research

Developing Computational Thinking Skills in Students

Can Man Survive on Mars?

Developing Computational Thinking Skills in Students in TransDisciplinary Units:

An Annotated Bibliography


FRAMING THE ISSUE

As a K-5 technology teacher, a large portion of my curriculum (8 weeks, 7/28 total classes, or 25% of my year) focuses on developing the computational thinking (CT) skills of my students with respect to computer programming. CT has many proposed definitions, however I am privy to Riley and Hunt’s (2014) idea that CT is the way computer scientists think and reason, as well as Garcia-Penalvo et al.’s idea of using an algorithmic, step by step approach, to solve any kind of problems (2016). As an American International School in South Korea, our school uses a standards based assessment approach and in Technology, we use the International Society for Technology Education (ISTE) standards for assessment. A power standard for ISTE is, “Computational Thinker” which requires students to:

Develop and employ strategies for understanding and solving problems in ways that leverage the power of technological methods to develop and test solutions” (“ISTE Standards for Students,” 2016).

This power standard breaks up CT to encompass four realms that relate to a student’s ability to problem solve and make decisions:

a: Formulation of problems and defining technology-assisted solutions

b: Data collection, analyzation, and representation

c: Chunking of problems to extract key information, and developing models

d: Understand automation and algorithmic thinking to create and test automated solutions (“ISTE Standards for Students,” 2016).

As CT is one of the standards that I assess, it is relevant to my practice to answer this question: How do we develop CT skills in elementary, specifically K-5, students?

It is also important to situate this question into the context of my particular school’s pedagogical and curricular framework. My school teaches in a similar style to British Columbian schools in respect to cross-curricular units, however, we call them “TransDisciplinary Units” (TDU). Teachers from the various specialist areas collaborate together on thematic units with the outcome of having students solve an authentic problem. For example, Grade 5 students this unit sought to discover, “what skills and understanding would humans need to survive on mars?” It is clear to see how science, mathematics, language arts, and various other STEM specialties could relate learning to this driving question. The development of CT skills could have great carryover to learning in Science, Mathematics, and Design & Engineering classes as CT models have been shown to be effective for learning math and science concepts (Hambrusch, Hoffmann, Korb, Haugan, & Hosking, 2009).


LITERATURE SEARCH

I started off my search examining “coding or programming AND computational thinking” (in the literature, coding, programming, and CT tend to go hand in hand), and refined my search keywords until I had less than fifty articles to sift through. I also wanted to focus on elementary students in my research, and required that the studies be empirical so as to know whether treatments were supportive of improving CT skills. In the end, these were my search parameters:

        Computational Thinking

AND Programming or Coding

AND Elementary or primary

AND students

NOT secondary

       Scholarly (Peer Reviewed) YES

       2003:2019 Publication Date

      Language: English & French




RESEARCH REVIEW

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.


DISCUSSION & CONCLUSION

Perhaps the most frustrating issue with discovering how to develop CT skills in students is that there is no clear, well defined definition of CT in the literature that has been agreed upon. The studies examined in this annotation seem to be privy to the 3 systems approach of CT that defined CT as both algorithmic thinking skills, using technology and automation to solve problems, and perceiving a situation like a computer scientist would; so it is good to see a resemblance of scholarly consistency when defining CT. Having a more consistent running definition of CT, or at least having the river of scholars beginning to flow in the same direction, will certainly help to aide future research.

There also needs to be a more reliable and valid measurement tool for measuring CT skills, rather than the current method of needing to extrapolate CT skill development outside of programming performance within a specific coding program. However, this may also be a limitation of the CT concept itself in that CT requires a computer programming software in order to fully understand the notion of CT in the first place. This will need to be further discovered.

Lastly, it is useful to see that CT skills can be developed outside of the computing environment like the Tran study suggested, and that CT skills can support further mathematics and science learning through generalized problem-solving skills. Having scientific data to support the divergent capabilities of programming knowledge will provide further support for the inclusion of programming courses within elementary curricula.

REFERENCES

D.D Riley, K.A. Hunt, (2014). Computational thinking for the modern problem solver. CRC Press, Boca Raton, FL, USA (2014)

F.J. García-Peñalvo, D. Reimann, M. Tuul, A. Rees, I. Jormanainen. “An overview of the most relevant literature on coding and computational thinking with emphasis on the relevant issues for teachers”. Belgium TACCLE 3 Consortium (2016)

García-Peñalvo, F. J., & Mendes, A. J. (2018). Exploring the computational thinking effects in pre-university education. Computers in Human Behavior, 80, 407–411.

ISTE Standards for Students. (2016). Retrieved from: https://www.iste.org/standards/for-students

Kolb, D. (1984). Experiential learning: Experience as the source of learning and development. Englewood Cliffs, NJ: Prentice Hall.

Kolb, D. (1999). The Kolb Learning Style Inventory, Version 3. Boston, MA: Hay Group.

Kong, S.-C. sckong@eduhk. h., Chiu, M. M. mingchiu@eduhk. h., & Lai, M. mlai@eduhk. h. (2018). A study of primary school students’ interest, collaboration attitude, and programming empowerment in computational thinking education. Computers & Education, 127, 178–189. https://doi-org.ezproxy.library.ubc.ca/10.1016/j.compedu.2018.08.026

Tran, Y. ytran@georgefox. ed. (2019). Computational Thinking Equity in Elementary Classrooms: What Third-Grade Students Know and Can Do. Journal of Educational Computing Research, 57(1), 3–31. https://doi-org.ezproxy.library.ubc.ca/10.1177/0735633117743918

S. Hambrusch, C. Hoffmann, J.T. Korb, M. Haugan, A.L.Hosking. “A multidisciplinary approach towards computational thinking for science majors”. Proceedings of the 40th ACM technical symposium on computer science education, SIGCSE '09, March 4-7, 2009, Chattanooga, TN USA, ACM, New York, NY, USA (2009), pp. 183-187

Sung, W. W. columbia. ed., Ahn, J. J. columbia. ed., & Black, J. B. columbia. ed. (2017). Introducing Computational Thinking to Young Learners: Practicing Computational Perspectives Through Embodiment in Mathematics Education. Technology, Knowledge & Learning, 22(3), 443–463. https://doi-org.ezproxy.library.ubc.ca/10.1007/s10758-017-9328-x

Wing, J. M. (2006). Computational thinking. Communications of the ACM, 49(3), 33–35. doi:10.1145/ 1118178.1118215.