Intro Stats: Policy Practioner

3 Credits · Level: Graduate, Undergraduate

This course provides a basic introduction to probability and statistics. Policy analysts and practitioners need not be statisticians, but they must be able to interpret data, calculate descriptive statistics, and understand the results of statistical models. Our classes will cover the mathematical building blocks but also focus heavily on research design, when to apply statistical techniques, how to translate regression outputs, and the assumptions and limitations of quantitative analysis. This will include assessing the generalizability and validity of arguments presented in articles that rely on statistical analysis. In the first few weeks of the course, we will use tools, including Excel and Stata, to explore data – measures of central tendency, measures of dispersion, and data distributions. Then, we will discuss probability – a fundamental tool for statistical sampling and hypothesis testing. When we cover sampling and hypothesis testing, we will focus on practical applications: when practitioners should use these tools and how to establish a sampling frame, set up fair tests, and determine sample size, for example. Of course, students will also learn the mathematical and computing techniques necessary to design and pull basic samples, produce estimates based on their results, and conduct hypothesis tests. During the final three weeks, students will learn when linear regression is appropriate, the assumptions researchers make when they use linear regression, how to interpret regression outputs and statistical tables, how to diagnose problems within linear regressions, the importance of replicability in research, the debate over p-values, and the basics of experimental design. This section will not require students to calculate regression by hand; instead, more than any other aspect of the course, it will focus on theory. The goal is for students to leave the class with the ability to critically read and understand articles that use quantitative methods.