Instructors, training managers, and executives can make more informed decisions and improve learning outcomes if they have better insight from learner data. However, data analytics software is traditionally not built for non-specialists, therefore creating meaningful interpretations of data often requires specialized data analysts.
A framework to analyze, interpret, and visualize micro-level, behavior-driven learning data by non-technical users. DAVE enables data analytics processes to interact with data shared within the Total Learning Architecture (TLA) using a simple and powerful web-based graphical user interface.
About the Project
The purpose of DAVE is to explore the use of a novel assessment tool in the context of the TLA, which helps inform the maturation of the “learning ecosystem” concept by generating unique learner data. This project extends xAPI within the TLA by developing models, prototypes, and specifications for analyzing, interpreting, and visualizing data. DAVE’s open source code, reusable by developers and learning engineers, is modular and aligned to the capabilities of xAPI and xAPI Profiles, as well as the needs of the TLA.
The data analytics and visualizations generated by DAVE are able to interact with various components of the TLA, and allow for the extraction, analysis, packaging, and presentation of xAPI data from within and across a TLA-enabled environment. By increasing consistency in the way data is represented and visualized throughout the TLA, this project will increase the value of xAPI as a data asset for implementers and end users of distributed learning.
DAVE’s open source prototype dashboard enables a user to design a custom data visualization in order to gain insight into a domain-based problem. For example, if an instructor wants to know which questions learners found to be the most difficult on an exam, DAVE can provide the distribution of correct and incorrect answers. This insight could also inform an instructor if a certain distractor answer in a multiple choice test was too effective or not effective enough. Through an intentionally designed user experience (a “wizard”), DAVE guides a user through the process of visualizing and reporting on data, including setting up an entire data portfolio known as a workbook. Once the workbook is created, DAVE can be used to update visualizations and reports automatically.
In 2018 an alpha version of the DAVE framework was released, including an online environment, a workbook of data analytics algorithms, and software which enables these capabilities to run against a data set. This alpha version features a master document detailing the prototype analytics algorithms and data visualization templates which enable any user to create meaningful data visualizations, regardless of technical background. The DAVE alpha prototype is available as an Apache 2.0 open source project on Github for demonstration purposes. The prototype demonstration is best served by the native exemplar dataset, which was designed specifically to allow for seamless exploration of the feature set of DAVE’s prototype data workbooks. In 2019, the DAVE project implemented a beta version of the DAVE framework and an expansion of the standards-based algorithms into a structured, component-based querying mechanism.