DEVELOPING THE SKILLS NEEDED FOR REQUIREMENT

ELICITATION IN GLOBAL SOFTWARE DEVELOPMENT

Miguel Romero

University of B´ıo-B´ıo, Department of Computer Science and Information Technologies

Avenida Andr´es Bello s/n 3780000 Chill´an, Chile

Aurora Vizca´ıno, Mario Piattini

University of Castilla-La Mancha, Alarcos Research Group - Institute of Information Technologies & Systems

Department of Information Technologies & Systems - Escuela Superior de Inform´atica

Paseo de la Universidad 4, 13071 Ciudad Real, Spain

Keywords:

Global Software Development, Requirement Elicitation, Training, Education.

Abstract:

The requirement elicitation stage is that which is most critical in the development of a software product. How-

ever, this stage is not covered on teaching courses with the required depth, nor is invested the necessary time in

training students and practitioners in these tasks. There is currently a trend towards global software develop-

ment (GSD) which complicates the process of elicitation requirements since, for instance, communication is

more difﬁcult because stakeholders are geographically distributed. Moreover, the elicitation in GSD involves

a variety of characteristics that are not often taught in software engineering courses. This paper presents some

of the most important factors which may affect elicitation in GSD. Furthermore, we propose techniques with

which to help students and software engineers to develop some of the skills needed to carry out the elicitation

process in GSD.

1 INTRODUCTION

Independently of the methodological approach used,

the requirement elicitation phase is the ﬁrst step in

the process of developing a software product. In such

stage requirements are discovered and documented

fulﬁlled which must satisﬁed by the software in or-

der to satisfy expectations and needs for clients and

users. The aforementioned stage is the most critical

phase in software development, because the mistakes

made at this stage are more expensive and difﬁcult to

resolve owing to their impact upon the other stages. In

fact, the 85% of the defects in software systems come

from the requirement elicitation stage: 49% from in-

correct assumptions, 29% from omitted requirements,

13% from requirement inconsistency and 5% from

ambiguous requirements (Young, 2002).

Given the importance of this stage is fundamental

to have professionals trained in this process, who are

capable of accomplishing top-quality requirements

elicitation. However, professionals who have recently

graduated from universities lack the skills and abil-

ities necessary to carry out this stage properly be-

cause during their degree little time is usually spent

on training in this phase of software engineering and

they often do not perform professional practices. In

fact, according to a survey dealing with university

curriculums carried out amongst 214 software engi-

neers, the requirements elicitation process is ranked

second among the topics that should be improved in

its teaching at universities and third among the top-

ics that may have the greatest need for training (Leth-

bridge, 2000). This is a problem for companies since

they have difﬁculties in recruiting qualiﬁed personnel,

and it is also a problem for new graduates who may be

discriminated against because of their lack of experi-

ence. Because students do not graduate with the skills

necessary for requirement elicitation, they must ac-

quire these skills during their company training, and

it is therefore companies which take the responsibility

for and assume the costs of such training.

With the arrival of new trends in software develop-

ment, the gap between what is taught in universities in

the ﬁeld of requirements engineering and what takes

place in practice is widening. One of these trends

is Global Software Development (GSD) (Herbsleb,

2007). GSD has grown considerably in recent years

due to the culture of globalization and other factors

393

Romero M., Vizcaíno A. and Piattini M. (2008).

DEVELOPING THE SKILLS NEEDED FOR REQUIREMENT ELICITATION IN GLOBAL SOFTWARE DEVELOPMENT.

In Proceedings of the Tenth International Conference on Enterprise Information Systems - DISI, pages 393-396

DOI: 10.5220/0001671103930396

 SciTePress

such as offshoring, and it will continue to do so (As-

pray et al., 2006). However, this practice is causing

some challenges (Herbsleb and Moitra, 2001): prob-

lems caused by inadequate communication, problems

caused by the cultural diversity, problems in knowl-

edge management, and problems caused by time dif-

ferences.

The objective of our work is to study what fac-

tors inﬂuence the global requirement elicitation pro-

cess and to propose a set of techniques through which

to develop some of the skills needed for this process

in students and practitioners.

