Developing Computational Thinking in Early Ages: A Review of the

code.org Platform

Rolando Barradas

1,2,4 a

, José Alberto Lencastre

, Salviano Soares

1,4 c

and António Valente

1,2 d

School of Sciences and Technology, University of Trás-os-Montes and Alto Douro, Quinta de Prados, Vila Real, Portugal

INESC TEC, Porto, Portugal

CIEd - Research Centre on Education, Institute of Education, University of Minho, Campus de Gualtar, Braga, Portugal

IEETA, UA Campus, Aveiro, Portugal

Keywords: Computational Thinking, Early-age, code.org, Technology-enhanced_learning.

Abstract: This article reports a pedagogical experience developed within the scope of a Ph.D. program in Electrical and

Computer Engineering with application to Education. Starting with a contextualization on the evolution of

computers and Computational Thinking, the article describes the platform used in this study - code.org -,

highlighting the strengths that captivate the students. In the Case Study topic, we describe the study carried

out, starting with a description of the students involved, followed by a description of the process and the

analysis of the results, ending with the evaluation process performed by the students. The article ends

concluding that code.org is a valid option to develop computational thinking at early-ages.

1 INTRODUCTION

Computational thinking is considered a key asset in

the XXI century, as it allows to increase the analytical

capacity of children in different areas of knowledge

(Wing, 2006; Resnick, 2012) as well as to promote

skills such as abstract, algorithmic, logical and

scalable thinking (Resnick, 2012). These skills,

associated with computer sciences, are transposed to

other areas of knowledge and, as a consequence, to

the daily lives of young people, making them more

reflective and critical, thus better prepared for the

world (Brennan & Resnick, 2012).

Within the scope of a Ph.D. program in Electrical

and Computer Engineering with application to

Education, to develop computational thinking at an

early age (Coelho, Almeida, Almeida, Ledesma,

Botelho & Abrantes, 2016) and using code.org, a

well-known coding platform for children, we

developed a pedagogical experience whose results are

revealed in this article.

https://orcid.org/0000-0001-9399-9981

https://orcid.org/0000-0002-7884-5957

https://orcid.org/0000-0001-5862-5706

https://orcid.org/0000-0002-5798-1298

2 CONTEXTUALIZATION

Computers are everywhere, from our desktops to our

pockets. The various physical forms in which we find

them make it easy to use them daily in fields as

diverse as industry, science, education, and

entertainment. Today, most of us carry in our pockets,

a smartphone with millions of times more computing

power than NASA (National Aeronautics and Space

Administration) had in 1969 when they placed a

mission on the Moon (Puiu, 2019). From the

construction of the first computer in 1946, the

military-developed Electronic Numerical Integrator

and Computer (ENIAC) to the present day, when the

use of computers is a constant in our daily lives, the

processing power of computers has continuously

increased. Yet we rarely think about this evolution,

since for most users today all this technology and

processing power it's taken for granted as they’ve

always lived surrounded by it.

The massification of computers has made them

easily usable by anyone, regardless of age,

educational background or professional occupation.

Barradas, R., Lencastre, J., Soares, S. and Valente, A.

Developing Computational Thinking in Early Ages: A Review of the code.org Platform.

DOI: 10.5220/0009576801570168

In Proceedings of the 12th International Conference on Computer Supported Education (CSEDU 2020) - Volume 2, pages 157-168

ISBN: 978-989-758-417-6

157

But at the same time has made it almost imperative

for everyone to have, at least, basic computer skills,

that include Internet and email, word processing,

graphics and multimedia, and spreadsheets. Thanks to

a wide range of intuitive applications and interfaces,

computers can now be used without expert

knowledge to solve complex problems or technical

tasks as well as for the most varied situations of

everyday life. However, as stated by Resnick (2012)

and Resnick, Maloney, Monroy-Hernández, Rusk,

and Astmon (2009), one of the biggest challenges that

users face nowadays is to need to stop from being

mere content consumers (programs, games, etc.) to

become creators of such contents. To do so, one needs

only to be aware of certain applications, sufficiently

curious to search for information on the Internet or

able to do so through a trial and error process.

Since the beginning of the millennium, ICT

(Information and Communication Technologies)

classes have become compulsory in schools.

However, the study of ICT focuses on the

transmission of knowledge about computer tools,

software and hardware, and their use to solve

everyday problems, such as writing a text, searching

for information on the internet and learning how to

communicate it efficiently. A natural evolution would

be to make users understand how these tools work and

how they are built.

Coding is a way of developing creative activities

with children. Still, it also allows them to gain a

broader view of computer uses, creatively solving

real-world problems, by focusing primarily on

design, planning, and implementation of a particular

project.

In this context, it becomes necessary to mention

an essential competence for the 21st century:

computational thinking (Wing, 2007).

2.1 Computational Thinking

Computational thinking can be defined as the set of

processes involved in formulating a problem and its

solutions so that a computer (human or machine) can

effectively solve it (Wing, 2017) and it’s more

connected to conceptualization than to coding itself

(Wing, 2007).

