Instructional Application for Programming and

Algorithmic Self-Learning

A Didactic Approach with Mobile Robotics as Pedagogical Context

Pedro G. Feijóo-García

and Fernando De la Rosa

Systems Engineering Program, Universidad El Bosque, Bogotá D.C., Colombia

Systems and Computing Engineering Department, Universidad de los Andes, Bogotá D.C., Colombia

Keywords: Programming Learning, E-Learning, Programming Teaching, E-Robotics, Instructional Technology, Visual

Programming Language.

Abstract: One of the main challenges related to algorithmic and programming teaching with novice students, is to

focus their process on acquiring concepts and developing problem solving skills in programming, without

spending time overcoming syntax-oriented learning curves of specific languages. The application here

explained is proposed as an instructional technology that, using the advantages of Visual Blocks

Programming, through virtual and remote mobile robotics’ scenarios, seeks to give playful and friendly

mechanisms for programming and algorithmic self-learning. This paper presents the pedagogical design and

approach of the tool, evaluated through a User Experience approach with high school students in the

Colombian educational context.

1 INTRODUCTION

In the last decade, there has been of projects focused

on designing educational tools to teach and enhance

programming and algorithmic skills. Several

solutions propose games’ design, narratives’

development and other types of approaches, offering

the student the possibility to learn to program using

powerful visual and interactive mechanisms.

However, even if these solutions offer novel

pedagogical approaches, they are usually limited

because they need the presence of a trained tutor to

articulate them in the student’s process, most of

which are applications that even depend on special

technical configurations. Many of these solutions are

complements to a programming course, which

means that the student lacks an environment that

favors self-learning scenarios.

Although there has been global advance in the

design and development of new instructional

technologies emphasized in programming and

algorithmic teaching, within the Colombian

educational context, these technologies’ promotion

and elaboration have not been encouraged as they

should. Programming and algorithmic skills are

unjustifiably delegated to higher education in the

national educational scheme. This situation

jeopardizes students’ motivation towards studying

professional careers related to Information

Technology; careers that nowadays are highly

demanded within the local and international

industry.

Through this article, we present the conception,

design, implementation, and validation of RoBlock;

a Web application proposed to offer self-learning

environments for algorithmic and programming

concepts for high school students. This, to answer

the following research question:

Is it possible to design an instructional

application, contextualized in mobile robotics and in

a self-learning visual programming approach, to

motivate high school students to learn to program

autonomously?

This paper has the following organization:

section 2 presents the theoretical framework that

supports this study; followed by section 3 which

presents related works and technologies. Later,

RoBlock’s modules’ design is presented in section 4,

justifying it with the pedagogical strategy described

in section 5. Section 6 describes the main results

obtained; and finally, the conclusions and future

work are exposed in section 7.

230

Feijóo-García, P. and De la Rosa, F.

Instructional Application for Programming and Algorithmic Self-Learning.

DOI: 10.5220/0006690002300237

In Proceedings of the 10th International Conference on Computer Supported Education (CSEDU 2018), pages 230-237

ISBN: 978-989-758-291-2

2 THEORETICAL FRAMEWORK

2.1 Behaviorism and Constructivism

Behaviorism is a trend that was quite popular in the

early twentieth century, with a peak of popularity in

1958 when Skinner and Holland included the

principles of the reinforcement theory in their first

programmed learning course (Galvis, 1992).

This theory proposes that teachers may try to

bring students from a first state (a starting

knowledge base), to a higher one, where there is a

knowhow or skill they should acquire. Different

paths can achieve this, but one of these, is through

the programming of instructions along the teaching

process. By mentioning "programming of

instructions", we refer on how the student's learning

process is carried out through a series of pedagogical

stages, designed to guide the student from a

reference knowledge base, towards a level in which

there is a new knowledge she/he does not have at

that specific moment.

What makes Behaviorism a programming

approach, is the treatment given to the student to

guide her/him from the starting point where she/he

is, to the higher knowledge stage desired. The basic

behavioral principles are listed below:

 An individual manages to learn by observing

the consequences of his/her actions.

 Reinforcements are those consequences that

strengthen the possibility of repeating an

action.

