Guided Participatory Research on Parallel Computer Architectures

for K-12 Students Through a Narrative Approach

Valentina Mazzoni

, Luigina Mortari

, Federico Corni

and Davide Bertozzi

Department of Philosophy, Education and Psychology, University of Verona, Verona, Italy

Department of Education and Humanities, University of Modena and Reggio Emilia, Modena, Italy

Department of Engineering, University of Ferrara, Ferrara, Italy

Keywords: Participatory Research, K-12 Students, Narrative Approach, Problem-oriented Project Work,

Parallel Computer Architecture, Networks-on-Chip.

Abstract: The approach to computer science (CS) education is typically geared towards the knowledge of the

principles behind information technology, but there are social indicators that it overlooks some important

educative aspects such as thinking competences and social attitudes. Such aspects play a fundamental role

when bringing CS education to the K-12 level. In order to enable a truly educational experience, we propose

to bring specific CS research problems within reach of K-12 students, because the active knowledge

construction process that takes place during research requires children to be engaged with all of their

knowledge, skills and attitudes. This poses the challenge of overcoming the knowledge gap of students,

which we address by means of a synergistic cooperation of CS experts and educators. More specifically, we

propose the narrative approach as the key enabler for CS participatory research with K-12 students.

1 INTRODUCTION

Computer science (CS) and the technologies it

enables now lie at the heart of the way students live

their lives, especially in school and entertainment

environments. The ubiquity of information

technology is frequently cited to support inclusion of

CS in secondary education. The starting point for

this work is that even learning-objective-oriented

approaches to CS education (Sawyer, 2009;

Pasternak, 2012) in many cases end up accounting

only for some educational elements (e.g.,

programming skills), while leaving the remaining

ones (e.g., social skills, self-confidence, motivation,

curiosity) to other disciplines (psychology,

pedagogy, sociology). In practice, however, any CS

curriculum develops social attitudes and thinking

skills, even though they are not explicitly accounted

for in curriculum design. The unmistakable proof of

this matter of fact is given by a generalized lack of

interest in science curricula in secondary and tertiary

education. Another side effect is that thinking

competences such as creative, critical and care

thinking are not evolved to the same extent of

technical knowledge and skills.

In order to overcome this gap, we value the

research experience as a highly educative one, since

it consists of an active knowledge construction

process where the subject is engaged with all of his

knowledge, skills and attitudes. For this reason, we

aim at bringing research experiences in CS within

reach of the cognitive capabilities of K-12 students.

It is in fact at this stage of education that long-term

attitudes start shaping up.

In order to overcome the common “black-box”

approach of children to the media-rich electronic

devices that are pervasive in their lives (e.g.,

smartphones, game consoles, laptops, etc.), the

object of the proposed research experience will be

the prototype implementation of a networked

parallel computer architecture, which provides the

needed “intelligence” to the above devices.

The main challenge we face in this project is to

adjust technical depth, contextualization, and

exemplification to the audience’s stage of cognitive

development. From a pedagogical perspective, an

effective way of making complex concepts

accessible to young students is the narrative

approach, since narrative thinking reflects the basic

and powerful forms in which we gain knowledge of

the world (Egan, 1986). Therefore, stories support

the possibility to explain phenomena by creating

111

Mazzoni V., Mortari L., Corni F. and Bertozzi D..

Guided Participatory Research on Parallel Computer Architectures for K-12 Students Through a Narrative Approach.

DOI: 10.5220/0004957601110117

In Proceedings of the 6th International Conference on Computer Supported Education (CSEDU-2014), pages 111-117

ISBN: 978-989-758-022-2

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

narrative forms of models of industrially created

objects (Fuchs, 2013).

For the sake of our research framework with

children, we identify grid street plans of modern

cities as an effective metaphor of on-chip

interconnection networks, which provide the needed

communication and integration function to modern

parallel computer architectures.

For the success of the proposed approach, the

synergistic and inseparable cooperation of

pedagogists and CS researchers at all stages of the

implementation is a mandatory requirement. In fact,

such cooperation should complement depth and

breadth of knowledge about CS with pedagogical

content knowledge (Cochran et al., 1993).