The following section describes the most impor-

tant factors that inﬂuence the success or failure of an

elicitation requirements in GSD. Various skills which

would be desirable in order to carry out this process

are outlined in Section Three. Our proposal is ex-

plained in Section Four, and ﬁnally in Section Five

some conclusions and future work are presented.

2 FACTORS INVOLVED IN THE

REQUIREMENT ELICITATION

IN GSD

Multiple factors determine the success or failure of a

GSD project. Herbsleb and Moitra (2001) present the

problem of GSD and raise the issue of the dimensions

that characterize it. These are: strategic issues, cul-

tural issues, inadequate communication, knowledge

management, project and process management issues,

technical issues.

These dimensions were studied by Damian and

Zowghi (2002) in order to identify how these prob-

lems affect each phase of engineering requirements.

The challenges identiﬁed for the elicitation stage are

shown below: differences in customer culture and

business (Cultural diversity), achieving appropriate

participation of system users and ﬁeld personnel (In-

adequate communication, Time difference), and man-

age conﬂict and have open discussion of interests (In-

adequate communication, Knowledge management,

Time difference).

Rafo and Setamanit (2005) present a list of im-

portant factors in GSD which is grouped into three

categories:

• The fundamental factors: Communication prob-

lems, Coordination and control problems, Cul-

tural differences, Language difference, Time-

Zone differences.

• The strategic factors: Development site, Product

architecture, Development strategy, Distribution

overhead, Distribution effort loss.

• Organizational factors: Team formulation, Team

dynamics.

Bhat et al. (2006) identify the success factors in

offshoring outsourcing projects based on root-cause

analysis. The factors which were identiﬁed are:

Shared goal, Shared culture, Shared process, Shared

responsibility, Trust.

In summary, the factors that have a major impact

in the requirementselicitation in GSD. The factors are

the following ones: cultural differences, communica-

tion, time difference, knowledge management, lan-

guage difference, trust, team formulation and team

dynamics.

These relevant factors are considered in the deﬁni-

tion of the basic skills for the requirements elicitation

in GSD presented in the following section.

3 BASIC SKILLS FOR

REQUIREMENTS

ELICITATION IN GSD

A list of 31 skills which are necessary for analyst

requirements is presented in Young (2006) . Some

of the skills presented are the following: criteria for

a good requirement, customer/user involvement with

requirements (joint team), identifying real require-

ments (from the stated requirements), anticipating

and controlling requirement changes, and so on.

This list of abilities is a good starting point, but it

does not consider particulars aspects of elicitation in

GSD.

Different skills for GSD are discussed in the lit-

erature. By example, the next list of skill is pre-

sented in Damian et al.(2006): international Team-

work, computer mediate communication, iterative de-

velopment in remote client-developer relationships,

living with ambiguity/uncertainty in remote teams,

and distributed project management.

Taking into account those skill and the factors

mentioned in the previous section, we identiﬁed the

following basic skills that a global requirement engi-

neer should have:

Knowledge of culture. A knowledge of the norms,

beliefs, business ethos, and skill in the native lan-

guage of the potential client is highly important

(Nguyen et al., 2006; Huang and Trauth, 2007).

Therefore software engineers should get used to deal-

ing with people who have different customs, culture

and even ways of interacting.

International Teamwork. Work in distributed

teams to solve large problems and collaborativedevel-

opment of shared understanding of project goal and

ICEIS 2008 - International Conference on Enterprise Information Systems

394

constraints (Damian et al., 2006), in addition the skills

to elicit requirements in distributed environment.

Communication skills. As was mentioned in the

previous section, one of the biggest problems in GSD

is communicationbecause stakeholders are geograph-

ically distributed. To face this problem the engineers

need the following skills:

• Conversational skills in English language, be-

cause it is the most common business language

(Adya, 2006; Huang and Trauth, 2007; Richard-

son et al., 2007).

• Knowledge of Communication protocols and

strategies, in order to avoid potential miscom-

munications and misinterpretations (Huang and

Trauth, 2007; Richardson et al., 2007).

