Alright, so my science teacher wants us to find three science project/experiment ideas that have variables, has at least four weeks of testing, and uses the scientific method. I am completely lost, and he is sorta..wacky, and doesn't know how to help me. Every time I ask him something he's like "Oh don't give me that" or "You're a smart kid, think of something". So...do any of you have some ideas? Thanks.

Four weeks of testing? Do some botany project or something about infared rays from the sun that has to take a while.
You can always do the old standby of mold growing on bread, and use variables like amount of light exposure, moisture, type of bread... the list goes on.
Some Ideas: *** from your explanation i wasn't sure whether or not you'll have to carry out the experiments so i'll give some realistic experiments

Rate of Oxidation on Guitar Strings Over a long period of time:
Use a couple different brands, ex d'addario, some polyweb elixers. Find mass before, after with oxides. Do your best to completely remove the rusts and measure again. This way you can find amount of oxide formed, do some stoichiometry and work backwards to find the amount of nickel/ other metals on the strings that reacted with oxygen. Note: don't play the strings because that introduces new factors caused by oils and acids from the sebaceous glands. Also, if they don't rust fast enough for your experiment, through in a catalyst like copper (II) iodide. You can make hypothesis's with this saying for example that you believe the d'addarios will have a higher rate of oxide formation then the polyweb elixers. The elixers may not be a good choice for strings, as i've had a set on my acoustic for almost half a year and there is no build up yet.

Effects of Soft Drinks on Teeth:
This could be a good one. If you can manage to get access to similar teeth somehow. Put the teeth in a control mixture of water. Then in coke, pepsi, sprite, and some control solutions like phosphoric acid, hydrochloric acid and maybe a week acid like acetic. These should all be easy enough to get. Your data could be mass of tooth decay for quantitave data, plus you could have qualitative data from your observations. I think this would be an interesting experiment.

Heres an idea, as a poor university student, I like to take part in weird studies to make some $$$. The weirdest one was getting an MRI while answering weird questions. You could come up with a long term study with questionaires. Example: effects of drug use on academic performance. Just get a pool of like 100 people from your school. Then look into some statistical methods for analyzing your data. This one would be hard to do, but if you just need an idea, you can score some serious browner points by showing some extra research.
Are you serious? It could be anything!

Growth of plants under different conditions.
Road salt causing rusting on a car.

Sit down for 5 minutes and you could easily get three of them.