Skill Level Advanced
- [Conrad] Hi, I'm Conrad Carlberg, and welcome to my course on confidence intervals. As you may know, confidence intervals allow us to summarize data in a way that pinpoints an outcome, and yet puts that precise finding in a context that highlights whole regions of possible values. The great statistical theorist John Tookey wrote that confidence intervals make clear the essential smudginess of experimental knowledge. That perspective seems especially important when the replication of experimental results is subject to question. And the scientific method tells us that's how it should be. Confidence intervals have a broad applicability, both across disciplines, and across numeric techniques. You'll find them used in metallurgical research, medical findings, and political reports. You'll find them used in the analysis of mean differences, regression coefficients, and polynomial distributions. And yet they make use of analytic techniques that are typically covered in basic courses on statistics and probability, like this course. I cover the fundamentals of confidence intervals here, and I focus on their use in the comparison of group means with hypothetical values, as well as the comparison of means with one another. Although I emphasize symmetric distributions, the concepts generalize smoothly to more complex situations such as binomial outcomes and other asymmetric distributions. Shall we get started?