 The more reinforcement the student receives,

the more likely she/he is to perform the

desired action successfully.

 The absence or delay of the reinforcement

after an action, limits the chances to repeat it.

 Differential reinforcement can gradually shape

a student’s learning behavior.

Unlike Behaviorism, it is fundamental in

Constructivism to understand the subject that learns.

This involves the understanding of his/her vital field,

as well as the interaction between his/her

environment (the influence context) and what the

subject has been, is, and desires to be. When

referring to the vital field of the student, the theory

refers to the student's understanding of his/her

environment, formed by his/her past, present and

future (Galvis, 1992).

Within this theory’s approach, the learning

process is considered born with the creation of

knowledge through interaction with environments

that give or allow the exploitation of curiosity,

experiential experiences and the enhancement of the

student's imagination. In other words, it focuses on

providing the learner with the ability to learn while,

at the same time, he/she constructs his or her own

mental models.

Based on this approach, Constructivism declares

that the learning obtained by a student, is not the

result of a predefined structure of operations

(operation is an internalized action that changes the

object of knowledge). According to this, the

structures construct based on the learning process

that the student carries, considering that the student

does not evolve in his knowledge through the

association of knowledge, but through the

assimilation of new mental models.

2.2 Mobile Robotics as Pedagogical

Context

Mobile Robotics is a branch of knowledge that is

oriented in the study associated with machines that

can move on land, through the air or in the water;

spatially in two or three dimensions (Matarić, 2007).

For this study, it is interesting to use this

pedagogical context for programming and

algorithmic teaching, because it incorporates

concepts that involve the automation of activities

and decision making, whilst considering the

interaction of a Robot (physical device or machine)

with its environment using sensors of different types

(Feijóo & De la Rosa, 2016).

An interesting characteristic of Mobile Robotics,

is the possibility to build a virtual programming

environment, consistent with a respective physical

environment. This, considering that both scenarios

are, as far as possible, equivalent, if and only if, both

respect the behavior of elements considered and

keep the similarity of the Robot’s positioning in a

two or three dimensions’ environment.

A virtual programming tool allows multiple

students to interact and learn simultaneously.

Furthermore, the proposal of a virtual environment,

developed considering the operation of a real robot,

offers the possibility to consider problems and

solutions that could be raised equally in an

experimental environment with a physical robot.

3 STATE OF THE ART

3.1 Academic Background

Aggarwal et al. (2017) conducted a research to

determine the effectiveness and usefulness of the

Instructional Application for Programming and Algorithmic Self-Learning

231

Microsoft’s Kodu Game Lab to support students’

learning for programming. For their study, they

worked for two 90-minutes sessions, with two

groups of elementary students, obtaining and

comparing results from the usage of Kodu’s tiles and

flashcards (first group), and from the usage of paper

sheets with color prints of the design patterns

considered for the learning process (second group).

Both groups were composed by students without

previous programming knowledge, assessing them

before and after the study. The results of this

research present the benefits and the drawbacks of

using physical manipulatives in different scenarios.

On one hand, the students who did not interact with

the tiles showed to had acquired a better

understanding of the rules’ execution from their

interaction with the programming environment. On

the other hand, students who interacted with the

tiles, demonstrated a better understanding about

rules’ syntax and construction.

Grover and Basu (2017) present the design and

development of assessments items to measure

students understanding in introductory CS courses,

by answering the following research question: “How

can learning outcomes for computing constructs

such as variables, expressions (arithmetic and

logical), and loops, be organized into a structured

assessment framework and measured with technical

quality?”. They worked with 100 students from

different middle school courses, applying an

assessment framework that was developed following

the Evidence-Centered Design (ECD). From their

study, they conclude that, even though Visual

Blocks Programming (VBP) simplifies

programming syntax, there are still conceptual

difficulties towards using and understanding key

structures in programs, such as variables,

conditionals and loops. Finally, they invite to put

additional effort on pedagogical strategies,

especially on formative and summative assessments,

to measure students’ understanding and

misconceptions, looking forward to refining

pedagogy and curricula.