The development of computational thinking

promotes skills such as (i) abstract thinking - using

different levels of abstraction to understand problems

and solving them -, (ii) algorithmic thinking - the

expression of solutions in different stages whose goal

is to find the most effective way to solve a problem -

(iii) logical thinking - formulating and excluding

hypotheses - and (iv) measurable thinking -

breaking up a big problem into small parts or

composing small parts to formulate a more complex

solution (Resnick, 2012).

Brennan and Resnick's studies on computational

thinking and the creation of interactive media

products gave rise to a framework of reference for

studying and evaluating the development of

computational thinking that encompasses three

dimensions: (i) computational concepts, (ii)

computational practices and (iii) computational

perspectives.

Regarding (i), computational concepts, Brennan

and Resnick (2012) identified seven, namely:

Sequences – A set of steps or instructions that can be

executed to complete a coding task, which is why it is

important to define the correct order of execution of

the commands since changing one of the commands

could lead to completely different results;

Loops - These are mechanisms that allow you to

execute the same sequence several times. When

solving certain problems, it is possible to identify

certain patterns of repetition;

Events - An event is a certain occurrence that causes

a certain action to happen;

Parallelism - To solve certain problems several

sequences may need to take place at the same time;

Conditionals - Conditionals allow a program to make

decisions, through the use of decision structures;

Operators - They are used to express and solve

mathematical and logical operations;

Data – Data structures are used to store, retrieve and

update values stored in variables.

The (ii) computational practices, associated

with the act of programming, are focused on the

construction process, on thinking and learning,

changing the focus from what is learned to how it is

learned (Brennan & Resnick, 2012). Thus, four sets

of practices were defined:

Being Incremental and Iterative - Action by which

children develop, check whether a project works and

continue to develop new approaches to the solution;

Testing and Debugging - It is through trial-error

processes, analysis of previously made situations, that

children check what does not work and correct errors;

Reusing and Remixing - You also learn when you

build something using old projects or projects that

others have already done;

Abstracting and Modularizing - The act of building

something big by joining sets of smaller parts, since

complex problems can be divided into smaller,

simpler problems.

CSEDU 2020 - 12th International Conference on Computer Supported Education

158

Regarding (iii) computational perspectives,

Brennan and Resnick identified three major

perspectives of children in their relation to

computing, categorized as:

Express - computing is a means of creation and self-

expression and allows students to start to see

themselves as builders and not just consumers;

Collaborate / Connect - computing allows you to

create with and for the others, making your creations

an inspiration for your new projects of those of others,

allowing the development of a critical spirit;

Questioning - Questioning technology with

technology. Trying to understand how certain

problems were solved can lead to questioning the

functioning of other situations in the real world

(Brennan, Chung & Hawson, 2011).

Therefore, it will be possible to evaluate the

development of computational thinking in young

people, analyzing the execution of projects/activities,

designed taking into account the three dimensions

defined.

Inevitably linked to computational thinking is the

problem-solving method (Jonassen, 2011). Using

coding as a way to develop computational thinking

also allows you to stimulate students' creativity by

solving real-world problems. According to Jonassen

(2011), problem-solving skills development activities

are the most relevant educational activities that

students can perform because the knowledge built

during the process is better understood and more

easily retained.

Using the problem-solving method leads students

to 'learn how to learn' (Papert, 1993) while looking

for the solution of a problem instead of waiting for an

answer given by the teacher, thus developing their

domain of the procedures (Echeverría & Pozo, 1998).

The use of this teaching method increases the

motivation of students who become the main agent in

the learning process.

3 THE code.org PLATFORM

Founded in 2013, code.org (Figure 1) is a nonprofit

dedicated to expanding access to computer science in

schools and increasing participation by women and

underrepresented minorities.

The platform code.org maintains a very broad set

of educative resources and tools that can be executed

in almost all platforms, including smartphones and

tablets which makes it very flexible and easy to use.

Its vision is that every student in every school should

have the opportunity to learn computer science, just

like biology, chemistry or algebra (Code.org, 2019).

Figure 1: Code.org Platform Home Screen.

But what makes the platform so attractive to students?

In our study we identified several details about the

platform that we consider important to mention:

3.1 The Easy Process of Accessing the

Content

Access to the platform has been built to respond to

almost any existing combination today. Students can

start testing the platform without creating an account.

More traditionally, students can use an account

created on the platform. Additionally, they can also

use a pre-existing account from other platforms, such

as Google, Facebook, and Microsoft, as can be seen

in Figure 2.

Figure 2: The various login possibilities on the code.org

platform.

Any situation in which the student logs in with an

account allows them to save their progress so that

they can continue solving tasks, and also allows the

teacher, through the creation of a class and assigning

students, to monitor the learning process of each one

individually.

Additionally, the fact that the platform is built as

a web application that runs entirely in a browser

environment means there is no need to pre-install or

configure the devices where the students will work.

The whole process of learning and development of

computational thinking is done through a graphical

environment in which students build their algorithms

Developing Computational Thinking in Early Ages: A Review of the code.org Platform

159

by dragging and dropping instruction blocks to solve

each of the challenges (Figure 3).

