Part One:
I've stopped recording my dreams for a little over a week after a bout of depression. While being depressed my dream recall has got significantly worse. However I can still remember a few dreams without writing them down.
I have amassed very few trends from my dream experiences. But I want to talk about correlations.

People:
Friends who I see all the time in real life can appear in almost any situation. They are as mobile as I am.
However, people I see less often tend to appear in the setting in which I see them most often.
People who I rarely see and who do not have a frequent location tend to appear when I am somewhere new or unrecognisable(like a theatre) or on public transport.
People who I know for their proffession appear in places of association e.g. teachers, tend only to appear in schools. (supporting schema idea).
People who I do not recognise at all sometimes appear, but my interactions with them are very limited and unstable (they tend to transform into someone else)

Places:
I usually have an idea of where I am in dreams. However, my idea is rarely accurate to the place. Often my dream environment is completely different to how it is IRL, but I assume that the places are the same.
I often have dreams where I am on days out in places which I do not recognise at all and do not assume to recognise.
Less often, I have dreams where I am in places which are very unusual looking and almost fantastical. If I am in such a place, the plot tends to be more farfetched.
Depending on how often I play an RPG, I can have dreams where I am half playing, half inside a game.

Plots:
Often relating to a recent or current concern or happening in my waking life.
I expected schemata to have a great deal of influence on the plot, but looking through my journal I can't see many obvious connections. The connections could just be a lot more random or subtle than I would expect.

Part Two:
I have come across the schema theory of learning in a psychology book I read a while ago. I can't remember it too accurately but I think it said children learn some small simple set of rules, and then they apply them to objects or situations of similarity. I think the example given was using a CD player. If we learn to use one CD player, we are better able to learn to use others because they follow similar rules. We apply our schema for the first CD player to the next one, and even if the buttons are located in different places, and the devices look completely different, we can learn to use it faster than someone who has not used any CD players before.
A schema is a set of rules created in association with a stimulus.
So, when applied to dreams, we can use schemata to create dream worlds which function differently and have their own rules, and because this is a process we do naturally we don't have to consciously deliberate over the minute details of our world (which would compromise its stability). It sounds like a quite convincing hypothesis despite my lack of evidence in my own dreams to validate it. The example about watching fantasy films was interesting. we consciously construct a schema, but we subconsciously use it once it has been made.

I think these will definitely be useful tools once I'm lucid, but inducing lucidity isn't easy for me.
This actually reminds me of Wittgenstein's 'language games'. Some language is meaningless for those who don't play the game. What does 'check mate' mean to someone who has never played chess? But it makes perfect sense to a chess player. Learning a schema could just be like learning a new game.