Applying Learning Analytics to Support Instruction

Abstract

This chapter highlights the ways in which learning analytics can be used to better understand and improve learning environments, instruction, and assessment (Siemens & Long, 2011). As a set of approaches for engaging in educational research, learning analytics and educational data mining represent relatively new modes of inquiry. The growth of these approaches maps closely to the availability of new forms of data being collected and stored in digital learning environments, administrative data systems, as well as sensors and recording devices. Moreover, the growth of these fields maps closely onto what the National Science Foundation refers to as “data-intensive research,” which encompasses more than learning analytics and educational data mining to include a broad range of social and physical sciences. As new forms of data have emerged (i.e., transaction level data from digital learning environments as well as digital forms of audio, video, and text) and been collected at ever increasing scales, there has been an explosion of efforts to make use of these data for the purposes of research. By and large, most early work beginning in the mid-2000s was directed at exploring research questions that were tractable within highly structured, well-designed digital learning environments like intelligent tutoring systems (ITS; e.g., Koedinger, Anderson, Hadley, & Mark, 1997; VanLehn et al., 2005). The tight alignment between the learning tasks students were expected to engage in and the data that were collected in these environments made them ideal for exploring not just the outcomes of learning but the various ways in which students engaged in learning activities. A basic insight from these early researchers continues to fuel research and efforts to improve instruction—data on students’ learning processes is as useful and sometimes more so than data on students’ learning outcomes. In this chapter, we expand upon this insight and highlight the ways in which data from digital learning environments, administrative data systems, and sensors as well as recording devices can be used to support instruction in real classrooms by reporting on students’ learning activities through various data products (e.g., dashboards). We do so across four cases that represent varying degrees of proximity to instruction. By highlighting these varying degrees of proximity, we intend to demonstrate the multiple ways in which learning analytics can be used to support instruction. Cases 1 and 2 describe efforts to use learning analytics to support instruction 9 Applying Learning Analytics to Support Instruction