Latent Learning for Production Engineers

If you do much dog training, you appreciate the power of latent learning. Good trainers know that drilling a dog over and over on a skill in a single training session can have all sorts of negative consequences: the dog learns lots of different wrong ways to do the skill (learns by repetition, possibly even by poorly timed reinforcement), the dog gets bored, the dog starts getting worried about repeated failures and shuts down.

I've mentioned that Mimi is a training sponge. By this I mean that she will happily train without a break for 20 minutes, 30 minutes, even longer. But even if I do maintain a training session with her for so long (rarely, but it does happen), I still don't drill the same thing over and over. Once she has a successful performance of a new skill, I'll move on to something else right away. (It is so difficult to resist the lure of "just one more time" but with experience and practice I'm getting better at it.)

The latent learning comes in at the next training session--that evening, the next day, perhaps a few days later--when I ask her to perform that particular skill again. Well-timed reward of a successful performance plus a bit of time can solidify knowledge and understanding of the skill. The next time the dog attempts it, she'll perform it correctly right away or at least with fewer attempts to get it right.

I am currently teaching a class on petroleum geology to a group of young Saudi production engineers. I selected the content and designed the exercises and am now delivering it for the first time. Yesterday, I turned them loose on a big case study that requires them to integrate core and wireline log data to characterize a particular reservoir. This workflow is one that geologists in Aramco apply. The outcomes of the workflow are used by reservoir engineers so the goal is for them to better understand where the stuff they use comes from. The class, divided into teams, worked on the exercise for about 3 hours yesterday afternoon. Most were clearly struggling to make sense of the data and to get their heads around even the statement of the problem to be solved. However, I made sure to spend a lot of time providing positive support for their efforts and questions, even if they started out chasing the wrong things.

I told them to go home and "sleep on it."

Two-thirds of the class were at it early this morning even before class started. I've made three passes through the room and there have only been a handful of questions focused on clarifying individual points. It is clear that they are no longer wrestling with the larger issues of integration of the data and making sense of the geology. They may not have the right answer but each team appears to have a plan for their presentation.

What a difference 8 hours makes. Well-timed reinforcement and a bit of time gives us latent learning in action.