AGB 327: Data Analysis for Agribusiness
I became interested in education in order to make quantitative concepts accessible to students, so this is one of my favorite courses to teach. I believe data analysis extends beyond simply performing statistical analyses and additionally encompasses the ability to understand and describe the components of a dataset, identify interesting questions to ask, transform the data to answer these questions, present the data in an accessible format and tie all the analyses together into a cohesive narrative. It is important for me that students understand the real-world implications of these tests, and for that reason I use real datasets during class, including some from my own research.
As the class takes place in a lab, I try to limit the time spent lecturing and instead focus on problem-solving. In her article on implementing an active problem-centered model, Jones-Wilson (2005) points out that for many students their first exposure to course material comes from the class lecture, while comprehension does not transpire until after the second viewing, which usually occurs when they are solving homework problems without the aid of an instructor. Instead, I expect students to perform the reading outside of class so that I can incorporate additional problem-solving into class. Specifically, every topic has a corresponding set of reading questions on Polylearn that students must complete before the class session. Students are given a dataset that resembles the example used in their textbook, but with different numbers, forcing them to actively solve the problems from their readings. The Moodle quiz prompts them for their calculated values, and can be taken as many times as desired, incentivizing both completion of the reading, as well as the process of identifying and fixing mistakes.
This ensures that by the time they enter the classroom they have a context for the more difficult, real-world problems we will be tackling during class. I spend the first part of every new module reiterating what they read, but most of the time is employed working on problems, with the ability to use me as a resource when they are confused. We then solve the problem together as a class to address any remaining concerns. This allows me to both use more complex datasets in my examples and spend increased time discussing the implications of our results.
This ties into my favorite component of the class, the final research project. Each quarter students are given a dataset with a corresponding question they are expected to have answered by the end. For instance, I have combined county-level agricultural and demographic data to ask the question “What impacts the number of organic farms in a county”. This assignment is structured to emphasize writing and narrative building. Specifically, students are expected to model an academic econometrics paper by introducing their topic and the data, incorporating relevant literature, using tables and graphs to augment their written data analysis, and devising a conclusion that ties each step of the analysis together. Students must also justify each analytical decision they make, whether through logic or outside sources, and explain how each one builds on the previous component.
The homework assignments are also structured to model the research project process. For each homework, students use the same farmers’ market dataset from my own research to perform statistical tests, allowing them to see how one dataset can be used to answer increasingly complex questions. The other main assessment tool is three exams. Students struggle most with the ability to judge what statistical test should be used to answer a specific question, as well as how to transform the data to successfully complete the problem. While the first midterm tends to be a rude awakening for students they soon learn the importance of understanding not just how to perform a test but also why they do it. I additionally implement the Muddiest Point exercise before each midterm, which is a metacognitive tool that requires students to both assess what they found confusing as well as compare and contrast their levels of uncertainty. This process allows me to better target review sessions and serves as a springboard for encouraging more introverted students to state their questions and concerns.
Over the summer I took the CTLT Introduction to Online Teaching and Learning workshop. One exercise I found especially useful was to break up the course material into separate modules, each with its own learning objectives and corresponding assignments that aid and assess students’ achievement of those objectives. By structuring the course in this manner I am able to clearly communicate to students what they are expected to learn as well as the underlying purpose of their assignments. The Polylearn course shell is now structured to represent a course roadmap that students will follow throughout the quarter, tying the class together.
Sample Materials:
Syllabus
Example of Reading Question
Homework Assignments
Final Project
I became interested in education in order to make quantitative concepts accessible to students, so this is one of my favorite courses to teach. I believe data analysis extends beyond simply performing statistical analyses and additionally encompasses the ability to understand and describe the components of a dataset, identify interesting questions to ask, transform the data to answer these questions, present the data in an accessible format and tie all the analyses together into a cohesive narrative. It is important for me that students understand the real-world implications of these tests, and for that reason I use real datasets during class, including some from my own research.
As the class takes place in a lab, I try to limit the time spent lecturing and instead focus on problem-solving. In her article on implementing an active problem-centered model, Jones-Wilson (2005) points out that for many students their first exposure to course material comes from the class lecture, while comprehension does not transpire until after the second viewing, which usually occurs when they are solving homework problems without the aid of an instructor. Instead, I expect students to perform the reading outside of class so that I can incorporate additional problem-solving into class. Specifically, every topic has a corresponding set of reading questions on Polylearn that students must complete before the class session. Students are given a dataset that resembles the example used in their textbook, but with different numbers, forcing them to actively solve the problems from their readings. The Moodle quiz prompts them for their calculated values, and can be taken as many times as desired, incentivizing both completion of the reading, as well as the process of identifying and fixing mistakes.
This ensures that by the time they enter the classroom they have a context for the more difficult, real-world problems we will be tackling during class. I spend the first part of every new module reiterating what they read, but most of the time is employed working on problems, with the ability to use me as a resource when they are confused. We then solve the problem together as a class to address any remaining concerns. This allows me to both use more complex datasets in my examples and spend increased time discussing the implications of our results.
This ties into my favorite component of the class, the final research project. Each quarter students are given a dataset with a corresponding question they are expected to have answered by the end. For instance, I have combined county-level agricultural and demographic data to ask the question “What impacts the number of organic farms in a county”. This assignment is structured to emphasize writing and narrative building. Specifically, students are expected to model an academic econometrics paper by introducing their topic and the data, incorporating relevant literature, using tables and graphs to augment their written data analysis, and devising a conclusion that ties each step of the analysis together. Students must also justify each analytical decision they make, whether through logic or outside sources, and explain how each one builds on the previous component.
The homework assignments are also structured to model the research project process. For each homework, students use the same farmers’ market dataset from my own research to perform statistical tests, allowing them to see how one dataset can be used to answer increasingly complex questions. The other main assessment tool is three exams. Students struggle most with the ability to judge what statistical test should be used to answer a specific question, as well as how to transform the data to successfully complete the problem. While the first midterm tends to be a rude awakening for students they soon learn the importance of understanding not just how to perform a test but also why they do it. I additionally implement the Muddiest Point exercise before each midterm, which is a metacognitive tool that requires students to both assess what they found confusing as well as compare and contrast their levels of uncertainty. This process allows me to better target review sessions and serves as a springboard for encouraging more introverted students to state their questions and concerns.
Over the summer I took the CTLT Introduction to Online Teaching and Learning workshop. One exercise I found especially useful was to break up the course material into separate modules, each with its own learning objectives and corresponding assignments that aid and assess students’ achievement of those objectives. By structuring the course in this manner I am able to clearly communicate to students what they are expected to learn as well as the underlying purpose of their assignments. The Polylearn course shell is now structured to represent a course roadmap that students will follow throughout the quarter, tying the class together.
Sample Materials:
Syllabus
Example of Reading Question
Homework Assignments
Final Project