In statistics notation a "hat" over a letter usually means estimate of a parameter. P means proportion. P-hat (usually written as a p with a little hat shape over it) thus means estimate of proportion. Parameters apply to populations, and statistics apply to samples. Usually, you cannot work with an entire population, so you use a sample to estimate the quantities you are interested in. For example, if you wanted to estimate the proportion of men who voted for Obama, you would ask a sample of men who they voted for.

### Step 1

Sample your population. For example, you might use random digit dialing to reach people, then ask any men who they voted for in the last election.

### Step 2

Count the number of subjects in your sample who have the property you are interested in. In the example, count the number of men who say they voted for Obama.

### Step 3

Divide this by the total size of your sample. This is p-hat.