Weintrop and Holbert (2017) present findings

from a study in which they test how learners use a

dual-modality environment, having the possibility to

choose to work either a Visual Blocks Programming

(VBP) approach, or a text-based approach. Pencil

Code, the proposed environment, is used to

understand which is the modality preferred by

learners, and why they move from one modality to

another in a same project. From this study, the

authors conclude that the dual-modality approach is

effective for programming learning, considering that

all the participant students completed successfully

the programming assessments. Furthermore, they

indicate that blocks are useful to introduce a new

programming environment, as well as support items

for conceptual comprehension.

Paramasivam et al. (2017) performed a research

applying end-user programming tools for functional

robots in Computer Science education, presenting

results according to a week-long introductory

workshop, in which eleven students with different

disabilities programed a Clearpath Turtlebot. For

their workshop, they used an end-user programming

tool (EUP) named CodeIt, using a text-based

interface rather than a visual blocks programming

approach, to increase accessibility for students with

less motor skills and visual disabilities. Their

findings report that EUP tools can be used to create

advanced robotics platforms, accessible and useful

for novice programming students. Furthermore, they

indicate that the pairing of robotics with EUP tools,

enhance students’ confidence and interest towards

programming and Computer Science topics.

Gucwa and Cheng (2017) present a methodology

to create challenge problems, using simulation

environments for hardware robot-based

programming competitions. For their proposal, they

center on the RoboPlay Challenge Competition,

which involves Linkbots and RoboSim as hardware

and simulation technologies. The authors argue that

this competitive context offers a unique opportunity

for students to apply learned skills. Furthermore,

they conclude that tools like RoboSim, are useful to

let students and teachers to work with robots,

without the need of setup of physical hardware.

They finally find that the students’ response to the

competition context with RoboSim is positive,

mainly because of the opportunity this tool gives

towards rapid code improvement and validation,

arguing that virtual scenarios let students gain

effective and useful feedback.

3.2 Similar Technologies and

Languages

The technologies and languages presented here are

previous solutions that have characteristics like

those of RoBlock:

 Scratch: Tool designed for people with no

notions of programming, for the design and

elaboration of 2D video games and animations

(MIT Media Lab & Lifelong Kindergarten

Group, 2006). This, using Visual Blocks

Programming as a playful mechanism of

interaction for learning. Scratch is Web, free

CSEDU 2018 - 10th International Conference on Computer Supported Education

232

and offers users’ profiling through

authentication mechanisms.

 RoboBlockly: This is a web based tool that

uses Visual Blocks Programming in a robot-

approach context. With this tool, students

program a robot in a simulated environment,

enhancing skills towards computing, science,

technology, engineering and mathematics (C-

STEM). The visual blocks can be downloaded

as C++\C code with the Ch interpreter, letting

students use their code with Lego Mindstorm

and Linkbot robots (Frankie et al., 2017).

 RoboSim: This is a standalone robot

simulation environment that allows students to

program virtual Linkbot and Lego Mindstorm

robots, working with the C++\C interpreter

Ch. The programs that students create with

this environment can be used to control

physical robots (Gucwa and Cheng, 2017).

 Pencil Code: This is a tool that uses a Visual

Blocks Programming approach to teach

programming through art, music and stories

elaboration. Its main characteristic is that,

differently from other approaches, this tool

offers a collaborative scenario for learning.

Additionally, the tool offers the possibility to

switch from a visual approach to a textual one,

which enhances programming skills from

algorithms design to syntax understanding

(Weintrop and Holbert, 2017).

4 MODULES’ DESIGN

Following the application’s final design reported by

Feijóo and De la Rosa, 2016, RoBlock articulates a

methodology encompassed between Behaviorism

and Constructivism, by guiding the student through

modules dedicated to basic concepts of

programming and algorithmics, that are applied in

tasks that must be solved by a robot in environments

with free cells for displacement, marks to be

discovered, obstacles to avoid, colors to identify,

among other elements included. In total, the tool has

five virtual modules and a remote module that

allows the student to interact with real scenarios

available in a remote laboratory (module that is still

in a prototype state). The modules appear in an

incremental way, requiring from the student to

achieve each module before moving on to the

following one.

