Summary of Introduction to Statistical Concepts and Methods
Introduction to Statistical Concepts and Methods Guide
Introduction
This study guide covers sampling distributions and basic statistical inference for means and proportions. You will learn how sample statistics relate to population parameters, how to build confidence intervals, and how to choose sample sizes to meet precision goals. The material uses intuitive explanations, worked examples and short exercises suitable for a student studying independently.
Definition: A sampling distribution is the probability distribution of a statistic (for example, a sample mean or proportion) computed from all possible samples of a given size drawn from a population.
1. Sampling distribution of the sample mean
1.1. Key properties
- The mean of all possible sample means equals the population mean: $\overline{X}_{\text{all}} = \mu$.
- The variability of the sample mean is measured by the standard error (SE). For a known population standard deviation $\sigma$:
$$\text{SE} = \frac{\sigma}{\sqrt{n}}$$
- As the sample size $n$ increases, the standard error decreases, so the sample mean becomes a more precise estimator of $\mu$.
Definition: Standard error is the standard deviation of a sampling distribution; it quantifies how much sample statistics vary from sample to sample.
1.2. When $\sigma$ is unknown
- Replace $\sigma$ with the sample standard deviation $s$ and use
$$\text{SE} = \frac{s}{\sqrt{n}}$$
- If $\sigma$ is unknown and $n<30$ while the population is normal, use the Student's t-distribution with degrees of freedom $\text{df}=n-1$.
- The t-distribution is more spread out than the normal distribution; as $n$ grows large it approaches the standard normal (Z) distribution.
2. Central Limit Theorem (CLT)
- If the population distribution is not normal, the distribution of sample means tends toward normal as $n$ increases.
- A common rule of thumb: $n\ge 30$ is usually large enough for the CLT to apply, so the sampling distribution of the mean is approximately normal.
3. Sampling distribution of a proportion
- Use the sample proportion $\hat{p}$ to estimate the population proportion $\pi$.
- Mean of sampling distribution: $\mathbb{E}(\hat{p}) = \pi$.
- Standard error when $\pi$ is known:
$$\text{SE}_{\hat{p}} = \sqrt{\frac{\pi(1-\pi)}{n}}$$
- When $\pi$ is unknown replace $\pi$ with $\hat{p}$:
$$\text{SE}_{\hat{p}} = \sqrt{\frac{\hat{p}(1-\hat{p})}{n}}$$
- Normal approximation (Z) is appropriate when $n\pi\ge 5$ and $n(1-\pi)\ge 5$.
4. Point estimation vs Interval estimation
4.1. Point estimation
- A point estimate is a single-value estimate of a population parameter: sample mean $\overline{x}$ for $\mu$, sample proportion $\hat{p}$ for $\pi$.
- Point estimates have sampling error; they are rarely exactly equal to the population parameter.
4.2. Confidence intervals (CIs)
- A confidence interval gives a range of plausible values for the population parameter with a stated confidence level $1-\alpha$ (e.g., 95%).
- General form for a CI for a mean:
$$\text{point estimate} \pm \text{critical value} \times \text{SE}$$
- Margin of error $\text{ME} = \text{critical value} \times \text{SE}$.
Definition: Confidence level $1-\alpha$ is the long-run proportion of similar constructed intervals that will contain the true parameter.
5. Which distribution / formula to use
Use the table below to select the correct interval formula.
| Scenario | Condition | Interval formula |
|---|---|---|
| Single mean, $\sigma$ known | Population normal or large $n$ | $\overline{x} \pm z_{\alpha/2},\frac{\sigma}{\sqrt{n}}$ |
| Single mean, $\sigma$ unknown, large $n$ | $n\ge 30$ (CLT) | $\overline |
Sign up for the full summaryAlready have an account? Sign in
Sampling Distribution & Inference
Klíčová slova: Statistical software, Descriptive statistics overview, Levels of Measurement, Central tendency & dispersion, Graphs & frequency distributions, Sampling methods & surveys, Sampling distribution & inference, Statistical distributions, Regression & correlation, Probability Foundations, Probability & Counting Methods, Probability Distributions, Hypothesis testing methods, Time series
Klíčové pojmy: Mean of sample means equals population mean $\mu$, Standard error of mean $=\sigma/\sqrt{n}$ when $\sigma$ known, Replace $\sigma$ with $s$ and use t-distribution when $\sigma$ unknown and $n<30$, CLT: for non-normal populations, sample means approx normal for large $n$ (rule: $n\ge 30$), SE for proportion $=\sqrt{\hat{p}(1-\hat{p})/n}$, Confidence interval: point estimate $\pm$ critical value $\times$ SE, Sample size for mean: $n=\left(\dfrac{z_{\alpha/2}\,\sigma}{E}\right)^2$, Sample size for proportion: $n=\dfrac{z_{\alpha/2}^2 p^*(1-p^*)}{E^2}$, Use Z when $\sigma$ known or $n$ large via CLT, use t when $\sigma$ unknown and small $n$, Normal approximation for proportion requires $n\pi\ge 5$ and $n(1-\pi)\ge 5$, Margin of error depends on confidence level and SE, Round sample size up to ensure required precision