Inlämning av Examensarbete / Submission of Thesis

Mikael Gislen , pp. 45. COM/School of Computing, 2012.

The work

Författare / Author: Mikael Gislen
mikael@gislen.com
Titel / Title: Dysfunctional aspects of Software Development - An analysis of how lip-service, deception and organisational politics may side-track the result of well-intended methodologies
Översatt titel / Translated title: Problematiska aspekter inom programvaruutveckling - En analys av hur läpparnas bekännelser, vilseledande och organisationspolitik kan förstöra välmenande metoder
Abstrakt Abstract:

This paper tries to identify and understand the human social obstacles for developing quality software. These include lip-service, cutting corners, deception and effects of internal politics. These obstacles can undermine the good intentions behind the software methodologies. The paper draws from the literature in different disciplines and uses an ethnographic research methodology to create a rich picture of the concerning aspects in the framework of one software development company in India. What stands out among the findings are that internal audits has mainly focused on finding errors in documentation procedures but study of the actual practices has often been shallow. In addition the understanding of business risks by the internal auditors have often been weak.

Context. The human based obstacles affecting Software Development Methodology analysed in the context of an ISO 9000 quality system in an Indian Software Development company working mainly with Swedish companies.
Objectives. Identifying and increasing the understanding of intrinsic negative social aspects such as lip service, cheating and politics which are affecting the results of Software Development Methodologies and if possible suggest some means to mitigate them. In particular to create a deeper understanding of why people cheat and pay lip service to methodologies and to try to understand the political aspects of methodology and quality systems. There are other positive social aspects, but they are not considered since the objective is about understanding the negative aspects and possibly mitigate them.
Methods. Ethnographic research using analysis of ISO 9000 and design artefacts, semi-structured interviews, participation in internal audits,
Results. Most focus in audits was on documentation and very less focus on underlying methodologies, some indications of lip-service to process and processes were also mainly managed on a higher level in the organization while the understanding and practices were less well established on lower level. It was hard to get a grip on the internal political aspects since the perception of the subject in the informants view was that it is mainly malicious and therefore embarrassing to speak about. Some conflicts between internal quality goals and customers’ needs were also identified.
Conclusions. An ethnographic research methodology gives a rich picture. The analysis gives deeper understanding of the problem areas, but not necessary solutions. The author suggests that at the heart of the problem is a difference in world view. Software professionals generally tend to resolve [technical] problems using a reductionist approach, while these intricate challenges cannot easily be resolved by this approach. A more holistic systemic approach is required and while the software methodology is useful to structure the development it does not resolve these dysfunctions. They have to be resolved on another level. It was also found that further studies is required in particular to better understand Internal politics, the effect of Positive and Negative Incentives, the effect of software metrics on quality performance and subjectivity in customers’ perception and expectation.

Populärvetenskaplig beskrivning / Popular science summary: Uppsatsen försöker förstå mänskliga sociala hinder för att utveckla programvara med hög kvalitet. Dessa hinder inkluderar "läpparnas bekännelser", ta genvägar, vilseleda och internpolitik. Sådana hinder kan underminera utvecklingsmetoder. Denna uppsats bygger på literatur från olika discipliner och använder etnografisk forskningsmetodik för att skapa en rik bild av dessa oroande aspekter inom ramen för ett IT-företag i Indien. Vad står ut är att intern kvalitetsrevision fokuserar för mycket på att finna fel i dokumentationsprocesser medan revision av hur utvecklingsarbetet faktiskt sker har varit för ytligt. Dessutom har de interna revisorernas förståelse av affärsrisker varit svag.
Ämnesord / Subject: Datavetenskap - Computer Science\Informatics
Datavetenskap - Computer Science\Software Engineering
Företagsekonomi - Business Administration\Management Control
Nyckelord / Keywords: Agile Software Development, Computer Supported Cooperative Work, Cultural Studies, Ethnography, Global Software Development, Human Computer Interaction, Information Systems, Internal Politics, ISO 9000, Micro Economics, Participatory Design, Psychology, Quality Systems, Software Engineering, Software Process, Software Quality, TickIT, The Human Role In Software Development

Publication info

Dokument id / Document id: houn-92q94s
Program:/ Programme Master of Science in Informatics
Registreringsdatum / Date of registration: 12/06/2012
Uppsatstyp / Type of thesis: Magisterarbete/Master's Thesis (60 credits)

Context

Handledare / Supervisor: Hans Tap
hans.tap@bth.se
Examinator / Examiner: Sara Eriksén
Organisation / Organisation: Blekinge Institute of Technology
Institution / School: COM/School of Computing

+46 455 38 50 00

Files & Access

Bifogad uppsats fil(er) / Files attached: bth2012gislén.pdf (586 kB, öppnas i nytt fönster)