With the multitude of ways for teachers to gather and analyze information about their classes, and the wealth of information available to instructors via the internet, becoming a data-driven instructor can be an overwhelming thought. There are many questions to consider when you are first getting started - “What does data-driven instruction mean?” “How should I collect data?” “What information is most important to me?”

This post is the first in our series on data-driven education, and through the posts to follow we hope to shed some light on this important topic and cover the key aspects relevant to those teaching Introductory Physics. In this post we will focus specifically on the type of data you should be collecting in your class. First, let’s begin with a general definition of data-driven instruction, and how this philosophy and practice plays out in the classroom.

Essentially, instruction that is data-focused builds off of a continuous cycle of assessment, analysis and action. The systematic analysis of student assessments drives new and thoughtful action on the part of the instructor. When planning a data cycle, the instructor should ask fundamental questions regarding each stage:

• “What should my students know?”
• “How will I measure their knowledge or show mastery on a topic?”
• “What pedagogical practices will I implement?”

Each of these questions, and more, will be addressed throughout the series, but let’s dive into the most fundamental - “What type of data should I collect?”

Concordia University in Portland provides a good starter list that we have expanded upon below. This will be different for each class, and it is important to tailor these sources of data for your own purposes.

• Short Quizzes, Q&A sessions, In-Class Activities

• These formative assessments can help guide the direction of a class in real time by slowing down and addressing misconceptions, or moving forward with material as necessary.

• Examinations

• Summative assessments are also a key component in determining the comprehension of key objectives, and (as we’ve identified previously) more frequent examinations may also reduce academic dishonesty.

• Individual or Group Projects, Lab Reports

• Long-term assignments taking several weeks to complete may provide insight into several key knowledge areas, and also offer data on additional competencies (time management, public speaking, etc. that you may want to analyze).

• Student-Reported Data
• Student reflections on progress toward achieving course milestones can be valuable, especially when compared to actual performance on key objectives.

• Homework

• Allowing students to practice, and reinforce concepts on their own, or in group study sessions, is another important way to track milestones, incentivize periodic self-assessments, and provide insights for concepts that need more attention.

Knowing what data to collect is an important jumping-off point when transitioning to a data-driven approach. Once you have determined the sources of data collection, the next step is to develop a plan for collecting, storing, and analyzing the data to ensure you get real-time feedback and useful insights about the progress of your students. Then in the final stage of implementing a data-driven approach, you will incorporate this feedback and integrate it into the course curriculum throughout the semester, and from year to year.

As data-driven education continues to evolve, leading the way is the conversation surrounding Big Data analytics. Read more about how to put these techniques to use in your classes.