How do I write a good hypothesis?

Assuming that you already have a topic for the research (e.g a description of an experiement or study), the first thing to think about is what that research is trying to find out. In other words, what is the question that you are trying to answer? Some examples of research questions would be: Does caffeine consumption affect concentration? Is there a link between personality and phobias? Do memory skills differ between genders? Once you know what the question is, you can formulate a hypothesis from your predicted answer to the question. For example, if your knowledge suggests that consuming lots of caffeine would impair concentration ability, your hypothesis might be "People who drink less caffeine will have better concentration". To improve the hypothesis further, make sure your hypothesis matches the description of the experiment. For example, if concentration is measured by number of mistakes on an attention task, a more specific hypothesis would be "People who drink less caffeine will make more mistakes on the attention task". Students also often ask about one-tailed or two-tailed hypotheses. Think of 'tails' as 'directions' for the purpose of hypothesis-writing ('tailed' is actually a term relevant to statistical testing, but that is not needed just now). A one-tailed hypothesis, like the example above, predicts one specific direction. A two-tailed hypothesis could go either way, for example: "The amount of caffeine a person drinks will affect their concentration" - this does not specify whether you think caffeine will make the concentration levels increase or decrease - it could go two ways, and is therefore a two-tailed hypothesis.