Educational Assessment: What to Watch in a Rapidly Changing World

Abstract

This article is a written adaptation of the Presidential address I gave at the NCME annual conference in April 2018. The article describes my thoughts on the future of assessment. I discuss eleven likely characteristics of future tests and, for each characteristic, why I think it is important and what to watch with respect to it. Next, I outline what is unlikely to change. The article concludes with a comment about the probable state of education 10 years on and what that state might mean for members of the assessment community.

From substitution to redefinition: A framework of machine learning‐based science assessment

“…Educational games (two articles) are very popular platforms used to facilitate students' science learning. Educational games could involve simulated scenarios and challenging tasks to elicit students' performance reflecting their problem‐solving ability (Bennett, 2018). During the game, log data could be collected and stored automatically so that teachers could investigate how students navigate in the game over time.…”

“…As emphasized by Bennett (2018), nonroutine cognitive labor increasingly characterizes high‐demand STEM careers. As a result, more innovative experiences grounded in diverse constructs need to be embedded in science education.…”

“…Of the 47 studies we reviewed, 51% developed contextualized assessment to enhance the assessment function. More than 85% of the assessments were administered using digital platforms, which align with trends in both large‐ and small‐scale assessment, improving testing efficiency and allowing for measurement of competencies that paper and pencil tests are not able to accomplish (Bennett, 2018).…”

“…However, these assessments are seldom delivered in a formal science learning environment. There is also a large gap in certain kinds of evidence (such as process data regarding cognition, affection, and sensory information, as well as socio‐economic and institutional data) which are important for educational decision making (Bennett, 2018) but are not yet utilized in ML‐based science assessment. ML algorithms can be more powerful compared with traditional statistical models for connecting the observations and interpretation due to fewer requirements to meet the assumptions of parametric models and the ability to deal with “big data”.…”

The Effects of MANSA Historical Board Game toward the Students’ Creativity and Learning Outcomes on Historical Subjects

“…Assessment must be able to accommodate all aspects of student potential. Future assessments will be technology-based, measure new constructs, build on fundamental and richer cognitive models, make better use of more complex tasks, be yourself, improve learning, be accountable, assessments made can be carried out for the long-term, using automated assessments, combining new approaches to modeling and analysis, and preparing more effective reporting (Bennett, 2018). One of the objectives of the assessment is to enable students to actively monitor and manage their learning processes (Xiao & Yang, 2019).…”