Causality X

POL SCI M170X - Causality X

<p>(Same as Political Science M170X.) Lecture, three hours; discussion, one hour. Enforced requisites: course 10 or 12 or 13 or Political Science 6, and course 20. Strongly recommended: course 100A or Mathematics 170E. Causal claims—claims that one thing affects another—are central to life choices, science, and policy. They are very difficult to substantiate and often misleading. The data revolution makes various forms of association easier to discover, but in many cases does not help to affirm whether an observed association reflects the implied causal relationship. The larger datasets grow, the more easily non-causal associations can rise to the level of being noticed and mistakenly used as causal. Study builds conceptual and formal analytical skills required to evaluate causal claims through a focus on conceptual work and more technical analytic methods. Reading and critiquing studies and reporting of studies at a conceptual level. Use of online tools to review statistics and R-coding. Improvement of conceptual understandings by examining the statistical tools used to understand, work with, and make causal claims in a variety of circumstances. P/NP or letter grading.</p>

Statistics

College of Letters and Science

Container 1.2

Container 1

Statistics and Data Science

STATS M169 - Causality X

You will learn statistical frameworks and “languages” -- namely potential outcomes and graphical causal models -- that we can leverage to better understand and work with causality in data science and statistics.

You will learn about both older and newer research designs, including randomized trials, instrumental variables, regression discontinuity, and more that are used in scientific disciplines to grapple with causal challenges.

You will develop skills for reading and critiquing studies and reporting of studies at a conceptual level through class discussions and in writing.

You will apply statistical concepts, including probability theory, conditional probability, confidence intervals and regression in methodology for causal analysis. You will also employ coding skills in R to execute these analyses.

You will build capacity for skeptical and skilled analysis of causal claims as they appear informally, in educational materials, in the news, and in scientific writing. 

General Catalog

Causality X

STATS M169

4 units