Goals vs. Instruction

I'm reading an article entitled "Teaching and Learning Argumentative Reading and Writing: A Review of Research" by Newell et al, and I (really quickly) wanted to record a bit of research they discussed:

When having students write argumentative essays, several studies found that students with GOALS wrote better essays than students without goals; for example, half the kids had specific goals to cater towards their audience, or to refute the counterargument, etc. and the other half had no specific goal. They also found (separate study) that students with more elaborate, specific goals wrote better than students with vague goals.

Finally, a study was done where half students made goals and the other half did not; out of the students who made goals, half were given instruction on how to formulate an argument, and the other half were not. The goals group, as a whole, did better than the no-goals group. There was no difference between the students who received instruction and those who did not.

I've read several studies now for this class that concluded that instruction doesn't really make much of a difference. One of the other articles with this theme found that students who practiced collaborative reasoning (kind of like the Socratic Seminars we did at Rudy) performed BETTER on an argumentative essay than students who practiced collaborative reasoning AND had direct argumentation instruction. (Well, kinda: it's complicated. The CR only kids had more ideas, and more and better quality arguments. The CR + lesson kids had fewer ideas, but their arguments were more organized. What's the compromise that fosters tons of great ideas AND organization?)

So basically, if you teach it, they won't learn. If you let them think it through, they will learn.