2 MOTIVATION

Today, teaching science at school is one of the major

challenges for education. The Eurydice Report on

the development of key competence at school in

Europe (2012, p. 43) stresses the following factors:

- young people lack of basic skills in mathematics

and science;

- there is a declining number of higher education

graduates in MST (mathematics, science and

technologies) fields.

Since education has a pivotal role in order to

reverse this negative trend, what seems to be urgent

is the necessity to impart new vigour to science

education in order to raise student’s interest,

knowledge and skills. Our contribution is thus a

methodology for the realization of an educative

experience in science education at school.

We consider the present situation of science

education to have its roots early in the school

system, where abilities, preferences and disabilities

start shaping up. Thus, we develop educational

strategies for students at the K-12 level.

While CS is firmly established in higher

education, introducing K-12 students to CS remains

a major challenge, since it implies to address the

pedagogical issues associated with adapting the level

of technical detail to students at varying levels of

cognitive development (Knobelsdorf, 2013). Recent

learning theories such as constructivism, activity

theory, and situated or distributed cognition theory,

as well as work conducted in the interdisciplinary

field of the learning sciences, are trying to tackle the

problem, although this effort is still in the early

stage.

One relevant evidence from the application of

the constructionism theory is that learners are more

likely to be intellectually engaged when they are

working on personally meaningful activities and

projects. In this direction, designing and creating

simple digital objects (a webpage, a small program

or a simple hardware device) was shown to rise the

curiosity to acquire the foundation of factual

knowledge (Knobelsdorf, 2011). However, one

common misunderstanding is that there is no

reaction of what is studied upon the development of

the person learning, upon the tastes, interests, and

habits that control student’s future mental attitudes

and responses. In practice, these personal elements

are collaterally formed. This is an evidence that

stems from sad matter of facts: for instance, CS

education may lead to students that are largely

engaged with computer programming, but at the

expense of the development of social abilities and

skills; moreover, this might not be reflected into the

formation of attitudes that decide the uses to which

the ability is to be put on.

Another common embodiment of constructivism

consists of setting up environments for learning

programming such as Logo, Scratch or Lego

Mindstorm. Although these learning environments

are engaging, students do not automatically obtain a

clear and systematic understanding of programming

concepts (Meerbaum-Salant et al. 2010). In our

approach, the entry point into CS is not simply

working with technology, but rather a research

experience that leads students to “discover” the basic

principles of complex electronic devices of common

use.

This choice stems from the awareness that

students succeed in developing domain-specific

competencies when their learning corresponds to

authentic situations, where tasks and problems arise

not from pedagogical concerns, but rather from the

real-world (Collins, 2006). Because CS knowledge

mostly consists of abstract concepts or problem-

solving strategies, we propose an effective way of

contextualizing this. We thus present a concrete

instance (a real prototyping platform) as well as the

underlying abstract principle (on-chip networking),

so that students not only develop a general

understanding of the concept in question, but also

learn to apply it in different situations.

3 PROPOSED APPROACH

3.1 Methodology

We target the incorporation of a project-based

CSEDU2014-6thInternationalConferenceonComputerSupportedEducation

112

experiential learning experience into the K-12

students’ curriculum in the form of a guided

research experience on selected design and

optimization issues of a real-life electronic device

prototype. The ultimate objective is to enable the

basic understanding of the working principle of

modern media-rich electronic devices, AND to

foster the “researcher mind-set” in the students. The

goal is that the latter get used to the knowledge

construction process of CS researchers, since we

identify in the research activity many attributes of an

authentic educative experience while teaching the

science behind IT.

This holds promise of increasing students’

motivation to learn as well, especially in the

technology field, thus triggering a positive feedback

on future school-choices and professional careers

that can potentially reverse current negative trends in

the long run. The ultimate reason is twofold. On one

hand, students get a “real-time reward” for their

uptake of the knowledge construction process: the

possibility of making inroads into the working

principles of those electronic devices that are

pervasive in their life. Students in fact tend to

consider them out-of-reach of their knowledge

capability because of their complexity. On the other

hand, students are brought to the stage where they

can explore part of the design space of such devices,

through a guided research experience, with positive

implications on self-confidence and curiosity.