To carry out the incorporation of the pedagogical

strategies previously described through this article,

each of the five virtual modules corresponding to

RoBlock, offers a different interactive approach.

Considering that the pedagogical context that is

intervened with this project is the Colombian one,

the designed application is in Spanish, being it the

native language of the Colombian population.

4.1 Virtual Modules

RoBlock offers a total of five virtual modules,

exposing tasks and scenarios for the student’s

learning process towards programming and

algorithmic concepts and skills. The modules

included respond to the following big topics:

variables, sensors, conditionals, loops, and

functions.

Mobile robotics offers a pedagogical context apt

to teaching every one of the topics of interest

because of the successful convergence with the

needs and operations of a mobile robot.

4.1.1 Module for Variables

For this proposed module, the student is asked to

solve exercises in which the robot must find a series

of marks that are distributed along a scene, within a

determined time frame (Figure 1).

Figure 1: Module for Variables (Content in Spanish).

Figure 2: Module for Variables – Blocks’ Menu

(Content in Spanish).

Instructional Application for Programming and Algorithmic Self-Learning

233

For this purpose, the tool gives the student a

series of Visual Blocks designed to manage the

robot’s position variables (translation and rotation),

in addition to an editor that helps with the

integration of these and the control of their

respective execution (Figure 2).

In this module, the student does not require yet to

reach all the marks in the scene in a single code

execution. Each achieved mark is saved as

"reached", letting the student to experience, use and

play with several of the blocks arranged for this

module.

4.1.2 Module for Sensors and Module for

Conditionals

Throughout these modules, the student is asked to

solve exercises in which marks are hidden from the

scene, so the student requires to use sensors and

conditionals to find them in the time determined by

the problem (Figure 3).

Both modules differentiate in the type of visual

blocks available to the student. In the case of the

module for sensors, self-contained blocks for the

sensors management are provided, by which the

student does not need to declare or define any

condition. On the other hand, for the module for

conditionals, each sensor is summarized in a boolean

block, asking the student to define the evaluation of

conditions with additional conditions’ blocks.

Figure 3: Module for Conditionals (Content in Spanish).

The learning process of the student is

incremental, so that, having already passed the

module of variables, the tool gives the student a

series of blocks to handle position variables of the

robot (Figure 4), in complement with a series of new

blocks that allow the student to manage sensors

(module for sensors) and conditions (module for

conditionals).

As in the module for variables, in this module the

student does not require yet to find all the marks in

the scene with a single code execution. Each mark

that found is visible in the scene and saved as

"reached", making it easier for the student to

experiment with the use of several of the blocks

arranged for this level.

Figure 4: Module for Conditionals – Blocks’ Menu

(Content in Spanish).

4.1.3 Module for Loops

Through this module, the student works with

exercises in which the robot must move through

routes and proposed alleys (Figure 5).

Figure 5: Module for Loops (Content in Spanish).

As in the modules already described, the tool

grants the student a series of visual blocks proposed

for the management of position variables of the

robot, in complementing with a series of elements

that allow the student to handle sensors and

conditions. Furthermore, and from this module, the

tool provides blocks that let the student model and

use control instructions for loops and paths’

following (Figure 6).

Unlike the previous modules, in this module the

student requires solving the task through the scene

(or maze) in a single code execution. This is

achieved because each time the position of the robot

restarts, those marks already reached remain

unmarked. The premise is that the exercise must be

solved, emphasizing the need to declare and use

loops.

CSEDU 2018 - 10th International Conference on Computer Supported Education

234

Figure 6: Module for Loops – Blocks’ Menu (Content in

Spanish).

4.1.4 Module for Functions

Throughout this last virtual module, the student

works with exercises in which all the types of

scenarios provided in previous modules appear. The

only difference is that, for this module, the student

must use functions (blocks enabled from this level)

that the tool validates from its declaration to its

invocation.

This module provides blocks for the declaration

and use of functions (Figure 7), in addition to those

blocks already grouped and offered to the student in

the previous modules (Figure 8).

Figure 7: Module for Functions – Declaration and Blocks’

Usage (Content in Spanish).

