What This Document Is
These are lecture notes from an Intermediate Statistics course (36 705) at Carnegie Mellon University, specifically covering the topic of hypothesis testing. The notes introduce the fundamental concepts and framework for formally evaluating evidence against a pre-defined claim about a population parameter. It lays the groundwork for understanding how to determine if observed data supports or contradicts a specific hypothesis.
Why This Document Matters
These notes are essential for students enrolled in an intermediate-level statistics course. Hypothesis testing is a cornerstone of statistical inference, used across numerous disciplines – from scientific research and engineering to economics and business analytics. Understanding these concepts is crucial for interpreting research findings, making data-driven decisions, and drawing valid conclusions from data. This material is typically covered early in a hypothesis testing unit, setting the stage for more advanced testing methods.
Common Limitations or Challenges
This document provides the *theory* behind hypothesis testing. It does *not* offer detailed calculations, specific software implementations, or real-world case studies. It also cautions against over-reliance on hypothesis testing, suggesting confidence intervals are often a more appropriate analytical tool. The notes are a starting point; applying these concepts requires practice and further exploration of different test types.
What This Document Provides
This lecture includes:
* A formal definition of null and alternative hypotheses.
* An explanation of Type I and Type II errors and their implications.
* An introduction to the concepts of test statistics and rejection regions.
* A discussion of error rates (alpha level, power) and how to evaluate tests.
* An overview of several common testing methods: Neyman-Pearson, Wald, Likelihood Ratio, and permutation tests.
* Illustrative examples of one-sided and two-sided hypothesis tests with normal distributions.
This preview *does not* include detailed derivations of formulas, complete worked examples, or in-depth coverage of each testing method. It focuses on establishing the foundational principles of hypothesis testing.