In practice, our approach aims at bringing

participatory-research with children (Christiansen

and James, 2008; Mortari, 2009) outside the

boundaries of humanistic studies, where it has been

mainly experimented so far. Some added values of

project-based learning based on a real problem from

CS research are:

- In the real world, knowledge (and its application)

is integrated, rather than split artificially into

subjects. Moreover, problem-solving is not a

school exercise with a predefined set of answers

but rather a complex engagement of an authentic

issue with multiple potential solutions (inquiry-

based learning). This feature, first characterized by

Dewey (1938), remains the distinctive hallmark of

experiential learning, central to our approach.

- It implicitly sets a broad range of learning

objectives that contribute to all of the pillars of

lifelong education, as identified by Delors (1996):

learning to know, learning to do, learning to live

together and learning to be.

The research experience needs to be guided by a

CS researcher from academia or industry for a

number of reasons, associated with his technical

expertise as well as with the different educational

interaction that he potentially raises in students

(Tenenberg, 2010; Fincher, 2013). He can point out

some key design choices that students would have

never thought about. He can also encourage students

to think more deeply about the problems, rather than

simply grasping “good enough” answers. Finally,

students experience increased relatedness to a

technology-related profession, and to the practice

that the researcher exposes them to. Simply

speaking, the feeling of “serious work”, of “complex

work made accessible”, and of “doing things right”

clearly increases students’ motivation.

3.2 Experimental Research Setting

Between 2000 and 2005 a fundamental design

paradigm shift took place in the field of computer

architecture. The application demand for more

performance-per-watt, especially in the embedded

computing domain, caused traditional monolithic

high-performance microprocessors to evolve toward

multicore architectures. In practice, the processing

workload started to be split among a number of

concurrent computation units, thus materializing

congruent multiples in performance speed-up and

power efficiency. This trend is currently well

underway, to the extent that manycore architectures

start to appear, that rely on hundreds of concurrent

processing units implemented onto the same

integrated circuit. The key component of a highly

parallel computing architecture consists of an on-

chip interconnection network (Network-on-chip,

NoC) capable of networking the processing cores

together onto the same parallel hardware platform.

Further technical details can be found in (Bertozzi

and Flich 2012). Our approach therefore aims at

familiarizing K-12 students with the paradigms of

computation parallelism and on-chip networking,

which are revolutionizing architectures and

applications in the embedded computing domain.

For the sake of keeping the research experience

focused, it will concern the routing problem in

NoCs. This latter consists of finding performance-

efficient routing paths for network packets to reach

their intended destinations. Feasible solutions to this

problem have to meet the deadlock avoidance

concern. Deadlock consists of a permanent blocking

condition of network traffic due to circular

dependencies in the routing channel request pattern.

Overall, during the research activity students will

have to devise feasible routing algorithms while

assessing the absence of such circular dependencies.

Moreover, such routing algorithms will be

GuidedParticipatoryResearchonParallelComputerArchitecturesforK-12StudentsThroughaNarrativeApproach

113

comparatively assessed from the viewpoint of their

effects on congestion and implementation

complexity.

At this point, the relevant problem of

overcoming the knowledge gap of K-12 students

arises. Addressing this problem requires a cross-

fertilization with findings from pedagogy research.

4 THE FIGURATIVE

STRUCTURE

4.1 Narrative Approach

Is it possible to teach very difficult scientific

concepts to young students? Different educators and

psychologist have considered this question and

diverse answers were given.

The psychologist Jerome S. Bruner (1966, p. 33)

wrote: “Any subject can be taught effectively in

some intellectually honest form to a child at any

stage of development”. Following Bruner, narrative

and propositional thinking are the ways in which

human beings structure their knowledge (Bruner,

1986). Usually, at school, sciences are taught using

formal language and logic-scientific thinking

(namely the paradigmatic one). Our proposal

considers a second opportunity: using the narrative

form to introduce formal and scientific knowledge.

Developing narrative understanding of the science

would complement the introduction of formal

explanations of how the world works.