Figure 8: Module for Functions – Blocks’ Menu.

In the same way, as in the previous module, in

this module the student requires navigating the

raised scene with a single code execution. This

occurs because each time the robot’s position

restarts, the marks already reached remain

unmarked. The latter, with the premise that the

exercise is solved, emphasizing the need to declare

and use functions.

5 PEDAGOGICAL APPROACH

The designed tool proposes to offer the student self-

learning environments for programming and

algorithmic concepts and skills. This is because,

according to the directors of the participant

educational institutions in this project, and

considering the national educational scheme, there is

not enough curricular time to cover programming

and algorithmics topics, without sacrificing time for

the strengthening of core areas required in

Colombia: mathematics, natural sciences, history,

and Spanish.

Pedagogically, RoBlock makes use of both

Behaviorism and Constructivism; inviting the

student to design and construct her/his own answers

to tasks proposed by the tool, while raising levels of

knowledge that lead the student to the appropriation

of concepts of different complexities. These

problems are based on the principle that their

solution is unique. In this way, the students obtain

stimuli against their answers, based on the personal

development of its algorithmic solutions.

Considering that different students may not learn

programming equally, it is fundamental that the first

levels allow free will, but at the same time, focus on

the functionality of the algorithm or solution towards

the understanding and management of basic

concepts. This implies that the student is implicitly

guided to find the solution of the tasks proposed, by

offering only, and in a strategic way, those elements

needed to solve the proposed tasks.

6 ROBLOCK’S EVALUATION

To test RoBlock as a learning technology, 46 school

students from three institutions (one public and two

private) actively participated in a comparative

pedagogical evaluation environment using Scratch

as a reference technology, generating interesting and

significant results in terms of the programming and

algorithmics learning process. Among the target

Instructional Application for Programming and Algorithmic Self-Learning

235

population, 27 students evaluated RoBlock,

providing feedback, given the functionalities offered

by the software. The results here presented, which

correspond to the qualitative assessment conducted

within the study, complement the quantitative results

previously reported by Feijóo and De la Rosa

(2016).

6.1 UX: Students’ Perception

The study group interacted with RoBlock in a period

equivalent to five hours of experimentation.

Throughout the established period, they solved

exercises for each of the five virtual modules

offered, working approximately one hour per

module. At the end of their experience, the students

made a qualitative evaluation, providing their

opinions towards the tool and its purpose.

The results here presented show that the tool was

well received by the target population and that, for

five hours of interaction with it, a large minority

(close to 50%) was partially or totally in agreement

that it is possible to learn programming and

algorithmics only with RoBlock. This is of interest,

especially given the characteristics of the study

group, previously indicated in this document.

The first question was raised to know if the

group considered RoBlock an interesting tool. Most

students (96.3%) who worked with RoBlock

considered the tool interesting or very interesting

(Figure 9). This is favorable for the study,

considering that these students are the target

population of the project, and that those who

considered the tool not interesting, do not exceed 4%

of the population surveyed.

Figure 9: Answer: Do you consider RoBlock an interesting

tool?

Complementarily, most students (96%)

considered RoBlock as a good tool for teaching

programming and algorithmics in a playful way

(Figure 10).

Figure 10: Answer: Do you consider RoBlock useful for

Programming and Algorithmics teaching?

Finally, a higher proportion of students (52%)

indicated that they did not consider RoBlock enough

to learn to program in a self-taught way (Figure 11).

Yet, 48% of the students who worked with the tool

are partially or totally in agreement that RoBlock

serves to learn autonomously, which is considered a

favorable response for RoBlock in this study.

Figure 11: Answer: Do you believe RoBlock is appropriate

for self-learning?

7 CONCLUSIONS AND FUTURE

WORK

From this study, we conclude that the first version of

RoBlock was a success, and that this is a friendly,

interesting, and pleasant tool for the target

population to which it is directed. This, based on the

results obtained from the tests of User Experience

and the acceptance presented by the students

towards the tool.

