Interval Estimation Notes
Module 06: Interval Estimation
Overview of Interval Estimation
- Interval Estimation for Population Mean:
- When population standard deviation (B3) is known
- When population standard deviation (B3) is unknown
- Determining sample size for a desired precision
Point Estimators and Sampling Distribution
- Point Estimator: A sample statistic that estimates a population parameter.
- Example: Sample mean (x̄) estimates population mean (μ).
- Sample Mean (x̄): A random variable with its own probability distribution, known as the sampling distribution of the sample mean.
- Mean of sampling distribution = μ, standard deviation (standard error) = σ/√n
- Shape: Normal distribution if n ≥ 30 (Central Limit Theorem - CLT)
- Unbiased Estimator: x̄ is unbiased for μ if E(x̄) = μ.
Interval Estimates
- An interval estimate provides information about how close the point estimate is to the actual population parameter.
- General Form: Point Estimate ± Margin of Error
- Calculate margin of error using either known or sample standard deviation.
Case 1: Population Standard Deviation Known
- Formula: μ = x̄ ± z(σ/√n)
- z = z-value for the confidence level
- Example: Kohl's customer spending scenario
- n = 100, σ = $20, x̄ = $82
- Margin of Error: ± 1.96(20/√100) = ±3.92
- 95% Confidence Interval: [78.08, 85.92]
Case 2: Population Standard Deviation Unknown
- Use sample standard deviation (s) to estimate population standard deviation.
- Formula: μ = x̄ ± t(s/√n)
- t-value aligns with the confidence level and degrees of freedom (n-1)
- Larger sample size yields a closer match to the normal distribution.
Confidence Intervals
- A 95% CI means we are 95% confident the interval includes the true population mean.
- For a 90% CI, use z = 1.645; for a 99% CI, use z = 2.576.
- Adjustments in confidence levels affect the margin of error: larger confidence levels increase margin of error.
- Example from Kohl's:
- 90% CI: ±3.29, resulting in [78.71, 85.29]
- 99% CI: ±5.15, resulting in [76.85, 87.15]
Sample Size Determination
- Formula when B3 is known: n = (z * σ / E)^2
- If σ is unknown: Use preliminary sample or best estimate.
- Example: For a desired margin of error of $5 at 95% confidence for Kohl's:
- n = (1.96 * 20 / 5)^2 ≈ 62
- Example for Credit Card Debts study:
- Given a margin of error of $500:
- n = (1.96 * 4007 / 500)^2 ≈ 247
- Known Population Standard Deviation: μ = x̄ ± z(σ/√n)
- Unknown Population Standard Deviation: μ = x̄ ± t(s/√n)
- Sample Size for Desired Margin of Error: n = (z * σ / E)²