the other hand, unlike teachers who evaluate and nominate students for gifted programs, AI does not know the students who generated the evaluated responses. Furthermore, bias in AI can be detected and potentially fixed, whereas doing so with teachers is an arduous task, especially when considering the increasing teacher turnover. Thus, challenges of AI use should be compared to the existing challenges without AI to have a better understanding of the extent of progress and lack thereof.
Regarding programming and differentiation, some successful uses of AI may require specific student data (such as student background information and interests) that are then utilized for differentiation. Leakage or theft of such data is always a potential concern. Schools would need to obtain parental consent before employing AI for differentiation or other purposes. While they would assure parents that they will adhere to protocols, in today’s technology landscape, there can be no guarantees regarding the protection of data confidentiality.
Conclusion
When it comes to AI and its impact, emotions vary from excitement to worry and there are good reasons for this range of emotions. The changes in education and specifically gifted education are largely inevitable yet the scope and transition process may vary. Schools are highly structured environments, and adaptation may take longer than other aspects of our lives. AI can change gifted education in many ways, but here I discussed two ways from the lens of creativity research: identification and programming, differentiation, and instruction. There are risks associated with each of these. Like in every novelty, there is ambiguity around the ramifications of AI use in gifted education and there are risks for misuses and ethical breaches, like in any technology. This is why a major consideration in AI research is ethical uses that must be mitigated by researchers as they conduct their research and expand its applications.
References
de Chantal, P.-L., & Organisciak, P. (2023). Automated feedback and creativity: On the role of metacognitive monitoring in divergent thinking. Psychology of Aesthetics, Creativity, and the Arts. Advance online publication. https://doi.org/10.1037/aca0000592
Acar, S., Dumas, D., Organisciak, P., & Berthiaume, K. (2024). Measuring original thinking in elementary school: Development and validation of a computational psychometric approach. Journal of Educational Psychology. Advance online publication. http://dx.doi.org/10.1037/edu0000844
Organisciak, P., Acar, S., Dumas, D., & Berthiaume, K. (2023). Beyond semantic distance: Automated scoring of divergent thinking greatly improves with large language models. Thinking Skills and Creativity, 49, 101356. https://doi.org/10.1016/j.tsc.2023.101356
Gifted Education Family Network 