• Using groupware tools such as: e-mail, video con-

ference, wiki, instant message, and so on to com-

municate. Therefore, one of the skills that a soft-

ware engineer should have is knowing how to use

this kind of tools in order to obtain the maximum

beneﬁt from them (Damian et al., 2006; Richard-

son et al., 2007).

4 TECHNIQUES TO DEVELOP

SKILLS

In order to develop these skills in software engineers

or even in students we propose using techniques of

simulation and artiﬁcial agents.

Simulation is a technique which has been used in

teaching for many years. It is successfully used, for

example, in medicine and aviation. The main advan-

tage is that it allows students to train themselves with-

out the risk of a real environment and at a lower cost.

Educational Virtual Environments (EVEs) are fre-

quently used for simulation. These environments use

Virtual Reality (VR) to create virtual worlds in which

students (Ieronutti and Chittaro, 2007):

• Directly experience certain physical properties

(e.g. shape, size and time duration) of objects and

events.

• Change their point of view in order to access new

and/or unusual perspectives.

• Interact with objects either to discover and study

their hidden elements or to evaluate the effects of

their manipulation.

In training EVEs can, moreover (Ieronutti and Chit-

taro, 2007):

• Provide a low-cost alternative to creating full-

scale physical training scenarios.

• Offer the opportunity to create a wide variety of

scenarios including those rarely (or never previ-

ously) encountered in the real world.

• Simulate training scenarios that can be run repeat-

edly.

• Include a monitoring of progress during training

sessions to evaluate learners’ skills.

In Sims (2007) an EVE is presented which uses a Vir-

tual Human with the goal of training students in the

Arabic language and in Arabic cultural familiariza-

tion. Some of the cultural aspects that are discussed

in this EVE are:

• Appropriate and inappropriate use of honoriﬁc

and family names.

• Colloquial terms for policemen, soldiers, and

strangers.

• Iraqi gestures that may be misinterpreted by

Americans, and American gestures that may be

misinterpreted by Iraqis.

• Methods to calm tense situations.

• Proper and improper ways in which to interact

with Iraqi women.

• Showing respect for family relationships.

Other EVE is Lok (2006) where the virtual Human are

use for simulating the patients who are interviewed by

the students of medicine. The interaction between vir-

tual patient and students are talk in a natural manner.

In the context of global requirement elicitation, a

simulation environment that uses virtual reality will

be useful in teaching students and practitioners how

they should perform interviews or maintain a conver-

sation with different types of stakeholders.

The artiﬁcial agents paradigm constitutes a natu-

ral metaphor for systems with purposeful interacting

agents, and this abstraction is close to the human way

of thinking about their own activities (Wooldridge and

Ciancarini, 2000). This foundation has led to an in-

creasing interest in social aspects such as motivation,

leadership, culture or trust (Fuentes et al., 2004).

We propose the use of artiﬁcial agents to model

the EVE because:

• Agents operate without the direct intervention of

humans or others, and have some kind of control

over their actions and internal states (Autonomy).

• Agents interact with other agents (and possibly

humans) via some kind of agent communication

language. This feature will be highly important

in the simulation of human interactions in dis-

tributed teams (Social Ability).

DEVELOPING THE SKILLS NEEDED FOR REQUIREMENT ELICITATION IN GLOBAL SOFTWARE

DEVELOPMENT

395

• Agents perceive their environment and respond in

a timely fashion (reactivity).

• Agents can take the initiative and achieve their

own goals (Pro-activeness).

It is therefore possible to implement an EVE with dif-

ferent agents which have different types of behavior

and simulate people of different cultures, characters

or language, thus assisting software engineering and

students to develop the desirable skills work in elici-

tation groups in GSD.

5 CONCLUSIONS

GSD is a current trend. However, although this

paradigm has several advantages it also brings cer-

tain disadvantages which have been summarized in

this paper. In order to decrease the effect of some

of these problems we propose using different tech-

niques to train students and practitioners in working

with people who are geographically distributed and

who may have different cultures, languages and time

zones.

As future work we are going to develop of a sim-

ulator based on virtual agents which will allow stu-

dents to be trained in the process of elicitation require-

ments in the area of global development. The agents

will permit the simulation of stakeholders’ behaviour

by considering cultural, cognitive and geographic as-

pects, and also time differences.

