skip to main content

Title: Analyzing Students' Design Solutions in an NGSS-aligned Earth Sciences Curriculum
This paper analyzes students’ design solutions for an NGSS aligned earth sciences curriculum, the Playground Design Challenge (PDC), for upper-elementary school (grade 5 and 6) students.We present the underlying computational model and the user interface for generating design solutions for a school playground that has to meet cost, water runoff, and accessibility constraints. We use data from the pretest and posttest assessments and activity logs collected from a pilot study run in an elementary school to evaluate the effectiveness of the curriculum and investigate the relations between students’ behaviors and their learning performances. The results show that (1) the students’ scores significantly increased from pretest to posttest on engineering design assessments, and (2) students’ solution-generation and testing behaviors were indicative of the quality of their design solutions as well as their pre-post learning gains. In the future, tracking such behaviors online will allow us to provide adaptive scaffolds that help students improve on their engineering design solutions.
; ; ;
Award ID(s):
Publication Date:
Journal Name:
Proceedings of the 20th International Conference on Artificial Intelligence in Education
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract Background

    Despite the importance of understanding the mechanism of natural selection for both academic success and everyday decision-making, this concept is one of the most challenging to learn in contemporary science. In addition to cumulative socio-cultural influences, intuitive cognitive biases such as the teleological bias—the early developing tendency to explain phenomena in terms of function or purpose—contribute to the difficulty of accurate learning when the process is taught in high school or later. In this work, we therefore investigate—for the first time—the viability of a teacher-led classroom-based storybook intervention for teaching natural selection in early elementary school. The intervention was designed to counteract teleological explanations of adaptation. In consequence, we specifically examined the nature and extent of elementary school children’s teleological reasoning about biological trait change before and after this intervention.


    Second and third grade students demonstrated a variety of misunderstandings at pretest, including teleological preconceptions. Most of these teleological ideas were explicitly accompanied by incorrect mechanistic ideas, confirming that the teleological reasoning observed in this young sample reflected fundamental misunderstandings of adaptation as a goal-directed event. Overall, learning from the classroom intervention was substantial, with students performing significantly better on all measures of natural selection understanding at posttest. Interestingly,more »explicit teleological reasoning displayed at the pretest did not have a differentially greater impact on learning than other kinds of marked pretest misunderstandings. One explanation for this might be that children displaying teleological misunderstandings at pretest also tended to demonstrate more biological factual knowledge than other students. Another explanation might be that pretest misunderstandings that were not overtly teleological were, nevertheless, implicitly teleological due to the nature of the mechanisms that they referenced. The differential impact of teleological preconceptions on learning might therefore have been underestimated.


    In summary, early elementary school children show substantial abilities to accurately learn natural selection from a limited but scalable classroom-based storybook intervention. While children often display explicit teleological preconceptions, it is unclear whether these ideas represent greater impediments to learning about adaptation than other substantial misunderstandings. Reasons for this, and limitations of the present research, are discussed.

