Teachers’ goals predict computational thinking gains in robotics
| Date | 13 May 2019 |
| Published date | 13 May 2019 |
| Pages | 308-326 |
| DOI | https://doi.org/10.1108/ILS-05-2018-0035 |
| Author | Eben B. Witherspoon,Christian D. Schunn |
| Subject Matter | Library & information science,Librarianship/library management,Library & information services |
Teachers’goals predict
computational thinking gains
in robotics
Eben B. Witherspoon
Department of Education,
University of Pittsburgh –Learning Research and Development Centre,
Pittsburgh, Pennsylvania, USA, and
Christian D. Schunn
University of Pittsburgh –Learning Research and Development Centre, Pittsburgh,
Pennsylvania, USA
Abstract
Purpose –Computational thinking (CT) is widely considered to be an important component of teaching
generalizable computer science skills to all students in a range of learning environments, includingrobotics.
However, despite advances in the design of robotics curricula that can teach CT, actual enactment in
classroomsmay often fail to reach thistarget. This study aims to understandwhether the variousinstructional
goalsteachers’hold when using thesecurricula may offer one potentialexplanation for disparitiesin outcomes.
Design/methodology/approach –In this study, the authors examine results from N= 206 middle-
school students’pre- and post-tests of CT, attitudinal surveys and surveys of their teacher’s instructional
goals to determine if student attitudes and learning gains in CT are related to the instructional goals their
teachersendorsed while implementing a shared robotics programmingcurriculum.
Findings –The findings provide evidence that despite using the same curriculum, students showed
differential learning gains on the CT assessment when in classrooms with teachers who rated CT as a more
important instructional goal; these effects were particularly strong for women. Students in classroom with
teachers who rated CT more highly also showed greater maintenance of positive attitudes toward programming.
Originality/value –While there is a growing body of literature regarding curricular interventions that
provide CTlearning opportunities, this study providesa critical insight into the role that teachersmay play as
a potential support or barrier to the success of these curricula. Implications for the design of professional
developmentand teacher educative materials that attendto teachers’instructional goals are discussed.
Keyword Robotics
Paper type Research paper
Introduction
Computer science education is now widely considered to be an integral part of a well-
rounded K-12 science, technology, engineering and mathematics (STEM) education. In the
USA, the “Computer Science for All”initiative urges that computer science (CS) learning
This work was supported by a grant from the National Science Foundation, Division of Research on
Learning in Formal and Informal Settings (DRL 1418199). The opinions are those of the authors and
do not represent the policies of the funding agency. The authors would also like to acknowledge Ross
Higashi and Josh Jarvis for their consultation and support with the development and design of the
curricular materials and Mary Kay Stein for her insights and comments on an earlier draft. This
research has been approved by the Human Research Protection Office at the University of Pittsburgh.
ILS
120,5/6
308
Received1 May 2018
Revised3 December 2018
Accepted2 January 2019
Informationand Learning
Sciences
Vol.120 No. 5/6, 2019
pp. 308-326
© Emerald Publishing Limited
2398-5348
DOI 10.1108/ILS-05-2018-0035
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/2398-5348.htm
opportunities be provided not only within specialized elective classes or after-school clubs
but also in general education classes that offer these experiences to every student (Smith,
2016). In part, thispolicy shift is driven by a growing need for some base level ofcompetence
in computing for students to remain competitive in a job market that increasingly requires
computational knowledge and skills, regardless of career trajectory. The USA Bureau of
Labor Statistics (2017) predicts that the fastest growing careers in the coming decade are
likely those that will require some degree of computational literacy and the ability to use
computers and programming logic to solve problems in a variety of applications.
Educational researchers have sometimes used the term computational thinking (CT) to
describe this particulartwenty-first-century skill. A canonical and complete definitionof CT
remains unsettled in the literature, leading some to advocate for the pragmatic approach of
identifying core and peripheralconcepts of CT; core aspects typically include decomposing
problems, designing algorithmic solutions and abstracting those solutions to multiple
contexts (Voogt et al.,2015).Therefore, while many definitions of CT exist, most emphasize
the importance of drawing on heuristicsfrom the field of computer science to solve problems
and applying the knowledge and skills of computer science to solve problems across a
variety of contexts and subjects(Barr and Stephenson, 2011;Wing, 2006).
Educational psychologists have studied the possible cognitive benefits of using
computer science in K-12to develop generalizable problem solving skills like CT for decades
(Klahr and Carver, 1988;Pea and Kurland, 1984). In particular, specific CT concepts from
computer science such as “commands execute in sequence,”“conditional statements
determine if and when to pass control of the program to a new set of commands”and
“programs repeat the commands a setnumber of times or until a condition is met”may be
generalizable across programming languages and contexts. However, still relatively little is
known about particular pedagogicalpractices that might be linked to effective instruction in
this class of generalizablecomputational skills.
Robotics is one field that has been studied by educational psychologists as a learning
environment that could potentially provide authentic opportunities to learn generalizable
computer programmingskills in an applied setting (Grover and Pea, 2013). Relatively recent
advances in the design of educational technologies, informed by research in the learning
sciences, have shown promise in providing students with generative learning experiences
that may help develop the generalizable programming knowledge and skills prioritized by
initiatives such as Computer Science for All (Lye and Koh,2014). For example, block-based
graphical programming languages can reduce syntax errors, allowingnovice programmers
to focus on the logic of their programs control structure (Kelleherand Pausch, 2005;Robins
et al.,2010). Specific to robotics educationalcurricula, virtual simulations such as those used
in the current study can reduce the mechanical errors often introduced by physical robots,
thereby reducing the cognitive load of beginning programmers. Such simulated virtual
curricula have been proven to teach programming and physical robotics, but more
efficiently (Liu et al., 2013b).Additionally, there is emerging evidence that certain features of
these virtual robotics learning environments may be associated with measurable gains in
generalizable CT knowledgeand skills (Witherspoon et al., 2017,2018).
In the context of Computer Science for All, educational robotics programs present
themselves as a convenientoption for school districts aiming to take up this initiative.In the
past few decades, robotics programshave become almost ubiquitous in middle schools and
high schools, both in elective after-school programs and more recently in compulsory
education as the required technology becomes more broadly affordable (Melchior et al.,
2005). However, in many K-12 settings, technology-rich programs like robotics are
implemented within Technology Education (“Tech Ed”) departments, which have
Teachers’
goals
309
To continue reading
Request your trial