Figure 3: Coding by blocks.

3.2 Course Organization and

Self-efficacy Control Tools

The Computer Science Fundamentals, Course 2, the

object of this study, is organized in 19 Lessons, both

online and offline, organized by logical order and

with increasing difficulty of contents. Note that for

this study, only online lessons finishing with Lesson

16 - Flappy Bird - were used, as students had

previously taken offline coding classes (CS

Education Research Group, 2015). Every lesson starts

with an introductory video that explains the lesson’s

objectives and gives students some insight into the

type of problems they will encounter. Starting with

simple sequences and covering all the computational

concepts identified by Brennan and Resnick (2012),

the course ends with slightly more complex concepts

such as nested loops and functions.

Figure 4: Viewing progress for self-efficacy control.

While performing a coding task, students are given

the maximum number of blocks that should be used

to find the optimal solution in each of the problems.

If a student can solve a particular problem using the

optimal solution, the problem number will be filled in

green, on their progress overview. If they do solve it

but not with the optimal solution, the problem number

is filled with a light green color, as seen in Figure 4,

so that the student can return to that program at a later

time and find the optimal solution.

This course overview allows students to have a

tool for monitoring their effectiveness, allowing them

to compare with their peers and serving as motivation

for task solving.

3.3 Known Characters

All challenges present in code.org were created using

animations, sounds, and characters according to the

age group of the target audience.

Figure 5: Example of a challenge involving familiar student

characters.

The fact that the platform uses, in many of its

challenges, familiar themes and characters such as

Angry Birds and Plants vs. Zombies, pictured in

Figure 5, makes the challenges even more attractive

to students.

3.4 Instant Feedback

Kapp, Blair, and Mesch (2012) talking about

gamified instruction scenarios, stated that the instant

feedback was one of the most important elements in

a Gamification system.

Figure 6: Instant Feedback Associated with Positive

Reinforcement.

The code.org platform has interpreted these

indications very well by designing a gamified system

that maximizes feedback, notifying the student at all

times about what they have already completed and

what mistakes they have made (as seen earlier in the

self-efficacy control tools). In addition to this task

completion message, the system provides immediate

corrective feedback. When students are successful in

their task they are immediately rewarded with

messages and sounds that indicate joy and success, as

can be seen in the screen shown in Figure 6.

CSEDU 2020 - 12th International Conference on Computer Supported Education

160

On the other hand, when something does not go

so well, the student is informed immediately with a

message of encouragement, as seen in Figure 7.

Figure 7: Instant feedback associated with non-success.

This instant feedback allows students to immediately

view their progress and compare it with others.

3.5 Reward System

In addition to the already mentioned elements of

motivation, and following the same principles of

Gamification and reward system set out by Kapp,

Blair, and Mesch (2012), code.org implemented a

feature that allows students to print a course

certificate, similar to the one shown in Figure 8, when

they complete all Lessons.

Figure 8: Example of a certificate issued by the platform.

4 CASE STUDY

The study group consisted of 130 students from five

4th grade classes (9 and 10 years old) over two school

years, divided as follows: 2017/2018, 28 students

from Class 1 and 28 students from Class 2;

2018/2019, 25 students from Class 3, 26 students

from Class 4 and 23 students from Class 5, out of a

total of 130 students. Of those, 62 were females and

68 were males.

In the academic year 2017/2018, the classes were

taught in a pedagogical pair regime, with the ICT

teacher always assisted by the headteacher of the

classes. However, it was not possible to maintain this

organization of the classes in the 2018/2019 academic

year, so they were taught only by the ICT teacher.

Initially, the basic concepts of the code.org

platform were taught to all students. - i.e., how the

blocks fit together, workspace locations, interface

details such as the action stage and execution buttons,

as well as the use of platform access credentials.

Subsequently, computational thinking was developed

through hands-on laboratory problem-solving

exercises (Jonassen, 2004) involving sequences,

loops, parallelism, events, conditionals, operators,

and data, which would allow students to create their

first Flappy game.

Data on the execution of these tasks were

collected through the platform's automatic records for

statistical processing. Since the goal set for the course

was for students to create their Flappy Bird, for this

statistical treatment we only considered the online

exercises from the first 16 lessons of code.org 's

Computer Science Fundamentals, Course 2. Also

used as instruments for data collection were the

descriptive syntheses of the students elaborated by the

teachers of the classes, an online questionnaire for the

students and the notes of the ICT teacher on how the

classes worked.

4.1 Results

The results of the 115 different problems were

evaluated for each of the 130 students involved in the

study, in a total of 14950 problems. The overall

results are summarized in Table 1.

Table 1: Overall Test Results.

No. of different

problems/total

analyzed

Completion

Rate

Completion

Rate with

non-optimal

solution

Global

Results

115/14950 85.8% 3.7%

In this first analysis, we found that the percentage of

problems solved is 85.8%, and only 3.7% of them did

not get the optimal solution to the problem, which is

a very positive result.

