ISSUES
"It is unfortunate to say that in times like this, racism and discriminations are embedded in our education system, so developers need to make efforts on developing algorithm and datasets which are not racial."
​
- Balaganur, 2019
Student Data
Schools already collect large amount of data that is stored in Student Information Systems. This data can range from grades to behavioural notes and parent place of employment. The issue with student data arises when parents are not informed about what data is collected and how it will be used. Every country has different laws related to student data. These laws should be reviewed before collecting or sharing student data.
Algorithm Bias
Predictive technology has made its way into the justice system, retail and more recently education. However, AI in the education sector faces specific issues that require more diligence and further exploration (Balaganur, 2019). Algorithms will mimic the same biases that are prevalent in the data. For example, if teachers tend to report more behavioural issues by minorities than the algorithm will interpret the data as such (Chastel, 2018). The designer of the algorithm must demonstrate that they have looked for sources of bias and removed them in order to provide an accurate assessment (Walker, 2018). No predictive algorithm will be perfect, but it must consider the potential of algorithm bias when making decisions, especially when those decisions impact students’ lives.
Ethics
Is it ethical to intervene with a student’s educational path based on a predication made by data?
The purpose of predictive algorithms is to make a positive impact on the education system by being more efficient and saving time. However, it has to be recognized that the wrong prediction may be made. Algorithms are not perfect and should be taken into consideration that the final decision should be made by a human being, at least for the time being.