ACKNOWLEDGEMENTS

This work is partially supported by the MELISA

project (PAC08-0142-3315), Junta de Comunidades

de Castilla-La Mancha, Consejer´ıa de Educaci´on

y Ciencia, in Spain; ESFINGE project (TIN2006-

15175-C05-05) Ministerio de Educaci´on y Cien-

cia (Direcci´on General de Investigaci´on)/Fondos Eu-

ropeos de Desarrollo Regional (FEDER) in Spain;

the CompetiSoft project (506AC0287, CYTED pro-

gram).

REFERENCES

Adya, M. P. (2006). Imparting global software development

experience via an it project management course: Crit-

ical success factors. In COMPSAC’06, pp. 51–52.

Aspray, W., Mayadas, F., and Vardi, M. (2006). Glob-

alization and Offshoring of Software. Association

for Computing Machinery (ACM), Job Migration

Task Force, March 2006. ACM 0001-0782/06/0200,

http://www.acm.org/globalizationreport.

Bhat, J. M., Gupta, M., and Murthy, S. N. (2006). Overcom-

ing requirements engineering challenges: Lessons

from offshore outsourcing. IEEE Software, 23(5):38–

44.

Damian, D., Hadwin, A., and Al-Ani, B. (2006). In-

structional design and assessment strategies for teach-

ing global software development: a framework. In

ICSE’06,pp. 685–690.

Damian, D. E. and Zowghi, D. (2002). The impact of

stakeholders’ geographical distribution on managing

requirements in a multi-site organization. In RE’02,

pp. 319–328.

Fuentes, R., G´omez-Sanz, J., and Pav´on, J. (2004). A Social

Framework for Multi-agent Systems Validation and

Veriﬁcation. Wang, S. et al (Eds.) ER Workshop 2004,

Springer Verlag, LNCS, 3289:458–469.

Herbsleb, J. D. (2007). Global software engineering: The

future of socio-technical coordination. In FOSE’07 at

ICSE’07, pp. 188–198.

Herbsleb, J. D. and Moitra, D. (2001). Guest editors’ in-

troduction: Global software development. IEEE Soft-

ware, 18(2):16–20.

Huang, H. and Trauth, E. (2007). Cultural inﬂuences

and globally distributed information systems develop-

ment: experiences from chinese it professionals. In

SIGMIS-CPR ’07, pp. 36–45.

Lethbridge, T. (2000). What knowledge is important to a

software professional? IEEE Computer, 33(5):44–50.

Lok, B. (2006). Teaching communication skills with virtual

humans. IEEE Computer Graphics and Applications,

26(3):10–13.

Ieronutti, L. and Chittaro L. (2007). Employing virtual hu-

mans for education and training in x3d/vrml worlds.

Computer & Education, 2007(49):93–109.

Nguyen, P., Babar, M., and Verner, J. (2006). Critical fac-

tors in establishing and maintaining trust in software

outsourcing relationships. In ICSE’06, pp. 624–627.

Raffo, D. and Setamanit, S. (2005). A simulation model for

global software development project. In The Interna-

tional Workshop on Software Process Simulation and

Modeling.

Richardson, I., Moore, S., Paulish, D., Casey, V., and Zage,

D. (2007). Globalizing software development in the

local classroom. In CSEET’07, pp. 64-71.

Sims, E. M. (2007). Reusable, lifelike virtual humans for

mentoring and role-playing. Computers & Education,

2007(49):75–92.

Wooldridge, M. and Ciancarini, P. (2000). Agent-Oriented

Software Engineering: The State of the Art. First Int.

Workshop on Agent-Oriented Software Engineering,

LNCS 1957:1–28.

Young, R. (2002). Recommended requirements gathering

practices. CROSSTALK The Journal of Defense Soft-

ware Engineering, 2002(April).

Young, R. (2006). Twelve requirements basics for project

success. CROSSTALK The Journal of Defense Soft-

ware Engineering, 2006(December).

ICEIS 2008 - International Conference on Enterprise Information Systems

396