Analyzing the results according to the three

dimensions of Brennan and Resnick's (2012)

framework, (i) computational concepts, (ii)

computational practices and (iii) computational

perspectives, we obtained the following results:

(i) Computational Concepts

It is important to note that the number of problems

indicated in Table 2 refers to the number of exercises

in which each of the concepts was approached and

that the same problem could address more than one

concept. It is also possible to observe a disproportion

between the number of problems that work some

Developing Computational Thinking in Early Ages: A Review of the code.org Platform

161

Table 2: Test results, grouped by computational concepts.

Concept

No. of

different

problems/total

analyzed

Completion

Rate

Completion

Rate with

non-optimal

solution

Sequences 46/5980 89.0% 2, 0 %

Loops 67/8710 86.5% 5.5%

Events 10/1300 69.7% 0%

Parallelism 10/1300 69.7% 0%

Conditionals 15/1950 72.8% 2.8%

Operators 25/3250 71.6% 1.7%

Data 10/1300 69.7% 0%

concepts. However, this fact was already expected

since this was an initiation course so more complex

concepts like Events, Parallelism and Data were only

addressed in the final part of the course. The

understanding of the Sequences was the skill that

students had less difficulty to acquire, with a

completion rate of 89.0%. Of these, only 2.0% of the

results were not an optimal solution. On the opposite

side, the Events, Parallelism and Data skills were

those in which students had the most difficulties.

Despite this, the completion rate is quite positive, at

69.7%, of which 100% reached the optimal solution.

(ii) Computational Practices

All computational practices mentioned by Brennan

and Resnick (2012) were addressed while solving the

proposed problems, although, due to the type of

problems present in the course, not all computational

practices were given equal emphasis. Also, in this

case, it is important to note that the number of

problems indicated in Table 3 refers to the number of

exercises in which each practice was addressed and

that the same problem could address more than one

practice.

Table 3: Test results, grouped by computational practices.

Practice

No. of

different

problems/total

analyzed

Completion

Rate

Completion

Rate with

non-optimal

solution

Being

incremental

and iterative

82 /10660 89.7% 4.0%

Testing and

debugging

23 /2990 78.9% 3.9%

Reusing and

remixing

67 /8710 86.5% 5.5%

Abstracting

and

modularizing

10 /1300 69.7% 0%

Analyzing the results grouped by computational

practices, it is possible to observe that the practices of

Being incremental and iterative were the most

addressed throughout the course with 82 different

problems. It was also in these practices that students

showed less difficulty with a completion rate of

89.7%. The students also demonstrated to be

comfortable with the practices that involved Reusing

and Remixing. In these problems, they managed to

obtain a completion rate of 86.5%, with only 5.5% of

these not reaching the optimal solution.

The practices that involved Abstracting and

modularizing were those in which students obtained

fewer good results. However, it was not possible to

assess whether this is due to actual abstraction

difficulties or whether, on the other hand, it is simply

because the tasks dealing with this practice were the

most complex exercises, therefore, the last to be

solved, and, given the slow pace of some students,

there was no time to solve them.

(iii) Computational Perspectives

The three computational perspectives - Express,

Collaborate, and Question - were cross-sectional

throughout the process, although they were not

objectively measured. Although the students

developed the problem-solving tasks following the

guidelines, in some tasks the freedom to create

something new and/or personalize the existing one -

Express - was implicit through the inclusion of

personal elements in the provided scenarios.

Although the work was mostly done individually,

typically as soon as a lesson ended, the students

endeavored to help their most delayed colleagues by

doing peer work - Collaborate. Also, curiosity about

the processes and the different problem-solving

methods led them to Question the technology and to

want to make new developments in the existing

challenges - or games, as they called it.

4.2 Flappy Bird Challenge

We defined as a final objective for the course a

Lesson in which, students had to build a Flappy game.

CSEDU 2020 - 12th International Conference on Computer Supported Education

162

Figure 9: Flappy Bird Lesson.

Overall, the results were very positive, with a 63.9%

completion rate of the challenge. Starting with the

scenario represented in Figure 9, students created

their personalized Flappy Bird, as shown in Figure

10.

Figure 10: Student-Customized Flappy Bird Examples.

Students were encouraged to share their games with

family and friends, via email, by sending a link that

could be executed on all platforms where code.org

works. In statistical terms, all students who managed

to complete the challenge did so with optimal

solutions. However, regarding the data analysis, this

fact was not considered of much relevance, since, in

the case of a work of creation, personalization, and

sharing, almost all the solutions presented by the

students could be considered optimal.

4.3 Daily Classes

As already mentioned in the description of the study,

in the academic year of 2017/2018 the coding classes

were held in a pedagogical pair regime, in which the

ICT teacher was always assisted by the main teacher

of each class. This was because the students were

very young and it was the first time that they worked

with computers and platforms such as code.org.

However, in the academic year of 2018/2019, it was

not possible to maintain this organization. Although

the overall results of the study were excellent, given

this pedagogical change from one academic year to

the next, we found it relevant to observe the partial

results per class / academic year and then try to draw

some conclusions.

