Artificial intelligence in formative assessment
- Author
- SEB Education
- Category
- Technology and Learning
- Reading time
- 7 min read
Share
Formative assessment happens during learning, not at the end of it. Its goal is not to give a grade but to find out what students have already understood and what needs another explanation — in time to change the next lesson.
The challenge has always been scale. In a class of thirty students, following each one’s reasoning on every exercise is humanly impossible. That is where artificial intelligence tools start to help.
What technology does well
Adaptive platforms can spot error patterns: the student who always forgets to invert the fraction, the group that confuses cause and effect in reading. Instead of a pile of tests to mark, the teacher receives a map of the class.
With that map, the next lesson can start with what actually got stuck, and support groups can be formed by type of difficulty rather than by final grade.
What remains the teacher’s job
No algorithm knows that a student did poorly because they slept badly, or that another got everything right because they copied. Data points the way; the teacher, who knows the class, interprets it. That is why, across the ecosystem’s schools, technology supports pedagogical decisions and never replaces them.
Data protection also matters. Information about children’s and teenagers’ performance is sensitive, and any tool adopted must comply with Brazil’s data protection law (LGPD) and the school’s privacy policies.