Balanced Designs for Deeper Learning in an Online Introductory Computer Science Course for Middle School Students

• In Event: 65.041 - Design-Based Research: Coming of Age for a Versatile Methodology Supporting Diverse Research and Learner Needs

Abstract

Objectives
What should a first-of-its-kind introductory CS course created on a MOOC platform for blended in-class learning in middle school look like? What is the best blend for balancing online and offline, individual and collaborative, open-ended and directed activities for active, deeper learning?

Using past research in the learning sciences and computing education as a foundation, this research adopted an iterative process to design, refine, and study an online curricular intervention—‘Foundations for Advancing Computational Thinking’ (FACT)—to answer questions on the development of deeper, transferable CT skills among young teens.

Theoretical Perspectives
DBR aids putting learning theory to work in engineering instruction for STEM courses and technology-enhanced learning environments, and investigating curriculum sequences, instructional approaches, activities, and assessments (Barab & Squire, 2004; Bell, Hoadley & Linn, 2004). Initial designs of interventions represent “embodied conjectures” (Sandoval & Bell, 2004) that are then refined, often with help from stakeholders as active “design partners” (Amiel & Reeves, 2008). This is especially vital given newly scaleable MOOC platforms.

Methods
Empirical inquiries on growth of computational thinking skills and perceptions of CS as a discipline were conducted over two iterations (Table 1.1) of designing and teaching FACT (Table 1.2) in a public middle school classroom.

In Study1, teaching the curriculum face-to-face without the constraints of the online medium allowed evaluation of the pedagogical content knowledge (Shulman, 1987) designs of FACT. Using the first iteration to also pilot a unit of the curriculum online allowed us to test the initial set of design and pedagogical conjectures about the use of FACT in a blended classroom setting with a mix of online/offline, individual/collaborative activities.

Student and teacher feedback in Study1 guided creation of FACT on OpenEdX for Study2. For example, video length was shortened from 8-12 minutes to 1-5 minutes; what students called “boring” programming assignments were replaced by game design projects that engendered higher student engagement and learning, a more engaging corpus of videos was curated for helping students understand computing in the world, and open-ended projects were added.

Data Sources
The main data-sources were pre-posttests of computational thinking, a “preparation for future learning” (Schwartz, Bransford & Sears, 2005) transfer test, surveys measuring prior computational experience, pre-post responses to ‘perceptions of computing’ question–“What do computer scientists do?” and surveys for student feedback on various aspects of the course.

Results
Instruments measuring computational thinking showed gains for all students in both studies. More importantly, mixed methods analyses revealed that Study2 students performed as well or moderately better on posttests and ‘perceptions of computing’ questions than Study1 students.

Results and feedback from Study2 are informing further revisions to FACT. Differentiation in pacing and assessments, and balancing learner choice within structured activities remain challenges to be tackled by future DBR iterations.

Scholarly significance
FACT exemplifies the application of learning theory and DBR to MOOC course design. More importantly, this course represents crucial first steps to leverage MOOC platforms in K-12 school settings to address a huge need in providing effective CS education at scale.

Authors

• Shuchi Grover, SRI International

• Roy D. Pea, Stanford University