This longitudinal study examined changes in Computational Thinking (CT) skills in first and second grade students exposed to a developmentally appropriate coding curriculum. The “Coding as Another Language” (CAL) curriculum spans seven weeks and uses the KIBO robot to engage students in learning that integrates programming and literacy concepts. We compared children receiving CAL (N = 667) to a control group (N = 181) who participated in typical classroom activities without coding (No-CAL). TechCheck, a validated “unplugged” CT assessment suitable for young children regardless of their coding experience, was used to measure CT. Over the course of the study, children who received CAL-KIBO improved on TechCheck (M_change = 0.94, p < .001) whereas the No-CAL group did not change significantly (M_change = 0.27, p = .07). Accounting for demographic factors, baseline performance and classroom (teacher) effects, CAL exposure was a significant predictor of post-test CT scores (p < .01). Improvements in CT measured by TechCheck over seven weeks of the CAL-KIBO curriculum were consistent with approximately six months of development without coding instruction. Secondary analysis stratified by grade revealed decisive evidence that CAL exposure improved scores in first grade and anecdotal evidence that second grade scores improved. The CT domains that showed improvement in children who received CAL-KIBO included algorithms, modularity, and representation. Young children who learned to code improved in solving unplugged problems that were not explicitly taught in the coding curriculum. This provides evidence that a developmentally appropriate curriculum for teaching young children to code can accelerate their acquisition of CT skills.

这项纵向研究调查了接触发展适合的编码课程的一年级和二年级学生的计算思维（CT）技能的变化。“另一种语言编码”（CAL）课程为期七个星期，并使用KIBO机器人让学生参与整合编程和扫盲概念的学习。我们将接受CAL的儿童（N  = 667）与对照组（N  = 181）参加了没有编码的典型课堂活动（No-CAL）的儿童进行了比较。TechCheck是一种经过验证的“不插电” CT评估，适用于幼儿，无论其编码经验如何，都可以用来测量CT。在研究过程中，使用CAL-KIBO的孩子在TechCheck（M_变化 = 0.94， p  <.001），而无CAL组没有明显变化（ M_变化 = 0.27， p  = .07）。考虑到人口统计学因素，基线表现和课堂（教师）影响，CAL暴露是测试后CT评分的重要预测指标（ p  <.01）。通过TechCheck进行的CT改进CAL-KIBO课程的七个多星期与没有编码指导的大约六个月的开发时间是一致的。按年级分层的次要分析显示，有决定性的证据表明CAL暴露可改善一年级的分数，而轶事证据则表明，二年级的分数会提高。在接受CAL-KIBO的儿童中表现出改善的CT域包括算法，模块性和表示形式。学会编码的幼儿在解决编码课程中未明确教授的未插拔问题方面得到了改善。这提供了证据，即开发适合幼儿教学的适合课程的课程可以加快他们对CT技能的掌握。