STAT 799 - Topics in Statistics - Applied linear mixed models in agriculture and life sciences

Spring 2026
Kansas State University

Instructor: Dr. Josefina Lacasa

Time: January 26 – February 6: MTWUF 2.30 – 3.45 pm; February 9 – May 15 by appointment.

Course Description: Hierarchical models (i.e., mixed models) are widely used for analyzing data generated in agriculture, as most data has some type of hierarchical structure. This course aims to help students gain understanding and develop the intuition for the most common assumptions in mixed models. The course content is most appropriate for students who already understand the fundamental principles of applied linear models and designed experiments, and want to learn how to analyze hierarchical data using mixed models. This is a project-based course with 2 intensive weeks with classes every day, followed by a 3 month-long semester project with periodic meetings. Students are expected to start the course with a modeling project in mind. Lecture topics to be covered include but are not limited to hierarchical data structures, linear mixed models, non-linear mixed models, heteroscedastic models, and generalized linear mixed models. Potentially, topics like hierarchical Bayesian models and generalized additive models may be covered. Applications of mixed models will be focused on modeling agricultural data, including but not limited to animal science, grain science, and plant science.

Main Goal of this Course:
The main objective of this course is to learn how to design and implement hierarchical models, interpreting the results of said models and making inference. By the end of this course, students should be able to: • Identify the data structure for a given dataset and write the statistical model that corresponds to said data structure using statistical notation. • Distinguish the benefits and disadvantages of different modeling approaches. • Write the Materials and Methods section in a paper (or graduate thesis) that describes the data generating process and the statistical model.

What this course is not: an R programming course (however, help will be provided).

Prerequisites: Instructor consent. Students are expected to have statistical background equivalent to STAT 705 and STAT 720.