Table 1.
Summary of desirable properties of a good instructional design with corresponding suggestions to improve human training protocols for BCI.
| Level | Properties of a good instructional design | Corresponding suggestions for BCI training protocols |
|---|---|---|
| Feedback | - Non-evaluative and supportive feedback (Hattie and Timperley, 2007; Shute, 2008) | Provide positive feedback (feedback only indicating when the user did right) only for beginners, and disconfirmatory feedback for advanced users |
| - Feedback that conducts to a feeling of competence (Ryan and Deci, 2000) | ||
| - Clear and meaningful feedback (Hattie and Timperley, 2007) | Start with a subject-independent classifier for users with poor initial performances | |
| - Explanatory and specific feedback (Hattie and Timperley, 2007; Shute, 2008) (Moreno and Mayer, 2007) | Provide more information about what was right or wrong about the EEG patterns produced by the user: | |
| - Feedback that signals a gap between current and desired performances (Hattie and Timperley, 2007; Shute, 2008) | - Provide as feedback the value of a few (less than seven) relevant EEG features | |
| - Provide as feedback some measure of quality of the mental imagery | ||
| - Multimodal feedback (Ainsworth, 2006) (Merrill, 2007) | Provide a multimodal feedback (e.g., visual + haptic), with the same granularity and specificity for each modality, with some redundancy between them | |
| - Engaging feedback and environment (Ryan and Deci, 2000) | Represent the feedback as an interaction with a game element (e.g., a 3D car) | |
| Instructions | - Goals should be clearly defined (Hattie and Timperley, 2007; Shute, 2008) | Expose the real goal of BCI training, i.e., to produce clear, specific and stable EEG patterns |
| - The meaning of the feedback should be explained (Ainsworth, 2006) | Explain what the BCI feedback means, particularly for non-intuitive feedback such as the classifier output. | |
| - Prior knowledge should be activated (Merrill, 2007; Moreno and Mayer, 2007) | - Instruct the users to remember situations in which they used the task they will imagine | |
| - The skill to be learned should be demonstrated (Merrill, 2007) | - Demonstrate successful BCI use and BCI feedback during correct task performance | |
| Tasks | - Progressive and adaptative tasks (Ainsworth, 2006; Merrill, 2007) | Use adaptive BCI training protocols with increasing difficulty (e.g., progressively increasing the number of mental tasks to be mastered) |
| - Tasks that are challenging but still achievable (Hattie and Timperley, 2007; Shute, 2008) | ||
| - Need for autonomy and work at the user's own pace (Ryan and Deci, 2000; Shute, 2008) (Moreno and Mayer, 2007) | Include more training sessions with free and/or self-paced BCI use | |
| - Motivation and positive emotions promote learning (Ryan and Deci, 2000; Um et al., 2012) | Using positive emotion-inducing training tasks e.g., including gaming mechanisms | |
| - Need for variability over tasks and problems (Sweller et al., 1998; Ainsworth, 2006) | Include variety in the mental tasks to be performed, e.g., change in speed or duration of the mental imagery | |
| - Adapt the training procedure to the student (Hattie and Timperley, 2007; Shute, 2008) | Matching BCI training protocols to users' characteristics |
It should be noted that such suggestions are only based on theory, and will need to be formally validated.