Indeed, the elements of narrative are not foreign

to formal scientific understanding (Fuchs, 2013).

Stories can be used to deepen our understanding of

some physical concepts (e.g., in Fuchs’ work: the

gestalt of forces and its aspects), because schematic

and metaphoric structures are part of our everyday’s

life, in particular of children’s one.

Story form is a cultural universal which ‘reflects

a basic and powerful form in which we make sense

of the world and experience’ (Egan, 1986, p. 2).

Children especially use personal narratives to order

and explain the complexity of their experiences of

the world (Engel, 1999). Gallas (1994) presents how

children talked and wrote about science, and reports

on the complexities of the language and the stories

they used to understand the world of science. Using

narrative forms help children to get introduced to the

complexity of the world through an approach that

respects the form of their knowledge and their

human mind.

In applying this approach to our research

experience on parallel distributed computing, we are

facing two distinctive and unprecedented

challenges:

- The definition of a figurative structure that makes

on-chip interconnection networks and their system

integration function accessible to K-12 students

through a suitable metaphor.

- The application of the figurative structure to an

open-situation (i.e., the research experience). Thus,

the figurative structure should be realized as a plot,

rather than as a full and “static” story serving the

purpose of bringing pre-defined concepts within

reach of the cognitive abilities of students. The

plot of the story would be in fact the metaphor for

the constraints and operating conditions of a real

multicore processing environment. Pre-defining

only the plot enables the students to evolve it and

complete the story, by following a driving question

provided by the CS researcher. Providing answers

to the stated question will enable students to

augment the plot with missing details, which

correspond to technical solutions to a specific

research problem in the physical domain. Solutions

to problems will be derived by students in a

collaborative way, under the guidance of the CS

researcher, who will lead the in-class research.

4.2 The Narrative Approach at Work

The in-classroom research framework we propose

will be structured into a seven step methodology:

1- Setting the path to the research experience by an

in-class presentation of the CS researcher

bridging the gap between students’ pre-

knowledge and the concepts needed to start the

experience.

2- Definition of a figurative structure capable of

overcoming the technical knowledge gap of

students with respect to parallel computing and

on-chip networking (see Section 4.3).

3- Presentation of the figurative structure to the

students, with a clear distinction made between

predefined vs. undefined elements. The former

ones are the outcome of pre-taken design

decisions (e.g., the figurative structure for the

routing mechanism) and/or operating conditions

in the physical domain (e.g., synchronous

operation), while the latter ones represent the

available degrees of freedom for design space

exploration (e.g., the figurative structure for the

routing algorithm).

4- In-classroom collaborative research, where the

students will work out their solutions to the

routing problem under the guidance of the CS

researcher. This will not be done directly, but by

CSEDU2014-6thInternationalConferenceonComputerSupportedEducation

114

reasoning on the corresponding problem in the

figurative structure.

5- Selection of the best candidate routing solutions

for prototyping on a real on-the-field

programmable hardware platform, and definition

of a set of quality metrics and experiments for

their comparative assessments.

6- Taking the field-programmable hardware

platform (FPGA) to the classroom, after the

implementation variants of the routing

framework have been pre-implemented and

made quickly interchangeable. Running the

experiments and collection of experimental

results that should then be properly structured for

discussion (tables, figures).

7- Discussion of experimental results, with the CS

researchers having the key role of stimulating the

association of observed macroscopic results with

the low-level details and effects taking place in

the figurative structure of the hardware platform.

The researchers will lead the research activity

supporting students to identify questions, formulate

hypothesis, design solutions and problem-solving

strategies through dialogue and collaborative work

groups.

In order to guarantee the feasibility of the

methodology, two fundamental requirements need to

be fulfilled in the hardware prototype:

- Implementation of a networked multicore system

with fast reconfiguration capability of the routing

function. Runtime reconfiguration of the routing

algorithm should not imply the recompilation of

the hardware platform, so to meet timing

constraints of a class lecture.

- The platform should be equipped with a

graphical user interface for the sake of specifying

hardware parameters, collecting statistics and/or

monitoring specific functional effects while the

system is running. For this purpose, the GUI

should reflect the chosen figurative structure, and

graphically associate events in the hardware

platform with those in the figurative structure.