Figure 11: Comparative chart of tasks performed by

computing concept.

Looking at Figures 11 and 12 - for a better reading of

the graph, viewing from left to center, the lines

sequentially represent the classes Class 1, Class 2,

Class 5, Class 3 and Class 4 - it can be seen that,

compared to the Classes 1 and 2, the lines referring to

Classes 3, 4 and 5 are almost always closer to the

center of the chart, moving away from the optimum

results.

Figure 12: Comparative chart of tasks performed by

computational practices.

Developing Computational Thinking in Early Ages: A Review of the code.org Platform

163

Objectively, and comparing some results of task

completion rates, in the academic year of 2017/2018,

Classes 1 and 2, only 14.28% of the students did not

complete the last challenge of the programmed

course. On the other hand, in the academic year of

2018/2019, this number rose to 55.4%. Although the

students were not the same, pedagogically speaking,

this was the only variable changed throughout the

study, so it should not be underestimated the

importance of working in a pedagogical pair regime

in this type of subject given the number of students in

the classroom and their age.

4.4 Evaluation by Students

To evaluate the work developed from the students'

voices, we adapted a questionnaire based on the

previous studies by Kalelioğlu (2015), a

questionnaire made available to students in an

anonymous online form, with some free text

questions, allowing students to express their feelings

about the developed work. The questionnaire

contained the following questions:

Question 1. Did you have difficulties with the

process of solving the problems? If you answered Yes

or Some, say what were the difficulties.

Question 2. Was it possible to solve the various

problems in several different ways? Did you try to do

it?

Question 3. Do you think the code.org platform is

easy to use?

Question 4. Did you have any problems with the

platform?

Question 5. What was your favorite Lesson of

code.org course 2?

Question 6. And what Lesson did you like least in

code.org course 2?

Question 7. What do you think were the benefits

(what did you learn) of coding using code.org?

Question 8. Do you like coding?

Question 9. What would you like to learn more about

coding?

Question 10. Would you like to improve your

knowledge?

After the end of classes for this study, students

were invited to answer the online questionnaire. 76

valid responses were collected and 2 were canceled

due to repetition of the answer and inconsistent data

in the responses, such as “I had difficulties” but “ I

had no problems ”.

Analyzing the obtained answers, it was possible to

retain the following conclusions:

When asked about (1) difficulties in the problem-

solving process, 23 students reported that they had no

problem in the problem-solving process. Only 5

students considered that they had problems. 48

students reported that they had some problems with

this process.

However, when asked to refer the actual difficulties,

we observed that only 3 of the 76 students reported

real problems with the process and the platform - “I

had difficulties in executing it”; “Had to reset to drag

the blocks”; “At the beginning, I had some difficulties

dragging the blocks”. All the remaining responses

referred to problems that are in no way related to the

resolution process, but rather to some issues of

previous knowledge, namely, notions of laterality

[“turning to the left”, “I had difficulties in those

problems that asked to turn right or turn left"],

mathematical questions ["knowing the angles”, “I had

difficulty finding the angles and number of pixels

needed”, “was calculating the pixels and degrees."],

or even problems interpreting the problems

["Sometimes it was more difficult to pass the level

because I didn't quite understand how it was supposed

to be done", "I had some difficulties in understanding

some exercises"]. Interpretation issues could be

related to the previously mentioned fact concerning

some of the incomplete or partially wrong translations

to the local language.

Regarding the fact that (2) it is possible to solve the

proposed problems in several different ways,

86.8% of the students reported that they tried to do it,

but only 84.2% of those reported that they managed

to do it in several different ways. This fact was also

observed in the analysis of the results provided by the

platform, where 6.9% of the problems were solved

with non-optimal solutions, compared to the total

number of exercises solved. This fact shows that

despite the students realized that it would be possible

to solve the problems in another way, as they

appeared signaled in a different color, not all students

were able to do so, to obtain the optimal solution.

Regarding the (3) ease of use of the code.org

platform, 78.9% of the students who answered the

survey found the platform to be easy to use

[“Code.org is very easy to use and I had no problem

with the platform. It is very easy, it is practical

because it is suitable for all electronic devices and to

enter it you just have to enter the password and enter,

and you have a solution if you forget your password”,

“Yes I think it is easy to use code.org platform”].

Interestingly, some of the students who considered

that the platform is not easy to use, consider that they

had no problems in its use because when asked if they

CSEDU 2020 - 12th International Conference on Computer Supported Education

164

had had (4) problems with the platform, 81.6% of

the students reported that they did not have problems.

