Big data and learning analytics

- There's one question I get asked very often. Exactly how big is big data? Think of it this way. Big data is a dataset so large that you can't just open it in an everyday spreadsheet program like Excel. Big data requires specialized analytics tools to process and make complex calculations. Now the definition of big data can vary from industry to industry. But we generally agree that it includes three dimensions. Volume, variety, and velocity. They're the three V's of big data. Volume generally refers to the amount of data you get, variety refers to the number of types of different data. And finally, velocity refers to the speed of data processing. How fast can you process the data. Recent technological breakthroughs have significantly reduced the cost of data storage and computation making it easy and cheaper to store data. And this is why you're probably seeing the term big data everywhere. Plus, technologies generate a log of data, or data exhaust as we call it. For workplace learning, many companies use learning management systems or sometimes it's called LMS to host and track learning. LMS records all kinds of data about the learners and about how they used the system. Data such as these things, the learner's profile, demographic information on age, gender, educational level, et cetera. And finally you get clickstream data. Datas like where did the learners go, if they clicked on the links, and how long did they spend on a page once they clicked on it. This data provides us with insights into learner behaviors and the learning experiences. For example, we can use data to help answer questions like how many people logged into the learning platform today. Which learning resources are most popular, how many attempts does it take on average for people to pass a particular quiz. Sometimes these datasets are small enough that we can process them using Excel. But there are other times when the datasets are so huge that we need to use specialized tools for making complex calculations and for making predictions. It is especially useful when we want to make sense of data from other platforms such as internal social networking sites, e-libraries, and online discussion forums. Of course it's great to know what your learners are doing online, and to have insights about them. But that is not all. Learning analytics need to go one step further. We need to do something with the insights we gain. We need to act on it. For example, managers can look at how engaged the staff are for each piece of content through a learning analytics dashboard and predict how likely they are to pass or fail the course. Managers can then provide support to staff that might fail, getting to know the reasons behind such low engagement, and to provide appropriate interventions. While big data and learning analytics have come a long way the combined usefulness is just at the beginning. Do you know what type of big data your organizations collects to improve learning? I encourage you to start exploring that today.