Introducing Computational Thinking to Young Learners: Practicing Computational Perspectives Through Embodiment in Mathematics Education
Woonhee Sung • Junghyun Ahn • John B. Black (2017)
The purpose of the authors’ study was to identify key factors in the design of elementary lessons that allow for the integration of CT skills into non-computing domains. Using a Pre-post mathematics test, the authors examined two K-1 classrooms that consisted of 66 underrepresented minority students. Using a randomly assigned, 2x2 factorial experiment (4 experimental groups), the authors designed a coding program using the iPad App, “Scratch Jr.”, to examine two factors:
Factor one: they embraced a constructionist, “embodied approach”, and looked at whether full body movement (embodied), role play, and hands on approaches were more powerful for learning abstract STEM subjects versus low-embody styles (hand gestures);
Factor two: they examined the importance of “computational perspective taking” (CPP), thinking like a computer scientist. High CPP had students programming a surrogate (machine or character) to solve the problem, while low CPP had students simply walk through the code themselves.
The authors found that the level of embodiment used had a statistically significant positive impact on student mathematics scores, as did high CPP. The Full embodied with CPP significantly outperformed the low-embody and no CPP group as well. The authors also found that high CPP instruction increased the accuracy of student programming skills.
The authors took significant rigor in randomly assigning their control, as with their experimental design. However, they introduced a major confounding variable in the fact that they themselves taught the four different lessons, and may have been influenced to be more enthusiastic about the high embodied/high CPP group than when they taught the low-embodied/low-CPP group, resulting in lower student achievement. If they had trained other teachers to teach the curriculum without knowing the goal of the study, the reliability of their findings could have been improved.
As the authors showed, CT skills and programming are important to the mathematics and science classrooms as programming also teaches planning abilities and the problem-solving process (Wing, 2006). This is evident by the statistically significant increases in the various groups, though as mentioned above, this should be viewed hesitantly as instructor bias was almost certainly present to influence the data.
Computational Thinking Equity in Elementary Classrooms: What Third-Grade Students Know and Can Do
Yune Tran (2019)
Tran’s study was concerned with two research questions (2019, p. 4):
What changes, if any, are evident in third-grade students learning of foundational CS concepts and CT over 10 weeks of coding lessons?
How can 10 weeks of coding lessons influence third-grade students’ CT in and out of school?
To answer these questions, Tran exposed over 200 elementary students to a 10-week, puzzle based code.org coding curriculum and examined a pre-post test assessment on CT and computer science (CS) skills. There was no control group as this was the first intervention of its kind in Oregon, USA, and the 13 third grade classrooms were located in suburban and rural areas. Code.org was an affiliate of this study and was present in the decision making process of classrooms chosen, a conflict of interest in this study as the curriculum used was Code.org’s.
Tran examined the students using Kolb’s constructivist style experiential framework of Feeling > Watching > Thinking > Doing (1984,1999).
Tran found that after her intervention, there was a significant improvement in CT skills based on her self-created pre-post test of CT and CS skills. Student motivation and positive outlook on coding was also significantly improved post test, as is evident from the interview findings; Lastly, students noted in interviews that their teamwork, cooperation, and resiliency skills improved from the partner coding challenges.
A large limitation of Tran’s study is the measurement of CT. Tran, in collaboration with her university, used a self created model for measuring pre-post test scores with an internal reliability of .63 and .61 on pre-post tests respectively (and she notes this is a problem). With low internal reliability, the findings should be viewed hesitantly.
As well, since CT has not been solididly defined in the literature with many competing opinions, measuring CT tends to be done on a program by program basis, and the aptitude a student possesses within this program. As such, having an in-depth review of CT skills is difficult with changing definitions from scholar to scholar. This muddied waters means that the improvements in CT skills should be taken with caution.
That being said, the improvements to positive attitudes towards CT programs, problem-solving skills, and interest in STEM fields seems well supported based on interview responses. Whether this increase will survive in the future for these students is uncertain.
As noted by Tran, CT development initiatives have been largely in secondary schools with little emphasis on elementary CT skill development, in the USA at least. However, we know that earlier engagement with STEM concepts increases student motivation and initiative to learn STEM skills (Tran Y., 2019). The importance of early CT skills development is likely to further CT further down a student’s educational journey.
A Study of Primary School Students' Interest, Collaboration Attitude, and Programming Empowerment in Computational Thinking Education
Siu-Cheung Konga • Ming Ming Chiub • Ming Lai (2018)
Building upon Seymour Papert’s conception of CT and its proposed ability to empower students, the authors of this study sought to define and measure “programming empowerment” to fill the gap in measurement of CT skills. Operationally, they define CT similar to the initial proposed definition in this paper, and they defined programming empowerment to compose of four components: meaningfulness, impact, creative self-efficacy, and programming self-efficacy (p. 1). Though part of a larger, unpublished as of this writing, study on the promotion of CT skills in elementary schools, this specific portion of the study sought to answer if greater interest in computers, and more positive outlooks on collaboration, led to greater programming empowerment in students.
The 30m likert-scale survey was completed online with 287 Gr 4-6 students. The survey was satisfactory in its rigorous analysis, as well as found to be reliable to measure the constructs designed to measure.
Researchers found that their data supported their initial hypothesis that a student with greater interest in programming also viewed programming as more meaningful, impactful, and had greater creative self-efficacy and programming self-efficacy. However, more positive attitudes towards collaboration suggested higher creative self-efficacy, but not greater programming self-efficacy. The data also supported the hypothesis that interest was critical to programming empowerment, and that older students viewed programming training as less meaningful, and that boys showed more interest in programming that girls did.
The minor flaw in this study is that the instrument used is only mentioned to be validated by experts, but what this means or what rigour was used in the study of the reliability of this tool was not discussed. The authors did include the full measurement tool for examination.