91Ó°ÊÓ

The sample proportion is a crucial concept in statistics, especially when making inferences about a population based on a sample. Imagine you want to know how many American adults think regular exercise is important. You can't ask everyone, so you survey 1,000 adults instead. In this example, out of the 1,000 adults, 753 said they believe exercise is important.
This percentage, or proportion, for the people you've surveyed is known as the sample proportion. You calculate it using the formula \( p = \frac{x}{n} \), where \( x \) is the number of "successes" (here, people who said exercise is important), and \( n \) is the total number of observations (or people surveyed). In this case, \( p = \frac{753}{1000} = 0.753 \).

• This tells you that 75.3% of the people in your sample believe exercise is important.
• It serves as an estimate or reflection of the population proportion.
• Understanding the sample proportion helps us make informed guesses about the larger group from whom the sample was drawn.

When making estimates using samples, it is critical to understand that these are, indeed, estimates, and not definitive conclusions. This is where the margin of error plays an important role. It gives us a range within which we think the true population parameter lies.
The margin of error is calculated using the formula \( E = z \cdot \sqrt{\left(\frac{p \cdot (1-p)}{n}\right)} \). Let's break this down:

• \( z \) refers to the z-score, which you select based on the desired confidence level (in our example, a 90% confidence level gives us a z-score of 1.645).
• \( p \) is the sample proportion (0.753 in our case).
• And \( n \) is the sample size (1,000 people surveyed).

Inserting these values into the formula gives us a margin of error of \( 0.027 \). What this means practically is that we are 90% confident that our estimate from the sample could potentially vary by this amount when estimating the true population proportion.

The z-score is a measure that reflects how many standard deviations a data point is from the mean. In the context of confidence intervals, the z-score helps us determine how extreme a sample proportion is relative to the population proportion.
For confidence intervals, the z-score quantifies the level of confidence we want to have in our interval estimate. For example, if you want to be 90% confident that your sample proportion falls within your calculated range, you'd use a z-score of 1.645.

• This z-score is derived from standard normal distribution tables or calculators.
• It influences the width of the confidence interval: higher z-scores (like 1.96 for 95% confidence, or 2.576 for 99%) lead to wider intervals, implying greater certainty that the interval contains the true proportion.
• So in our example, a z-score of 1.645 supports a 90% confidence that the interval (0.726, 0.78) contains the real population proportion.

Understanding how to use and choose the correct z-score is key to accurately estimating confidence intervals and making reliable statistical inferences.