Think of hypothesis writing as a three-step process. This will help you to organize the whole process in your head and it will function as an effective guide.
Come up with a general hypothesis, which is basically stating the relationship between the two variables in your experiment--a rough observation based on everything that has been seen and collected in the information gathering stage. If you're conducting an experiment revolving around the effect of salt rubbed into an open wound, a good general hypothesis might be "rubbing salt into an open wound may affect the level of pain."
Form a specific, or directional, hypothesis. A general hypothesis provides the experiment with no direction or indication of where the experiment should be taken. These important details regarding the experiment should be answered by the specific hypothesis. For example, "rubbing salt into an open wound may increase the level of pain." The affect that one variable has on the other is made clear.
Come up with a good measurable hypothesis. Further elaborate on the relationship between two variables by eliminating other factors to make the experiment testable. This might be "rubbing salt into an open wound may increase the level of pain because salt contains sodium chloride." This third step enables you to design your research around your hypothesis and to conduct your experiment.