4.3 The Grid Street Plan Metaphor

The figurative structure of our on-chip

interconnection network is realized through the

metaphor of a grid street plan of a modern city (such

as New York City). Grid street plans are the

metaphor of 2D mesh topologies for NoCs in the

physical domain. The students will therefore explore

the design space of NoC routing algorithms by

taking routing decisions in a grid street plan. The

metaphor is so effective that in the early stage of on-

chip networking routing mechanisms were directly

inspired by the paradigm of driving directions

(Borkar, 1988). The key requirement for the

metaphor to hold consists of an initial alignment of

the metaphor to the feasibility space of NoCs. In

fact, the direct transposition of the grid street plan

implementation details (e.g., crossings, roundabouts,

traffic lights) to the NoC domain does not result in

efficient solutions. For instance, street crossings

managed via traffic lights or via the right-hand

precedence rule would result in poor communication

bandwidth in the electronic domain, since some

traffic streams would block other ones although

heading to different destinations. As a consequence,

the students will move from this consideration and

will be guided to design street crossings and grid

networks for which the metaphor holds, although the

resulting solutions will certainly be a cost-overkill in

a real city. In this direction, crossings will be

engineered in such a way that every arriving

direction is theoretically connected with all other

directions in a collision-free way. This implies the

implementation of multi-layer street crossings,

following the paradigm pictorially illustrated in

Figure 1.

Figure 1: Multi-layer street crossing as a metaphor of NoC

switches.

5 PREVIOUS RELATED WORK

The challenge to apply the narrative approach to

science education has been tackled by several

authors in the past (Fuchs 2013, 2007; Corni et al.,

2010). We refer to them in order to base our

proposal on a reliable pedagogical framework,

which is based on the narrative and story structure of

human knowledge (Egan, 1986; Bruner, 1986).

In his work, Fuchs (2007) creates figurative

conceptual structures for understanding physical

processes as a collection of force-dynamic gestalts

(quantity, quality, and power). These aspects are

structured with the help of metaphoric projections of

GuidedParticipatoryResearchonParallelComputerArchitecturesforK-12StudentsThroughaNarrativeApproach

115

image schemas. The application of analogy to the

various fields of continuum physics lets him

recognize a fundamental yet simple conceptual

structure - the same as that used in much of human

reasoning, not only in physics but also in

psychological and social situations.

Another example is provided by Falk, Herrmann,

Job, and Schmid (1983), who developed an approach

to teach Gibb’s thermodynamics stressing the use of

substance-like quantities.

We find that the exploitation of the narrative

approach for science education is only in the early

stage. Our novel contribution consists of using it as a

key enabler for a research experience with K-12

students. This implies not just the investigation of a

figurative structure for multicore processors and

their interconnection system, but also of its

suitability for “on-the-field” evolution.

6 CONCLUSIONS

We propose an innovative approach to CS education

at the K-12 level. Our main idea consists of bringing

research experience on parallel computer

architecture within reach of K-12 students, thus

jointly developing their knowledge level of the

matter as well as their personal attitudes. The key

enabler consists of the narrative approach, which we

exploit to overcome the technical knowledge gap of

the target students.

Our future work consists of further developing

the NoC metaphor and the HW/SW prototype for the

research experience, and of testing it in Italian

middle schools.

An educative research (Creswell, 2002; Mortari,

2007) will be conducted on this experience in order

to produce qualitative evidences about the impact of

the educative experience on the children’s thinking.

Qualitative research tools such as video and

audiotapes, interviews and written tasks will permit

to collect data about the experience itself and the

subjective student’s responses. A qualitative analysis

of these data will guarantee the possibility to

describe and assess the expected impact.

REFERENCES

Bertozzi, D., Flich, J., 2012. Designing Network-on-Chip

Architectures in the Nanoscale Era. CRC Press.

Borkar, S., et al., 1988. iWarp: An Integrated Solution to

High-Speed Parallel Computing. Proc.

Supercomputing.

Bruner, J. S., 1966. Toward a theory of instruction,

Harvard University Press. Cambridge.