When referring to their (5) favorite Lesson of the

course, the students elected first place, with 16 votes,

Lesson 3, MAZE: SEQUENCE, the first they worked

on, tied with Lesson 16, FLAPPY, the last considered

for this study. Lessons 7: ARTIST LOOPS, 8: BEE

LOOPS and 13: BEE: CONDITIONALS, were also

the ones that most pleased the students, with 8, 10 and

7 votes respectively. Although many students chose

as favorite the Lessons that focused on simpler

concepts, 63.2% of the students who answered the

survey chose as a favorite Lesson one of the ones that

already involved more complex computing concepts

such as Loops, Events, Parallelism, Conditionals,

Operators, and Data. This result is in line with the

general results of the study, and it demonstrates that

students appreciated the complexity brought to the

problems by the application of computational

concepts with a higher level of difficulty. They

enjoyed using their recently developed computational

thinking.

On the other hand, when they referred to the (6)

Lesson that they liked least, students who had

preferred lessons with more elaborated concepts

tended to like less the first lessons of the course with

only basic concepts. The opposite was also noted,

with students who preferred Lesson 3 saying they did

not like the more complex lessons.

But it was when they were asked about their (7)

learning through the code.org platform that

students gave the most diverse answers. After

analyzing the content and after a first floating reading

(BARDIN, 1979), the main categories that emerged

were included in Table 4.

Analyzing the collected data, it was possible to verify

that, in general, students learned “that one should

never give up and that we must have patience”, in a

clear reference to the development of autonomy and

resilience inherent in solving problems. Also, the fact

that they learned "to think better", "to solve problems

in another way" and alone, "to do things in different

ways and with creativity" shows the development of

problem-solving skills.

In terms of acquired knowledge, students generally

refer to the fact that “learning to work better on the

computer and understanding the computer codes”,

although they also use the term “Programming” a

significant number of times. It is also important to

mention the fact that some students consider that they

had learned basic notions of laterality and

Mathematics

while solving exercises that involved

Table 4: Excerpt from the Free Text Responses.

Category Evidence (examples) Frequency

Programming “I learned to program more or

less well.”; I learned "how to

program games."; “The

benefits were learning to

program.”; “I learned to

program very well.”; “I

learned how to create some

things”

Autonomy “I learned to be more patient

and not ask for help right

away.”; “I learned to solve

problems on my own”; “I

with code.org learned that

we can never give up.”; “I

learned to work alone.”; “I

learned that one should never

give up and that we have to be

patient”

Learn to work

with

computers

“Learn how to work with

computers”; “I learned to

work better on the computer”

Learn Math “I was also able to look at a

certain angle and identify it”;

“I learned a lot of math.”; “I

learned to use angles”

Play “I learned to play code.org

games.”; “play and at the

same time learn.”; “I learned

to play games on the site.”; I

learned “to work, to play and

to have fun.”

Think faster

and better to

solve

problems

“At code.org I was able to:

(...) think faster to execute

problems”; “I learned to

solve problems”; “By coding

at code.org: I learned to think

better (...).”; “I learned to use

the brain more easily.” ; “I

was amazed to do things in

different ways and with

creativity”

Notions of

Laterality

“I learned right and left” 4

mathematical concepts such as angles and pixel

counts since it was important not only to learn to

program but also to learn others contents while

programming.Also important is that the students are

learning while having fun since, as they say, it was

'play and learn at the same time'.

To the question (8) Do you like to program?, we

obtained 66 positive responses (88.2%) confirming

that students were enjoying programming, at least in

the visual form they have known so far. We also

obtained 8 responses from students (10.5%) who said

they like to program, but only certain types of

Developing Computational Thinking in Early Ages: A Review of the code.org Platform

165

problems. As for negative responses, we obtained

only one.

As for (9) improving/deepening knowledge,

students tended to answer that they would like to learn

more about two different topics: building and coding

robots and coding games. Content analysis was also

performed to this free-text answers (Bardin, 1979)

and the main categories found were noted and

replicated in Table 5.

Table 5: Excerpt of categories for the content analysis of

the answers on deepening knowledge.

Category Evidence (examples) Frequency

Coding

Robots

“In terms of coding I liked

to control robots and many

more things about coding

and would like to improve

my knowledge.”; “I would

like to know if the robots

are programmed the same

way we program the

code.org characters and

yes I would like to improve

my knowledge.”; "I would

like to assemble robots,

invent a game to work in

tablets, use computers, etc.

“

Coding

Games

“I would like to learn

everything that I have to

learn and I would like to

improve my knowledge. ”;

“I would like to make other

games”; “Program games

like this.”; “Make games.

Yes, I would like to know

more. ”; “I would like to

invent my games”; “I

would like to know how to

program my own game”

Regarding (10) learning more/improving

knowledge, we obtained 97.4% of affirmative

answers, with phrases similar to “yes I would like to

know more”, “I would love to learn more” or “Yes, I

would like to”. As for negative responses, we

obtained only one - from the same student who said

he does not like to program - and another student

stated that he would like to learn “more or less”.

5 CONCLUSIONS

We call Computational Thinking to the ability to

formulate a problem and find a solution, whether

executed by a computer or not (Cuny, Snyder &

Wing, 2010). To assess the development of

computational thinking, Brennan and Resnick (2012)

created a framework of references that identifies

concepts, practices and computational perspectives.

