Have been working under the assumption that reduction of stress, and an associated a sense of perceived control, can increase the chance of a student participating – and learning.
Based on that assumption, we often discuss the mistakes we – and many now famous people – have made. That appears to reduce the expectation of what the students need to do to be ‘perfect,’ and so open the doors to growth.
We discuss studies of learning and memory that can reduce stress – including taking breaks, using sleep to aid understanding, ways to replicate learning contexts, and the advantages of overlearning, given how stress can limit cognitive skills during an exam.
Also, there are more participation points for disagreeing with what is being taught. That really helps. Encouraging disagreement does something for the students. It leads to discussions and full-blown arguments (non-confrontational so far) in which the faculty plays no role for a while.
Always, surprising them with weird and original knowledge and ideas seems to help them forget their problems for a while and engage in the problem, without the worry of what it means in terms of mere grades.