Abstract
Digital games provide engaging opportunities to support and assess implicit learning—the development of tacit knowledge and practices that may not be explicitly articulated by the learner. The assessment of implicit learning reveals learning not captured by traditional tests and may be critical to meet the needs of a broad range of neurodiverse learners. This chapter describes tools and methods designed to build implicit game-based learning assessment (GBLA), where research-grounded automated detectors identify implicit learning in gameplay. The detectors are based upon theoretical and empirical underpinnings, including extensive hand-labeling. The authors present a detailed overview of a six-step process for emergent GBLA, which has been applied and refined across multiple game-based learning studies. This chapter also includes a description of the data architecture and tools the authors designed and developed specifically for this approach.
TopBackground
Implicit knowledge is argued to be foundational to the development of explicit knowledge (Polanyi, 1966) and key to understanding how we support and measure learning (Brown, Roediger, & McDaniel, 2014; Busch, 2008; Collins, 2010; Kahneman, 2011; Reber, 1993; Underwood, 1996). This work draws from the theory of implicit (or tacit) knowledge, knowledge demonstrated in everyday activity that may not be articulated by the learner verbally (Berry, 1994; Collins, 2010; Kahneman, 2011; Reber, 1989; Underwood & Bright, 1996). As stated by Polanyi (1966) “We know more than we can tell” (p.4).
In educational research, however, implicit knowledge remains ill-defined and not well studied. Ample research has previously shown that traditional academic assessments do not measure all the cognitive abilities demonstrated by learners in everyday activities (Ginsberg, Lee, & Boyd, 2008; Sternberg, 1996). Examples of early work that shaped the ideas of implicit mathematical abilities include studies of gamblers at the race track (Ceci & Liker, 1986); street children using early algebra skills in their vending of fruits and snacks (Nunes, Schliemann, & Carraher, 1993) and housewives calculating “best buys” at the supermarket (Lave, Murtaugh, & de la Roche, 1984).
Key Terms in this Chapter
Classification Algorithm: Algorithms used to classify new observations into a set of categories based on a training set of data in which categories are known.
Educational Data Mining (EDM): An emerging research discipline that provides a suite of methods and research approaches from statistics, machine learning, and data mining to analyze data collected during teaching and learning.
Computational Thinking: A systematic approach to problem-solving required for computational problems that includes problem decomposition, pattern recognition, abstraction, and algorithm design.
Features: Variables engineered or created from large-scale educational data and used as input to build an educational data mining model.
Detector: An automated model built using educational data mining techniques to identify a construct at scale. Classification algorithms are one type of model used to create detectors.
Game-Based Learning Assessment (GBLA): A method to measure learning that takes place within the natural course of gameplay, usually using log data generated through digital gameplay.
Emergent Assessment: Measures of learning that are designed through observation and analysis of learners’ activity as opposed to an assessment for which correct responses or behaviors have been pre-determined.
Playback Tool: An analysis tool for game-based learning that includes simulated video replay generated from digital log data with convenient tools for efficient observation and hand-labelling of data.
Implicit Learning: The development of foundational knowledge and skills that are not necessarily expressed explicitly by the learner, and thus may not be captured on traditional tests and schoolwork.