Questions
Write a hypothesis in the form of a question. For example, you could ask whether soil salinity affects plant growth. Questions aren't the most effective way to frame hypotheses, however, because they are open-ended. They don't make predictions about what you expect to observe. Nonetheless, questions can serve as a starting point for writing a hypothesis. You can use your observations to turn a question into a conditional statement or an if-then scenario.
Conditional Statements
A conditional statement is another way to state a hypothesis. For example: soil salinity may be related to plant growth. This kind of statement should always be conditional. If you say that increasing soil salinity WILL impact plant growth, you're not stating a hypothesis, you're making a statement of fact based on an assumption about the way nature works. You can never assume your hypothesis is true without testing it first.
Problems with Conditional Statements
A conditional statement is more useful than a question, but it's still not a useful hypothesis because it doesn't tell you what you'd expect to see in an experiment if the hypothesis is true. It may be, for example, that soil salinity affects plant growth, however, your conditional statement doesn't tell you how you might go about testing this idea. Word your hypothesis in such a way that it remains tentative while making a prediction you can evaluate based on experimental results. The best way to state a hypothesis is by formalizing it as an if-then statement.
If-Then Statements
An if-then statement might read: if soil salinity affects plant growth, then increased soil salinity will cause decreased growth. This statement has two clauses. The first clause is similar to your conditional statement, it suggests a link between two phenomena. The second clause tells you what you would expect to see if the first clause is true. Stating your hypothesis this way helps you design your experiment. You know what you need to test to evaluate your hypothesis. If-then statements are the best and most formal way to state a scientific hypothesis.