Research On The Design Of Performance Assessment For Deep Learning

The 2017 Horizon Report(Higher Education Edition)proposes to change deep learning from mid-term trends to long-term trends,highlighting the need for higher education teaching to be based on deep learning to cultivate talents in the knowledge economy era.Deep learning,as learning that promotes effective learning transfer and complex problem solving,is a kind of learning that points to higher-order ability and requires corresponding care in learning evaluation,and performance assessment is an important method for evaluating higher-order learning.Based on this,performance assessment is proposed as a way to promote deep learning in this study.Currently,performance assessment has received widespread attention in pre-service teachers’ teaching practice,but its value in assessment curriculum learning has not been fully explored.Against this background,this study adopts the design-based research paradigm to lore how performance assessment can facilitate deep learning and its facilitative effect,taking the normal students’ public course on educational technology as an example.Based on identifying the key elements of performance assessment(objectives,performance tasks,rubrics,and feedback),a theoretical framework of performance assessment for promoting deep learning is constructed and the principles of design and implementation of each element are distilled,supported by Situated Cognition theory and expert-novice research,and combined with Biggs’ 3P model of learning.In the first round of the study,the objectives were organized around the big ideas of"technology" and "What should be done in teaching integrated technology",and the corresponding performance tasks and rubrics are designed.During the implementation process,performance assessment is embedded into curriculum and instruction to practice the concept of ’teaching-learning-assessment’ integration and use the improvement of students’work as a clue,while focusing on student self-regulated learning.The study analyzed the impact of performance assessment on normal students’ deep learning processes and outcomes in four ways:firstly,using the revised two-factor Study Process Questionnaire(RSPQ-2F)of Biggs as a measurement tool,paired samples T-test was performed on the collected pre-test and post-test data,the results show a significant increase in normal students’ propensity for deep learning motivation and strategies and no significant change in surface learning motivation and strategies.Secondly,improvement of the performance task outcome(PPT production)and paired samples T-tests on the first and last scores,the results indicated that normal students’ PPT design and production skills had significantly improved;Thirdly,two researchers evaluated the level of cognitive structure reflected in the first draft and finished product of the informational teaching design based on the SOLO evaluation form and found that after the mid-term presentation and teacher feedback,eight(out of ten)of the improved works were above the relational structure,indicates that students are generally achieving deep learning;Fourthly,semi-structured interviews were used to further demonstrate the occurrence of deep learning from students’ perspectives and to uncover the reasons why performance assessment promoted deep learning.The second round of the study focused on optimizing the design and implementation.Based on a stable initial design,iterative improvements based on the principle of"refinement" for issues reflected in the results of the first round of implementation,including task design that is more responsive to students’ needs,integrated design that focuses on the integration between performance tasks,self-reflection based on the rubrics and optimization of course arrangements of the course.The deep learning process,cognitive level,and interview data were analyzed by mixed research methods,and the results showed that the optimized performance assessment promoted the deep learning process and results of normal students.This study summarized the key points in the design and implementation of performance assessment to promote deep learning as follows:the objectives of the performance assessment point big ideas;performance tasks that are contextualized,open-ended,integrated,and adapted to students’ needs are designed to match the objectives;Rubrics should be used throughout the performance assessment to promote the processing process of performance tasks;the implementation phase is guided by work improvement,with reflection and presentation sessions.