计算机科学
构造(python库)
调试
语法
背景(考古学)
差异(会计)
计算机化自适应测验
数学教育
人工智能
程序设计语言
心理测量学
心理学
发展心理学
生物
会计
业务
古生物学
作者
Sally A. M. Hogenboom,Felienne Hermans,Han L. J. van der Maas
标识
DOI:10.1080/08993408.2021.1914461
摘要
Background and Context. Valid assessment of understanding of programming concepts in primary school children is essential to implement and improve programming education. Objective. We developed and validated the Computerized Adaptive Programming Concepts Test (CAPCT) with a novel application of Item Response Theory. The CAPCT is a web-based and resource-efficient adaptive assessment of 4489 questions measuring: the understanding of basic sequences, loops, conditions (if & if-else statements), debugging, multiple agents, procedures, and the ability to generalize to a new syntax. Method. Data was collected through an existing online adaptive practice and monitoring system called Math Garden. We collected 14 million responses from 93,341 Dutch children (ages 4 - 13). Findings. The CAPCT demonstrated good psychometric qualities because 75% of the variance in question difficulty was explained by differences in item characteristics. The CAPCT demonstrated robustness against adding new participants and adding new items. Differences in player ability (i.e., understanding of CS concepts) were due to differences in age, gender, the number of items played, and prior mathematical ability. Implications. The CAPCT may be used by teachers to identify the level of programming concept understanding of their pupils, while researchers may use the CAPCT to construct and validate effective teaching resources.
科研通智能强力驱动
Strongly Powered by AbleSci AI