Observation simply means noticing things about the world around us. Observation begins with the five senses: sight, hearing, taste, touch and smell. Because the five senses can only tell them so much, scientists use many tools and techniques to help their senses reach farther. Scientists use microscopes to see things that are too small for their eyes to see and telescopes to see things that are too far away. X-rays and MRIs allow them to see inside the human body. Submarines and robots enable them to see deep under the ocean and cameras and recording devices help them observe things in many places at once.
A hypothesis is a special kind of guess--one you make after doing lots of observation, and one you can test. Scientists observe as much as they can, and then come up with a hypothesis (i.e., a guess) about why the things they have observed happen the way they do. For example, you may observe your teacher wearing red boots five times. All five times, it is a Tuesday. Each time, it is snowing. Each time she wears a different coat. Therefore, your hypothesis might be that your teacher wears red boots not because of what day of the week it is, and not to match her outfit, but to keep her feet warm in bad weather.
Once you formulate your hypothesis, you make a prediction based on it. If your prediction comes true, your hypothesis is likely to be true. If your prediction does not come true, your hypothesis cannot be right. In the above example, if your teacher shows up even once wearing red boots on a sunny day, your hypothesis is mistaken.
Scientists design experiments to test hypotheses. For example, a scientist thinks mice that eat candy may be less healthy than mice that don't, so he designs an experiment. He puts six mice in one cage and six in another. He feeds the mice in Cage A two candy bars every week and gives the mice in Cage B only healthy mouse food. Every week, he weighs and measures both groups of mice, checks them for diseases and has them run mazes. After several months, he can begin to see a pattern. If the two groups stay equally healthy, his hypothesis is probably wrong. If the mice that eat the candy develop diseases and the others don't, his hypothesis may be right. The more times he repeats the same test and gets the same result, the more confident he feels that his hypothesis is accurate.