Examples of the Null Hypothesis
- Hyperactivity is unrelated to eating sugar.
- All daisies have the same number of petals.
Examples of an If, Then Hypothesis
- If you get at least 6 hours of sleep, you will do better on tests than if you get less sleep.
- If you drop a ball, it will fall toward the ground.
Improving a Hypothesis To Make It TestableWhile there are many ways to state a hypothesis, you may wish to revise your first hypothesis in order to make it easier to design an experiment to test it. For example, let's say you have a bad breakout the morning after eating a lot of greasy food. You may wonder if there is a correlation between eating greasy food and getting pimples. You propose a hypothesis:
Eating greasy food causes pimples.
Next you need to design an experiment to test this hypothesis. Let's say you decide to eat greasy food every day for a week and record the effect on your face. Then, as a control, for the next week you'll avoid greasy food and see what happens. Now, this is not a very good experiment because it does not take into account other factors, such as hormone levels, stress, sun exposure, exercise or any number of other variables which might conceivably affect your skin. The problem is that you cannot assign cause to your effect. If you eat french fries for a week and suffer a breakout, can you definitely say it was the grease in the food that caused it? Maybe it was the salt. Maybe it was the potato. Maybe it was unrelated to diet.You can't prove your hypothesis. It's much easier to disprove a hypothesis. So, let's restate the hypothesis to make it easy to evaluate the data.
Getting pimples is unaffected by eating greasy food.
So, if you eat fatty food every day for a week and suffer breakouts and then don't breakout the week that you avoid greasy food, you can be pretty sure something is up. Can you disprove the hypothesis? Probably not, since it is so hard to assign cause and effect. However, you can make a strong case that there is some relationship between diet and acne.
If your skin stays clear for the entire test, you may decide to accept your hypothesis. Again, you didn't prove or disprove anything, which is fine.