Lesson 2 - Measuring affect in education and industrial settings

SUMMARY

In the previous lesson, we established that learning is shaped not only by what learners do, but also by how they feel. This lesson takes the next step: how can those affective states be measured in practice?

Measuring affect is far from straightforward. Emotions, engagement, and learning states cannot be observed directly, and every method used to capture them involves assumptions, trade-offs, and limitations. In educational and industrial settings, these challenges are further amplified by time constraints, learner stress, and ethical responsibilities.

In this lesson, you will explore the main data sources used to measure affect, including self-reports, behavioural data, and physiological or multimodal signals. You will learn how commonly used questionnaires align with different affect models, why some tools are appropriate in certain contexts but not others, and how poor measurement design can lead to biased or unreliable data. The lesson also highlights practical and ethical considerations that are especially relevant in vocational training and workplace learning.

By the end of this lesson, you will be equipped to think critically about affect measurement, not as a purely technical task, but as a design decision that must balance validity, feasibility, and responsibility. This foundation will prepare you to use affective information meaningfully and ethically in adaptive learning systems.

About The Author

Dr Enrique Hortal is an Assistant Professor in the Department of Advanced Computing Sciences at Maastricht University. His academic and research work focuses on affective computing, brain–computer interfaces, and machine learning for the analysis of physiological and brain signals. He brings extensive experience in computational intelligence and human-centred AI, combining theoretical foundations with practical applications in emotion recognition and intelligent systems.