I want to first split my sample into two different subgroups: a first group includes only people that were rather satisfied with their current learning process, and another subgroup only including learners that are rather not satisfied with their current learning process. Then, I want to find out with what probability learners of both groups faced a motivational vs. a cognitive challenge during their learning process. Then, I want to find out with what probability each learner used a specific type of strategy to regulate a given problem/challenge (e.g. encouragement strategies, organization strategies, and so on). Is this a classification or regression procedure? Each participant mentioned only a motivational OR a cognitive problem he or she faced, but each participant mentioned more than one strategy type to regulate the specific problem (e.g. participant 2 was satisfied, he faced a motivational problem, and he regulated it with two encouragement strategies and with three evaluation strategies). [![enter image description here][1]][1] [1]: https://i.sstatic.net/Eqq2A.png