In addition, we conclude that the research

question of interest to this project is satisfactorily

answered. With RoBlock, it is evident that it is

possible to design a technological tool to motivate

high school students to learn independently

CSEDU 2018 - 10th International Conference on Computer Supported Education

236

programming and algorithmics. This, considering

the results based on the second and third question

(Figures 10 and 11).

From this pilot study, there is a broad horizon for

future work involving RoBlock. In this study we

evaluated only the virtual scheme that RoBlock

offers through its first five modules, so it is worth

evaluating the use of physical and remote scenarios

with students, through the sixth module of the tool.

A future study will include the latest module offered

by RoBlock, and will evaluate the remote interaction

of students with physical robots in preconfigured

scenarios.

Also, based on the evolution of mobile

technologies, RoBlock could evolve towards to

operate with mobile devices, such as tablets and

smartphones. Likewise, it would be interesting to

carry out an impact study that indicates which

technological approach is most striking to the target

population, evaluating not only the appreciation for

the tool, but also the level of learning obtained by

the users.

ACKNOWLEDGEMENTS

We sincerely thank the educational institutions that

helped us in this first version of the project: Colegio

Provinma–Providencia Inmaculada (Bogotá,

Colombia), Centro Educativo y Cultural Reyes

Católicos (Bogotá, Colombia), and Programa

Progresa Fenicia (Universidad de los Andes, Bogotá,

Colombia). Their collaboration was essential for this

pilot study to succeed, as they efficiently

collaborated with their students and with all the

necessary computational resources.

REFERENCES

Aggarwal, A., C. Gardner-McCune, and D. S. Touretzky.

2017. Evaluating the Effect of Using Physical

Manipulatives to Foster Computational Thinking in

Elementary School. In Proceedings of the 2017 ACM

SIGCSE Technical Symposium on Computer Science

Education (SIGCSE '17). doi:

https://doi.org/10.1145/3017680.3017791

Feijóo, P. G., and F. De la Rosa. 2016. RoBlock – Web

App for Programming Learning. International Journal

of Emerging Technologies in Learning (iJET), Vol.

11, No. 12, 45-53. doi:

https://doi.org/10.3991/ijet.v11i12.6004

Frankie, T., D. Wesley, J. Gappy, and H. Cheng. 2017. C-

STEM: Engaging Students in Computing with

Robotics (Abstract Only). In Proceedings of the 2017

ACM SIGCSE Technical Symposium on Computer

Science Education (SIGCSE '17). doi:

https://doi.org/10.1145/3017680.3017832

Galvis Panqueva, A. 1992. Ingeniería de software

educativo. Bogotá: Uniandes.

Grover, S., and S. Basu. 2017. Measuring Student

Learning in Introductory Block-Based Programming:

Examining Misconceptions of Loops, Variables, and

Boolean Logic. In Proceedings of the 2017 ACM

SIGCSE Technical Symposium on Computer Science

Education (SIGCSE '17). doi:

https://doi.org/10.1145/3017680.3017723

Gucwa, K.J, and H.H. Cheng. 2017. Making Robot

Challenges with Virtual Robots. In Proceedings of the

2017 ACM SIGCSE Technical Symposium on

Computer Science Education (SIGCSE '17). doi:

https://doi.org/10.1145/3017680.3017700

Matarić, M. J. 2007. The Robotics Primer. Cambridge,

MA: MIT Press.

MIT Media Lab, and Lifelong Kindergarten Group. 2006.

Scratch. MIT. Retrieved from http://scratch.mit.edu

Paramasivam, V., J. Huang, S. Elliott, and M. Cakmak.

2017. Computer Science Outreach with End-User

Robot-Programming Tools. In Proceedings of the

2017 ACM SIGCSE Technical Symposium on

Computer Science Education (SIGCSE '17). doi:

https://doi.org/10.1145/3017680.3017796

Weintrop, D., and N. Holbert. 2017. From Blocks to Text

and Back: Programming Patterns in a Dual-Modality

Environment. Proceedings of the 2017 ACM SIGCSE

Technical Symposium on Computer Science

Education (SIGCSE '17). doi:

https://doi.org/10.1145/3017680.3017707

Instructional Application for Programming and Algorithmic Self-Learning

237