Write Down Everything
Whether you decide on a project with a formal hypothesis or not, when you perform your project (take data), there are steps you can take to make the most of your project. First, write everything down. Gather your materials and list them, as specifically as you can. In the scientific world, it is important to be able to duplicate an experiment, especially if surprising results are obtained. In addition to writing down data, you should note any factors that could affect your project. In my salt example, it is possible that the temperature could affect my results (alter solubility of salt, change the body's rate of excretion, and other factors I might not consciously consider). Other factors I might note could include relative humidity, age of participants in my study, a list of medications (if anyone is taking them), etc. Basically, write down anything of note or potential interest. This information could lead your study in new directions once you start taking data. The information you take down at this point could make for a fascinating summary or discussion of future research directions for your paper or presentation.
Don't Discard Data
Perform your project and record your data. When you form a hypothesis or seek the answer to a question, you probably have a preconceived idea of the answer. Don't let this preconception influence the data you record! If you see a data point that looks 'off', don't throw it out, no matter how strong the temptation. If you are aware of some unusual event that occurred when the data was being taken, feel free to make a note of it, but don't discard the data.
Repeat the Experiment
If I want to determine the level at which I taste salt in water, I can keep adding salt to water until I have a detectable level, record the value, and move on. However, that single data point will have very little scientific significance. It is necessary to repeat the experiment, perhaps several times, to get a significant value. Keep notes on the conditions surrounding a duplication of an experiment. If I duplicate my salt experiment, perhaps I would get different results if I kept tasting salt solutions over and over than if I performed the test once a day over a span of several days. If your data takes the form of a survey, multiple data points might consist of many responses to the survey. If the same survey is resubmitted to the same group of people in a short time span, would their answers change? Would it matter if the same survey was given to a different, yet seemingly similar group of people? Think about questions like this and take care in repeating a project.


