Should duration and team size be used for effort estimation?

Takeshi Kakimoto, Masateru Tsunoda, Akito Monden

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Project management activities such as scheduling and project progress management are important to avoid project failure. As a basis of project management, effort estimation plays a fundamental role. To estimate software development effort by mathematical models, variables which are fixed before the estimation are used as independent variables. Some studies used team size and project duration as independent variables. Although they are sometimes fixed because of the limitation of human resources or business schedule, they may change by the end of the project. For instance, when delivery is delayed, actual duration and estimated duration is different. So, although using team size and project duration may enhance estimation accuracy, the error may also lower the accuracy. To help practitioners to select independent variables, we analyzed whether team size and duration should be used or not, when we consider the error included in the team size and the duration. In the experiment, we assumed that duration and team size include errors when effort is estimated. To analyze influence of the errors, we add n% errors to duration and team size. As a result, using duration as an independent variable was not very effective in many cases. In contrast, using maximum team size as an independent variable was effective when the error rate is equal or less than 50%.

Original languageEnglish
Pages (from-to)91-105
Number of pages15
JournalStudies in Computational Intelligence
Volume721
DOIs
Publication statusPublished - 2018

Fingerprint

Project management
Software engineering
Scheduling
Personnel
Mathematical models
Industry
Experiments

Keywords

  • Estimation error
  • Productivity
  • Project management
  • Software effort prediction

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Should duration and team size be used for effort estimation? / Kakimoto, Takeshi; Tsunoda, Masateru; Monden, Akito.

In: Studies in Computational Intelligence, Vol. 721, 2018, p. 91-105.

Research output: Contribution to journalArticle

Kakimoto, Takeshi ; Tsunoda, Masateru ; Monden, Akito. / Should duration and team size be used for effort estimation?. In: Studies in Computational Intelligence. 2018 ; Vol. 721. pp. 91-105.
@article{8d4fca6ccf1c40f0acf8e21400430ad1,
title = "Should duration and team size be used for effort estimation?",
abstract = "Project management activities such as scheduling and project progress management are important to avoid project failure. As a basis of project management, effort estimation plays a fundamental role. To estimate software development effort by mathematical models, variables which are fixed before the estimation are used as independent variables. Some studies used team size and project duration as independent variables. Although they are sometimes fixed because of the limitation of human resources or business schedule, they may change by the end of the project. For instance, when delivery is delayed, actual duration and estimated duration is different. So, although using team size and project duration may enhance estimation accuracy, the error may also lower the accuracy. To help practitioners to select independent variables, we analyzed whether team size and duration should be used or not, when we consider the error included in the team size and the duration. In the experiment, we assumed that duration and team size include errors when effort is estimated. To analyze influence of the errors, we add n{\%} errors to duration and team size. As a result, using duration as an independent variable was not very effective in many cases. In contrast, using maximum team size as an independent variable was effective when the error rate is equal or less than 50{\%}.",
keywords = "Estimation error, Productivity, Project management, Software effort prediction",
author = "Takeshi Kakimoto and Masateru Tsunoda and Akito Monden",
year = "2018",
doi = "10.1007/978-3-319-62048-0_7",
language = "English",
volume = "721",
pages = "91--105",
journal = "Studies in Computational Intelligence",
issn = "1860-949X",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Should duration and team size be used for effort estimation?

AU - Kakimoto, Takeshi

AU - Tsunoda, Masateru

AU - Monden, Akito

PY - 2018

Y1 - 2018

N2 - Project management activities such as scheduling and project progress management are important to avoid project failure. As a basis of project management, effort estimation plays a fundamental role. To estimate software development effort by mathematical models, variables which are fixed before the estimation are used as independent variables. Some studies used team size and project duration as independent variables. Although they are sometimes fixed because of the limitation of human resources or business schedule, they may change by the end of the project. For instance, when delivery is delayed, actual duration and estimated duration is different. So, although using team size and project duration may enhance estimation accuracy, the error may also lower the accuracy. To help practitioners to select independent variables, we analyzed whether team size and duration should be used or not, when we consider the error included in the team size and the duration. In the experiment, we assumed that duration and team size include errors when effort is estimated. To analyze influence of the errors, we add n% errors to duration and team size. As a result, using duration as an independent variable was not very effective in many cases. In contrast, using maximum team size as an independent variable was effective when the error rate is equal or less than 50%.

AB - Project management activities such as scheduling and project progress management are important to avoid project failure. As a basis of project management, effort estimation plays a fundamental role. To estimate software development effort by mathematical models, variables which are fixed before the estimation are used as independent variables. Some studies used team size and project duration as independent variables. Although they are sometimes fixed because of the limitation of human resources or business schedule, they may change by the end of the project. For instance, when delivery is delayed, actual duration and estimated duration is different. So, although using team size and project duration may enhance estimation accuracy, the error may also lower the accuracy. To help practitioners to select independent variables, we analyzed whether team size and duration should be used or not, when we consider the error included in the team size and the duration. In the experiment, we assumed that duration and team size include errors when effort is estimated. To analyze influence of the errors, we add n% errors to duration and team size. As a result, using duration as an independent variable was not very effective in many cases. In contrast, using maximum team size as an independent variable was effective when the error rate is equal or less than 50%.

KW - Estimation error

KW - Productivity

KW - Project management

KW - Software effort prediction

UR - http://www.scopus.com/inward/record.url?scp=85021947579&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021947579&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-62048-0_7

DO - 10.1007/978-3-319-62048-0_7

M3 - Article

AN - SCOPUS:85021947579

VL - 721

SP - 91

EP - 105

JO - Studies in Computational Intelligence

JF - Studies in Computational Intelligence

SN - 1860-949X

ER -