Hi, I learned about this strategy from Computer Science ed researcher Mark Guzdial years ago at a paper talk and used in my teaching ever since. I too am having a difficult time finding a satisfactory reference.
I know that it’s a prevalent strategy in science classes and (apparently, I learned) for reading comprehension with younger children. Science example: for physics, "I’m holding a tennis ball and a bowling ball at even heights. I’m going to drop them at the same time. predict: Which will hit the ground first?"
For programming, I would usually write some code in front of my students and before I hit run ask them to predict what will happen. I'd get a few suggestions and then ask everyone to vote. I'll tell you it really makes kids want to see the code run. The payoff comes afterward in figuring out why whatever happened did since kids now have grounds for an investigation.
I'll ask around to research-y folks for a better reference.