Using complexity measures to evaluate software development projects: A nonparametric approach

Qing Cao, Vicky Ching Gu, Mark Andrew Thompson

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In this article, we use newly developed complexity metrics for software development projects that are more useful than traditional measures such as lines of code and functional points. Next, we present an approach to assessing the relative efficiency of software projects using these complexity measures as outputs. Due to the nature of the complexity measures, the constant returns to scale assumption often used in data envelopment analysis (DEA) is not appropriate. We relax this assumption and estimate the DEA model assuming variable returns to scale. This two-step approach provides project managers with a decision support tool to assess project productivity, categorize projects, and evaluate critical success/failure factors in software development projects.

Original languageEnglish (US)
Pages (from-to)274-283
Number of pages10
JournalEngineering Economist
Volume57
Issue number4
DOIs
StatePublished - Dec 1 2012
Externally publishedYes

Fingerprint

Data envelopment analysis
software development
development project
Software engineering
data analysis
Managers
Productivity
productivity
manager
efficiency
Development projects
Software development

ASJC Scopus subject areas

  • Education
  • Engineering(all)
  • Economics and Econometrics

Cite this

Using complexity measures to evaluate software development projects : A nonparametric approach. / Cao, Qing; Gu, Vicky Ching; Thompson, Mark Andrew.

In: Engineering Economist, Vol. 57, No. 4, 01.12.2012, p. 274-283.

Research output: Contribution to journalArticle

@article{5aa848d17e194e299a255d92ad1cf2f0,
title = "Using complexity measures to evaluate software development projects: A nonparametric approach",
abstract = "In this article, we use newly developed complexity metrics for software development projects that are more useful than traditional measures such as lines of code and functional points. Next, we present an approach to assessing the relative efficiency of software projects using these complexity measures as outputs. Due to the nature of the complexity measures, the constant returns to scale assumption often used in data envelopment analysis (DEA) is not appropriate. We relax this assumption and estimate the DEA model assuming variable returns to scale. This two-step approach provides project managers with a decision support tool to assess project productivity, categorize projects, and evaluate critical success/failure factors in software development projects.",
author = "Qing Cao and Gu, {Vicky Ching} and Thompson, {Mark Andrew}",
year = "2012",
month = "12",
day = "1",
doi = "10.1080/0013791X.2012.729878",
language = "English (US)",
volume = "57",
pages = "274--283",
journal = "Engineering Economist",
issn = "0013-791X",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

TY - JOUR

T1 - Using complexity measures to evaluate software development projects

T2 - A nonparametric approach

AU - Cao, Qing

AU - Gu, Vicky Ching

AU - Thompson, Mark Andrew

PY - 2012/12/1

Y1 - 2012/12/1

N2 - In this article, we use newly developed complexity metrics for software development projects that are more useful than traditional measures such as lines of code and functional points. Next, we present an approach to assessing the relative efficiency of software projects using these complexity measures as outputs. Due to the nature of the complexity measures, the constant returns to scale assumption often used in data envelopment analysis (DEA) is not appropriate. We relax this assumption and estimate the DEA model assuming variable returns to scale. This two-step approach provides project managers with a decision support tool to assess project productivity, categorize projects, and evaluate critical success/failure factors in software development projects.

AB - In this article, we use newly developed complexity metrics for software development projects that are more useful than traditional measures such as lines of code and functional points. Next, we present an approach to assessing the relative efficiency of software projects using these complexity measures as outputs. Due to the nature of the complexity measures, the constant returns to scale assumption often used in data envelopment analysis (DEA) is not appropriate. We relax this assumption and estimate the DEA model assuming variable returns to scale. This two-step approach provides project managers with a decision support tool to assess project productivity, categorize projects, and evaluate critical success/failure factors in software development projects.

UR - http://www.scopus.com/inward/record.url?scp=84870618879&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84870618879&partnerID=8YFLogxK

U2 - 10.1080/0013791X.2012.729878

DO - 10.1080/0013791X.2012.729878

M3 - Article

AN - SCOPUS:84870618879

VL - 57

SP - 274

EP - 283

JO - Engineering Economist

JF - Engineering Economist

SN - 0013-791X

IS - 4

ER -