Suppose we have a set of data and we would have frequency histogram. If the histogram resembles the one below, we can approximate it using the normal curve. We can see that if we place the normal curve (bell curve) over the top of the histogram, it seems to follow the same shape. With actual data, it will never be perfectly normally distributed, but if it is approximately normal (normally distributed), then we can use the standard normal distribution (z-scores) to help us answer questions or perform probability calculations about the data.