Nonprogramming Interactive Activities for Preliminary Engagement With Foundational Programming Concepts

• In Event: Power in the Plurality of Pedagogies and Partnerships in "Computer Science for All"

Abstract

OBJECTIVE
Research suggests that novice programmers struggle with understanding of key concepts such as variables, looping, abstraction, and creating expressions involving variables and operators. We examine whether and how designed non-programming digital and unplugged activities for conceptual exploration can support preliminary engagement with and learning of key concepts among middle school students

THEORETICAL FRAMEWORK
Our pedagogical approach in designing student-centered conceptually focused interactive activities drew on constructivism with its emphasis on knowledge construction as well as material engagement and manipulation (Ackerman, 2001). Recognizing the correlations between conceptual ideas in mathematics and programming, we drew inspiration from mathematics education research, where “dynamic mathematics” technologies such as Geometer’s Sketchpad (Jackiw, 1991, 2001), have demonstrated positive effects in conceptual mathematics learning for diverse student populations (Roschelle, et al., 2010). We also draw from learning sciences research to scaffold novice learners’ guided inquiry (Bransford, Brown, & Cocking, 2000) and exploration of concepts in a context separate from programming before they employ them in programming.

METHODS & DATA
Working with middle school teachers, we co-designed and developed a suite of four digital and two unplugged non-programming activities (Table 1) to be embedded in a Scratch-based programming curriculum. Designed as a short and preliminary engagement with variables, expressions, loops, and abstraction (VELA), each of these activities were designed for exploration over one-to-two class periods among student pairs (interspersed with whole-class teacher-led discussions). We also developed worksheets to scaffold students’ guided inquiry and lesson plans to guide teachers. These were piloted with different groups of students over the course of the design effort. Additionally we used Evidence-Centered Design to iteratively design and refine an ‘Middle School Introductory Programming Assessment’ as a pre-post measure. Teacher feedback, student cognitive think-alouds, and language simplification by an ELL expert contributed to refinements.

We then conducted mixed-methods research in three middle school ‘Introductory Computer science (CS)’ classrooms in a diverse urban school district (Table 2). Teachers participated in two days of PD prior to implementing the 3-week curriculum. We gathered data on student demographics and prior academic indicators (State test scores on math and English), pre-posttest, final open-choice Scratch projects, and surveys on student engagement with the activities.

FINDINGS
All three classrooms (N=71) showed significant gains on a pre-post assessment (Table 3). These gains showed no significant correlation with gender, grade, or prior mathematics (when controlling for pretest score) and English scores, thus demonstrating promise of the VELA approach and curriculum to reach all students regardless of gender, SES, or prior academic preparation. Additionally, students scored significantly higher in end-of-course open-choice programming projects compared to a control population from the same district.

SIGNIFICANCE
CS and programming are seen as essential learning in K-12 schools for all learners. In the spirit of epistemological pluralism (Turkle & Papert, 1990) and the mantra of CS for All K-12 learners, we explored innovative pedagogical approaches to aid student success in deeper learning of programming. Our work demonstrates the promise of novel approaches such as interactive non-programming activities for explicit engagement with programming concepts.

Authors

• Shuchi Grover, Looking Glass Ventures

• Patrik Lundh, SRI International