This study used the lessons provided by the code.org

platform to analyze the development of

computational thinking at an early age. As all the

tasks proposed to the students included the search for

the solution of a problem, concepts, practices and

computational perspectives it is possible to say that

computational thinking was promoted. Students

achieved very positive results, at the same time they

trained problem-solving skills, building and retaining

knowledge better (Jonassen, 2011).

The main conclusion from this pedagogical

experience is that the code.org platform is a valid

option to develop computational thinking at an

early age and a good way for students to start

solving real-life problems by stimulating the

capacity of abstraction using simulated and

experienced practice. We believe that this

pedagogical experience will provide these children

with essential skills for increasingly complex life in

the 21st century, which includes creativity and

innovation, critical thinking and problem solving,

communication and collaboration (Partnership for

21st Century Skills, 2009).

Code.org is designed very carefully, and is

suitable for children, making it a very useful tool for

the introductory coding classes. The use of some of

the gamification strategies, such as narratives,

trophies and instant feedback, works as an

involvement factor for students. Also, the fact of

having clues for solving problems, the possibility of

partially solving the exercises and later being able to

return to complete them at 100%, makes code.org a

flexible and appropriate tool for the age group under

study. It was also possible to observe that students felt

involved in recognizing some of the characters used

in the challenges and that they were learning

computer science concepts while having fun.

There are, however, other conclusions to be taken

from the experience, namely concerning the

importance of the organization of this type of class in

a pedagogical pair regime or in smaller groups of

students, since the proposed tasks require very close

monitoring. which becomes extremely difficult to

achieve by a single teacher.

CSEDU 2020 - 12th International Conference on Computer Supported Education

166

Regarding its limitations, this study results are

limited by the number of students involved and also

by the size of the classes, which may have affected

the statistical results. Additionally, the study is also

limited by its duration, since, taking advantage of

more time, it would have been possible to transmit the

students a deeper knowledge about the fundamentals

of coding and to evolve to more complex scenarios.

5.1 Final Reflection

Every student in every school should have the

opportunity to learn how to program. The existence

of these initiatives can help to bridge the technical gap

of human resources in the IT area, developing

children's skills and potential earlier, through

carefully designed scenarios that allow them to

understand the principles of operation of computers

and their software. Rather than trying to teach a

specific coding language, the primary purpose of

early-age coding classes is to provide students with

problem-solving skills. In this learning process,

children learn many other things. They are not simply

learning to program, they are coding to learn

(Resnick, 2013).

Such experiences demonstrate the importance that

the use of this type of platform has for children. The

early introduction of coding activities in the

curriculum is essential as it contributes to the

development of children by turning them into

producers and not simply consumers of content and

technology.

ACKNOWLEDGMENTS

This work was partially financed by the Portuguese

funding agency, FCT - Fundação para a Ciência e a

Tecnologia, through national funds, and co-funded by

the FEDER, where applicable.

This work was partially funded by CIEd –

Research Centre on Education, project

UID/CED/01661/2019, Institute of Education,

University of Minho, through national funds of

FCT/MCTES-PT.

We would like to thank the Colégio Paulo VI

(Gondomar, Portugal) the authorization to carry out

this study on its premises, and students of the 4th

grade of the academic years of 2017/2018 and

2018/2019 by their collaboration.

REFERENCES

Bardin, L. (1979). Análise de conteúdo. Lisboa: Edições 70.

Barradas, R., & Lencastre, J. A. (2017). Gamification e

Game-Based Learning : Estratégias eficazes para

promover a competitividade positiva nos processos de

ensino e de aprendizagem. In Revista Investigar em

Educação (Issue Mundo digital e Educação, pp. 11–37).

Sociedade Portuguesa de Ciências da Educação.

Barradas, R.; Lencastre, J. A.; Soares, S.; Valente, A.

(2019). Usability Evaluation of an Educational Robot

for STEM Areas. 11th International Conference on

Computer Supported Education. Heraklion: INSTICC

and University of Crete.

Bell, T., Witten, I. H., & Fellows, M. (2015). CS

Unplugged. University of Canterbury, NZ.

http://csunplugged.org/wp-content/uploads/2015/03/

CSUnplugged_OS_2015_v3.1.pdf

Brennan, K., Chung, M., & Hawson, J. (2011). Scratch

Curriculum Guide Draft. Nature, 341(6241), 73.

Brennan, K., & Resnick, M. (2012). New frameworks for

studying and assessing the development of

computational thinking. Annual American Educational

Research Association Meeting, Vancouver, BC,

Canada, 1–25. https://doi.org/10.1.1.296.6602

Code.org (2019). About | code.org. Retrieved April 2019,

from https://code.org/about

Coelho, A.; Almeida, C.; Almeida, C.; Ledesma, F.;

Botelho, L.; Abrantes, P. (2016). Iniciação à

Programação no 1.o Ciclo do Ensino Básico, Linhas

Orientadoras para a Robótica. Lisboa: DGE. Retrieved

September 17, 2017, from http://www.erte.dge.mec.pt/

sites/default/files/linhas_orientadoras_para_a_robotica

.pdf

Costermans, J. (2001). As actividades cognitivas -

raciocínio, decisão e resolução de problemas. Coimbra:

Quarteto Editora.

