Unit 5: Sampling Distributions

📚Study Guide: Sampling Distributions

Unit 5: Sampling Distributions

Sampling distributions bridge descriptive statistics and inferential statistics by describing how sample statistics vary across repeated samples. This unit introduces the sampling distribution of the sample mean x_bar and the sample proportion p_hat. You will learn the conditions under which these sampling distributions are approximately normal (Central Limit Theorem) and how to compute their means and standard deviations. Understanding sampling distributions is essential because every confidence interval and hypothesis test relies on knowing the behavior of the statistic under repeated sampling. The 10% condition and Large Counts condition are critical validity checks.

Key Concepts

• Sampling Distribution: The distribution of a statistic (like x_bar or p_hat) across all possible samples of size n from a population.
• Central Limit Theorem (CLT): For large n, the sampling distribution of x_bar is approximately normal regardless of the population shape. Common threshold is n >= 30.
• Mean of Sampling Distribution: mu_x_bar = mu; mu_p_hat = p. Statistics are unbiased if their expected value equals the parameter.
• Standard Deviation of x_bar: sigma_x_bar = sigma / sqrt(n). Requires the 10% condition (n <= 0.10N) if sampling without replacement.
• Standard Deviation of p_hat: sigma_p_hat = sqrt[p(1-p)/n]. Requires Large Counts: np >= 10 and n(1-p) >= 10.
• Unbiased Estimator: The mean of the sampling distribution equals the population parameter.
• Variability: Larger sample sizes decrease the standard deviation of the sampling distribution, making estimates more precise.

Vocabulary

• Parameter: A number describing a population (e.g., mu, p).
• Statistic: A number describing a sample (e.g., x_bar, p_hat).
• Sampling variability: The natural variation in statistics from sample to sample.
• Standard error: The estimated standard deviation of a sampling distribution when the population value is unknown.
• 10% condition: Sample size should be no more than 10% of the population when sampling without replacement.
• Large Counts condition: Requires at least 10 expected successes and 10 expected failures for normal approximation of p_hat.

Formulas

• mu_x_bar = mu
• sigma_x_bar = sigma / sqrt(n)
• mu_p_hat = p
• sigma_p_hat = sqrt[ p(1-p) / n ]
• z = (x_bar - mu) / (sigma/sqrt(n))
• z = (p_hat - p) / sqrt[ p(1-p)/n ]

Common Mistakes

• Applying the CLT to small samples from non-normal populations without justification.
• Using the population standard deviation formula sigma/sqrt(n) without checking the 10% condition for finite populations.
• Confusing the sample size n with the number of samples; sampling distributions are built from many samples, each of size n.
• Forgetting that p_hat is approximately normal only when the Large Counts condition is met.

AP Exam Strategies

• Always name the sampling distribution and its parameters (mean and standard deviation) before computing probabilities.
• Explicitly verify conditions: random sampling, normal population or large n for means; random sampling and Large Counts for proportions.
• Use the 10% condition to justify using sigma/sqrt(n) when sampling without replacement from finite populations.
• Remember that increasing n decreases standard deviation but does not eliminate bias.

Real-World Applications

• Polling: Sampling distributions quantify the margin of error in political and market surveys.
• Manufacturing: Quality control uses sampling distributions to set acceptance criteria for production batches.
• Medicine: Clinical trial designs rely on sampling distributions to determine necessary sample sizes.