Factors Influencing Learning of Computational Thinking Skills in Young Children

• In Event: Understanding and Supporting Young Children's Computational Thinking

Abstract

Objectives or purposes
There is increasing recognition of the importance of computer science education for young children. Computational thinking (CT) skills are the cognitive processes involved in solving problems with computers and other technologies. It is vital that all students, regardless of their background, have an equal opportunity to acquire CT skills. However, little is known about the factors that influence learning and development of this important skill set. This paper explores how demographic and environmental factors relate to students’ acquisition of CT skills.

Theoretic Framework
Wing, 2006 suggested that CT skills are not only beneficial in computer science but also in many other fields. However, women and certain minority groups are under-represented among those who pursue a degree in Computer Science (National Center for Women and Informational Technology, 2020; Google & Gallup, 2016). Females are reportedly more likely to develop technology phobias, which can in theory impact on acquisition of CT skills (Özyurt & Özyurt, 2015; Sullivan and Bers, 2016; Cheng, 2019). Access to technology may vary as a function of race, socio-economic status and other factors (Google & Gallup, 2016). Little is currently known about whether and how these factors influence the acquisition of CT skills in young children.

Methods
Nearly 600 1st and 2nd grade students from the Norfolk, Virginia school system completed a six-week Coding as Literacy (CAL) coding curriculum that included 12-14 hours of instruction with the KIBO educational robotics platform.

Data Sources
Students received the TechCheck assessment (Relkin, et al., 2020) before, during, and after the CAL curriculum. TechCheck is an “unplugged” CT assessment that is developmentally appropriate for children ages 5-9. “Unplugged” refers to the use of challenges that are analogous to those that arise in the course of computer programming that do not require knowledge of computer programming to be solved. Students were asked to share their self-identified gender. Other background and classroom information were collected from the schools with parental consent.

Results
667 students in first and second grade (ages 5-9, Mage= 7.41) from 8 schools and N=47 first and second grade educators participated in this research. 315 students identified as male, n=346 students identified as female, n=6 did not specify a gender. The majority of students identified as Black/African American (42.89%) followed by White (39.73%), Mixed (9.30%), Hispanic (4.19%), Asian/Pacific Islander (3.14%), and Native American (0.75%). We used regression analysis to explore how children’s performance on TechCheck related to their gender, grade, age, race, school, and teacher. We found that grade, age, teacher and school contributed significantly to the regression model. Gender and race/ethnicity were not significant predictors of CT outcomes.

Significance
We examined how students from a variety of demographic and environmental backgrounds perform on an unplugged CT assessment while participating in a coding curriculum designed to promote literacy and CT skills. The finding that age, grade, teacher and school are significant predictors of CT learning can potentially inform the design of better computer science curricula, professional development courses for teachers and CT teaching tools for young children.

Authors

• Emily Relkin, Tufts University

• Marina U. Bers, Tufts University