What is Your Data Strategy?
Are you collecting intelligent data?
Guest contributor, Sean Putman, shares his insights and expertise in developing a data strategy to get the most value from your learning analytics. Sean has been an instructor, instructional designer, and developer for over 15 years. He has spent his career designing and developing training programs, both instructor-led and online, for many different industries but he has had a strong focus on creating material for software companies. Sean has also spent the last few years focusing on the use and deployment of the Experience API (xAPI) and its effect on learning interventions. Sean has spoken at many industry conferences on the subject and is co-author of “Investigating Performance” a book on using the Experience API and analytics to improve performance. You may find Sean’s book on Amazon and connect with him via Twitter.
I have received funny looks when I tell people new to learning analytics that collecting data is the easy part. Collecting intelligent data is where it gets tough. That means collecting data that provides value. No matter how you collect the data – xAPI, cmi5, Google Analytics, others – the data can come in fast and furious. It can overwhelm you quickly at times. Most of the specifications are easy to implement – there is not much coding involved. Where your time should be spent is planning what you want to collect and how that data will inform whether or not behavior change or performance improvement has occurred.
What Do You Want to Measure?
When people ask me how to get started with a data project or how to use xAPI, the first thing I tell them is to sit down and plan what you want to measure and why. Ask the question, what does behavior change or performance improvement really mean? Why is collecting this data important? What is this data going to show? If you can answer these questions you are off to a good start on your project. You are developing a data strategy.
What Is Data Strategy?
Data strategy is figuring out what type of information you need to collect (data) to answer your ultimate questions, was behavior changed or performance improved? Data strategy is not restricted to data from the L&D department. It might mean that you need to collect data from other departments to complete your analysis.
Let’s look at an example of what this means. I am creating a course to inform users about security measures that need to be in place when passing documents to others. Violations are tracked in a database to keep records of when the security procedures are not followed. The Learning & Development department produces content to train users on the security procedure. To track if behaviors really changed, the L&D department needs to see if violations are reduced after people take the course. The data that is external to L&D is critical to finding this out. Part of the data strategy for this project would be to collect data on the content being accessed and if necessary what items within the content are viewed.
Through collection of this type of data we can see what items are working, what items people are skipping over, and where security violations are still occurring. This allows the L&D department to make design decisions, improve course content, and identify the material that really works. Using a combination of data collected from the production system and correlating it to the learning data provides a better indicator of project success.
How Do I Define a Data Strategy?
Sit down with all the stakeholders of a project and determine what the desired outcome(s) for a project are. This activity provides a roadmap for the expectations of the project. Once the end results are agreed upon, the data that needs to be collected can be documented. A data collection document helps define the analytics needed at the end of the project. It defines what dashboards are needed and help determine how difficult it will be to gain access to data outside the learning organization. Sometimes data from outside sources can be difficult to access. Mapping this out in the early stages of a project helps to avoid roadblocks. Any data access issues are resolved early or you can develop a plan B before development begins. If you wait until the end, you could end up not having the piece of data that is critical to the project’s success.
Remember that specifications like xAPI and cmi5 are just the mechanisms for collecting data. The critical piece to a successful analytics project starts with planning. Developing a sound data strategy helps define and measure project success. Wading through unnecessary data to find important information is time-consuming and often leads to failure. Spend the time in the early stages of a project to develop a sound strategy to guide the rest of the project.