CS Education Research Group (2016). Computer Science

Unplugged. Retrieved September, 2015, from

http://csunplugged.org

Cuny, J.; Snyder, L.; Wing, J.M. (2010). Demystifying

computational thinking for non-computer scientists.

Retrieved from http://www.cs.cmu.edu/~CompThink/

resources/TheLinkWing.pdf

Echeverría, M.; Pozo, J. (1998). Aprender a resolver

problemas e resolver problemas para aprender. In J.

Pozo, A solução de problemas: aprender a resolver,

resolver para aprender. Porto Alegre: Artmed.

Eguchi, A. (2014). Robotics as a Learning Tool for

Educational Transformation. International Workshop

Teaching Robotics, Teaching with Robotics &

International Conference Robotics in Education, 27–34.

https://doi.org/10.4018/978-1-4666-8363-1.ch002

Figueiredo, M., Torres, J. (2015). Iniciação à Programação

no 1.o Ciclo do Ensino Básico Linhas Orientadoras.

Retrieved October 1, 2016, from http://www.erte.dge.

mec.pt/sites/default/files/Projetos/Programacao/IP1CE

B/linhas_orientadoras.pdf

Jonassen, D. (2004). Learning to solve problems - an

instructional design guide. São Francisco: Pfeiffer

Developing Computational Thinking in Early Ages: A Review of the code.org Platform

167

Jonassen, D. (2011). Learning to solve problems. A

Handbook for Designing Problem-Solving Learning

Environments. New York: Routledge

Brennan, K., Chung, M., & Hawson, J. (2011). Scratch

Curriculum Guide Draft. Nature, 341(6241), 73.

National Research Council. (2010). Report of a Workshop

on The Scope and Nature of Computational Thinking.

Washington: The Nacional Academies Press.

Nunes, D. (2011). Ciência da Computação na Educação

Básica. Jornal da Ciência. Retrieved from http://

www.jornaldaciencia.org.br/Detalhe.php?id=79207

Papert, S. (1993). The Children Machine. New York:

BasicBooks

Partnership For 21ST Century Skills. (2009). Framework

for 21st Century Learning. Retrieved from http://

www.p21.org/storage/documents/docs/P21_framewor

k_0816.pdf

Pedro, A.; Matos, J. F.; Piedade, J.; Dorotea, N. (2017).

Probótica: Linhas Orientadoras. Lisboa: DGE. Re-

trieved from http://www.erte.dge.mec.pt/sites/ default/

files/probotica_-_linhas_orientadoras_2017.pdf

Pólya, G. (2003). Como resolver problemas. Lisboa:

Gradiva

Kalelioğlu, F. (2015). A new way of teaching programming

skills to K-12 students: Code.org. Computers in Human

Behavior, 52, 200-210

Resnick, M., Maloney, J., Monroy-Hernández, A., Rusk,

N., Eastmond, E., Brennan, K., Millner, A.,

Rosenbaum, E., Silver, J. a Y., Silverman, B., & Kafai,

Y. (2009). Scratch: Programming for All.

Communications of the ACM, 52, 60–67.

https://doi.org/10.1145/1592761.1592779

Resnick, M. (2012). Point of View: Reviving Papert’s

Dream. Educational Technology, 52(4), 42–46.

Resnick, M. (2013). Learn to Code, Code to Learn.

Retrieved February 7, 2019, from https://www.

edsurge.com/news/2013-05-08-learn-to-code-code-to-

learn

School, C. A. (2012). A curriculum framework for

Computer Science and Information Technology.

Retrieved from http://www.computingatschool.org.uk/

data/uploads/Curriculum%20Framework%20for%20C

S%20and%20IT.pdf

Stahl, G.; Koschmann, T.; Suthers, D. (2006). Computer-

supported collaborative learning: An historical

perspective. In R. K. Sawyer, Cambridge handbook of

the learning sciences (pp. 409-426). Cambridge, UK:

Cambridge University Press.

Wing, J. M. (2006). Computational Thinking.

Communications of the ACM, 49, pp. 33-35.

Wing, J. M. (2007). Computational Thinking. Retrieved on

May 1, 2019, from https://www.cs.cmu.edu/afs/cs/usr/

wing/www/Computational_Thinking.pdf

Wing, J. (2017). Computational Thinking’s Influence on

Research and Education for All. Italian Journal of

Educational Technology, 25(2), 1–12. https://doi.org/

10.17471/2499-4324/922

Puiu, T., (2019). Your smartphone is millions of times more

powerful than all of NASA’s combined computing in

1969, Retrieved on February 21, 2019, from https://

www.zmescience.com/research/technology/smartphon

e-power-compared-to-apollo-432/

Kapp, K. M., Blair, L., & Mesch, R. (2012). The

Gamification of Learning and Instruction Fieldbook.

San Francisco, CA: Wiley

CSEDU 2020 - 12th International Conference on Computer Supported Education

168