“Jigsaw” activities don’t work

Maybe there is a way to make them work, but I haven’t seen it yet. 

A jigsaw activity as I have experienced it involves a group of people all needing to learn the same thing. The new learning is divided by the facilitators into some number of discrete pieces.

Suppose there are four different components to a concept or skill that participants want to learn about. Each of those components becomes a station in the room. The learners are then divided into groups of at least four, and each person within the group is assigned one of the four stations to become an “expert” at that component. 

The participants scatter to their stations, and they engage in dialogue to become experts at their concepts or skills. They then return to their home group to share their learning with their peers.

The trouble comes here. The experts have had a lot of time to think and reflect upon a concept or skill, while the remaining members of the home group have to simply accept and absorb each experts’ final learning. 

The deep learning comes when working through a concept, not by simply observing or hearing it. Instead of a jigsaw, the participants might as well simply read an article with the “answer.” It would be more efficient if acceptance was the goal. 

But the learning is in the work, not in the receipt of knowledge, so each person needs to be part of each expert group. 

If this is true, jigsawing is counterproductive. 

Am I missing something? Are we just doing it wrong?