    « less
  2. Children’s math self-concepts—their beliefs about themselves and math—are important for teachers, parents, and students, because they are linked to academic motivation, choices, and outcomes. There have been several attempts at improving math achievement based on the training of math skills. Here we took a complementary approach and conducted an intervention study to boost children’s math self-concepts. Our primary objective was to assess the feasibility of whether a novel multicomponent intervention—one that combines explicit and implicit approaches to help children form more positive beliefs linking themselves and math—can be administered in an authentic school setting. The intervention was conducted in Spain, a country in which math achievement is below the average of other OECD countries. We tested third grade students ( N = 180; M age = 8.79 years; 96 girls), using treatment and comparison groups and pre- and posttest assessments. A novelty of this study is that we used both implicit and explicit measures of children’s math self-concepts. For a subsample of students, we also obtained an assessment of year-end math achievement. Math self-concepts in the treatment and comparison groups did not significantly differ at pretest. Students in the treatment group demonstrated a significant increase in math self-concepts from pretestmore »to posttest; students in the comparison group did not. In the treatment group, implicit math self-concepts at posttest were associated with higher year-end math achievement, assessed approximately 3 months after the completion of the intervention. Taken together, the results suggest that math self-concepts are malleable and that social–cognitive interventions can boost children’s beliefs about themselves and math. Based on the favorable results of this feasibility study, it is appropriate to formally test this novel multicomponent approach for improving math self-concepts using randomized controlled trial (RCT) design.« less
  3. In the ConnecTions in the Making project, researchers and district partners work to develop and study community-connected, integrated science and engineering curriculum units that support diverse elementary students’ science and engineering ideas, practices, and attitudes. In the community-connected units, students in the third, fourth, and fifth grades use human-centered design strategies to prototype and share functional solutions to a design challenge rooted in the students’ local community while scientifically exploring the phenomena and mechanisms related to the challenge. One of the units is “Accessible Playground Design,” a grade three unit that engages students in designing a piece of accessible playground equipment. It comprises 10 lessons, approximately 1 hour each, including a launch lesson, followed by four inquiry and four engineering design lessons, and a final design exposition.
  4. > Context • In 1972, Papert emphasized that “[t]he important difference between the work of a child in an elementary mathematics class and […]a mathematician” is “not in the subject matter […]but in the fact that the mathematician is creatively engaged […]” Along with creative, Papert kept saying children should be engaged in projects rather than problems. A project is not just a large problem, but involves sustained, active engagement, like children’s play. For Papert, in 1972, computer programming suggested a flexible construction medium, ideal for a research-lab/playground tuned to mathematics for children. In 1964, without computers, Sawyer also articulated research-playgrounds for children, rooted in conventional content, in which children would learn to act and think like mathematicians. > Problem • This target article addresses the issue of designing a formal curriculum that helps children develop the mathematical habits of mind of creative tinkering, puzzling through, and perseverance. I connect the two mathematicians/educators – Papert and Sawyer – tackling three questions: How do genuine puzzles differ from school problems? What is useful about children creating puzzles? How might puzzles, problem-posing and programming-centric playgrounds enhance mathematical learning? > Method • This analysis is based on forty years of curriculum analysis, comparisonmore »and construction, and on research with children. > Results • In physical playgrounds most children choose challenge. Papert’s ideas tapped that try-something-new and puzzle-it-out-for-yourself spirit, the drive for challenge. Children can learn a lot in such an environment, but what (and how much) they learn is left to chance. Formal educational systems set standards and structures to ensure some common learning and some equity across students. For a curriculum to tap curiosity and the drive for challenge, it needs both the playful looseness that invites exploration and the structure that organizes content. > Implications • My aim is to provide support for mathematics teachers and curriculum designers to design or teach in accord with their constructivist thinking. > Constructivist content • This article enriches Papert’s constructionism with curricular ideas from Sawyer and from the work that I and my colleagues have done« less
  5. This study investigated patterns in the development of computational thinking practices in the context of the Exploring Computer Science (ECS) program, a high school introductory CS course and professional development program designed to foster deep engagement through equitable inquiry around CS concepts. Past research indicates that the personal relevance of the ECS experience influences students' expectancy-value towards computer science. Expectancy-value is a construct that is predictive of career choices. We extended our research to examine whether expectancy-value influences the development of computational thinking practices. This study took place in the context of two ECS implementation projects across two states. Twenty teachers, who implemented ECS in 2016–17, participated in the research. There were 906 students who completed beginning and end of year surveys and assessments. The surveys included demographic questions, a validated expectancy-value scale, and questions about students' course experiences. The assessments were developed and validated by SRI International as a companion to the ECS course. Overall, student performance statistically increased from pretest to posttest with effect size of 0.74. There were no statistically significant differences in performance by gender or race/ethnicity. These results are consistent with earlier findings that a personally relevant course experience positively influences students' expectancy for success.more »These results expanded on prior research by indicating that students' expectancy-value for computer science positively influenced student learning.« less