Chapter 8 (Spüler et al., 2017) reports on the use of adaptive brain-computer interfaces in learning contexts. These interfaces try to assess the mental workload of a learner during task performance through online measurement of brain activity (electroencephalography). On the basis of these “neural signatures,” the system then automatically adapts the difficulty level of arithmetic exercises to the assessed mental workload. As these adaptations take place without the learner noticing them, the brain-computer interfaces described in this chapter provide a nice example of directive adaptivity.

In this chapter, we present recent developments to utilize Brain-Computer Interface (BCI) technology in an educational context. As the current workload of a learner is a crucial factor for successful learning and should be held in an optimal range, we aimed at identifying the user´s workload by recording neural signals with electroencephalography (EEG). We describe initial studies that identified potential confounds when utilizing BCIs in such a scenario. Taking into account these results, we could show in a follow-up study that EEG could successfully be used to predict workload in students solving arithmetic exercises with increasing difficulty. Based on the obtained prediction model, we developed a digital learning environment that detects the user´s workload by EEG and automatically adapts the difficulty of the presented exercises to hold the learner´s workload level in an optimal range. Beside estimating workload based on EEG recordings, we also show that different executive functions can be detected and discriminated between based on their neural signatures. These findings could be used for a more specific adaptation of complex learning environments. Based on the existing literature and the results presented in this chapter, we discuss the methodological and theoretical prospects and pitfalls of this approach and outline further possible applications of BCI technology in an educational context.