Bruner, J. S., 1986. Actual minds, possible worlds,

Harvard University Press. Cambridge.

Christiansen, P. and James, A., 2008. (eds) Research with

children. Perspective and practice, Routledge.

Cochran, K. F., DeRuiter, J. A., King, R. A., 1993.

Pedagogical Content Knowing: An Integrative Model

for Teacher Preparation, J.Teacher Education, 44 (4),

pp.263-272.

Collins, A., 2006. Cognitive Apprenticeship. In R.

K.Sawyer, (ed) The Cambridge Handbook of Learning

Sciences, Cambridge University Press. Cambridge,

pp.47-60.

Corni F., Giliberti E., Mariani, C., 2010. A story as

innovative medium for science education in primary

school. GIREP conference, Reims.

Creswell, J. W., 2002. Educational Research. Planning,

Conducting and Evaluating Quantitative and

Qualitative Research, Pearson Education. New Jersey.

Dewey, J., 1938. Experience and Education, Collier

Books. USA.

Delors, J., 1996. Learning: The Treasure Within. Report to

UNESCO of the Int. Commission on Education for the

21st century.

Egan K., 1986. Teaching as Story Telling, The University

of Chicago Press. Chicago.

Engel, S., 1999. The story children tell. Making sense of

the narratives of Childhood, Freeman and Company.

USA.

European Commission/EACEA/Eurydice, 2012.

Developing Key Competences at School in Europe:

Challenges and Opportunities for Policy. Eurydice

Report. Luxembourg: Publications Office of the

European Union.

Falk G., Herrmann F., and Schmid G. B., 1983. Energy

forms or energy carriers? Am.J.Phys. 51(12), 1074-

1077.

Fincher, S., Knox, D., 2013. The Porous Classroom:

Professional Practices in the Computing Curriculum,

IEEE Computer, 46 (9), pp.44-51.

Fuchs H. U., 2013. From Image Schemas to Narrative

Structure in Science, ESERA Conference, Cyprus.

Fuchs H. U., 2007. From Image Schemas to Dynamical

Models in Fluids, Electricity, Heat, and Motion,

Zurich University of Applied Sciences at Winterthur.

https://home.zhaw.ch/~fusa/LITERATURE/Literature.

html.

Gallas, K. 1994. The languages of learning: How children

talk, write, dance, draw, and sing their understanding

of the world. New York: Teachers College Press.

Knobelsdorf, M., Vahrenhold, J., 2013. Addressing the

Full Range of Students: Challenges in K-12 Computer

Science Education. IEEE Computer, 46 (9), pp.32-37.

Knobelsdorf, M., 2011. Biographische Lern- und

Bildungsprozesse im Handlungskontext der

Computernutzung. Doct. dissertation, FU Berlin.

Meerbaum-Salant, O., Armoni, M., Ben-Ari M., 2010.

Learning Computer Science Concepts with Scratch. In

Proc. 6

Int. Workshop Computing Education

CSEDU2014-6thInternationalConferenceonComputerSupportedEducation

116

Research (ICER 10), ACM, pp.69-76.

Mortari, L. 2009. (ed) La ricerca per i bambini. Milano:

Mondadori.

Mortari, L. 2007. Cultura della ricerca e pedagogia,

Carocci. Roma.

Pasternak, A., Vahrenhold, J., 2012. Design and

Evaluation of a Braided Teaching Course in Sixth

Grade Computer Science Education. In Proc. 43

ACM Technical Symp. Computer Science Education

(SICGSE 12), ACM, pp.45-50.

Sawyer, R. K., 2009. The New Science of Learning, in R.

K. Sawyer, (ed) The Cambridge Handbook of

Learning Sciences, Cambridge University Press.

Cambridge, pp.1-16.

Tenenberg, J., 2010. Industry Fellows: Bringing

Professional Practice into the Classroom, Proc. 41

ACM Techn. Symp. Computer Science Education

(SIGCSE 10), ACM, pp.72-76.

GuidedParticipatoryResearchonParallelComputerArchitecturesforK-12StudentsThroughaNarrativeApproach

117