Experimental Design and Analysis in Oceanography

Catalog description

Credits 3. 3 Lecture Hours. COURSE OVERVIEW. This course will develop your ability to design, execute and critically analyze hypothesis-driven studies in oceanography. You will progress from foundational probability to modern statistical modeling, linking experimental design, data collection logistics and rigorous inference. Topics include how to structure field sampling to support causal inference, which models best capture patterns and uncertainty in geoscience data, and how to partition variance to test hypotheses and interpret multivariate structure. Core concepts include experimental design, linear regression and ANOVA within the general linear model and principal component analysis for dimensionality reduction and interpretation. Expect hands-on coding, data wrangling, data visualization and a capstone analysis project. RELEVANCE AND APPLICATION. Ocean science increasingly depends on defensible experimental design and transparent, reproducible analytics. Skills gained, including statistical reasoning, programming, graphical communication and critical interpretation, translate directly to data-centric roles across the geosciences, environmental consulting, resource management and policy analysis. By practicing end-to-end workflows from hypothesis formulation to evidence-based conclusions, you will be better prepared for thesis projects, interdisciplinary collaboration and informed decision-making in professional and civic contexts. KEY TOPICS AND THEMES. You will practice estimation and inference, then connect these ideas to the scientific method, hypothesis framing and parametric analyses used in the geosciences. Design principles for field studies, including replication, independence, randomization and sampling design, provide the foundation for modeling with the general linear framework, progressing from simple to multiple linear regression, diagnostics and the ANOVA table. You will extend these skills to analysis of variance with fixed or random factors, and interaction terms. The multivariate portion introduces principal component analysis for dimensionality reduction, interpretation and ordination. You will also strengthen computational fluency by writing programs in Python, R or MATLAB to analyze data, generate clear figures and tables and communicate results effectively. These skills culminate in a capstone project in which you select a dataset, articulate and test hypotheses with reproducible code and present your findings to the class in written and oral formats. Prerequisite: Graduate classification or approval of instructor.