Back to search

North Dakota State University

Using Human Error Models to Improve the Quality of Software Requirements

Abstract

dc:description.abstract

Creating high quality software is a primary concern for software development organizations. Researchers have devoted considerable effort in developing quality improvement methods that help software engineers detect faults early in the development lifecycle (when the faults are cheapest to detect and repair). While useful, the available approaches still cannot make sure that Software developers are able to identify all or even a significantly large portion of faults. This is because they do not help software developers identify errors (i.e., underlying cause of faults) that may have led to the insertion of the faults (i.e., manifestation of error). This lack of focus on errors causes some faults to be overlooked which impacts quality of software produced. Requirements engineering is the most people-intensive phase of software development. Thus, requirements engineering is more prone to human error when compared to other phases of software development. To that end, this dissertation focuses on understanding the human error causes of requirements faults. The central idea that drives this dissertation is that, knowledge of errors that commonly occur during the requirements engineering process can help software developers in detecting faults that are otherwise overlooked when using traditional approaches and also help them to avoid making errors when developing requirements. Human error research focuses on understanding and classifying the fallibilities of human cognition. This dissertation combines requirements error information (gathered from Software Engineering literature) with the general accounts of human error and human error models (gathered from the Psychology literature). There are three steps to this work: development of a requirements phase human error taxonomy, empirical validation of the taxonomy’s usefulness for understanding requirements faults and errors, and development and subsequent validation of a formal software inspection technique based on the taxonomy. As a result of this dissertation, a structured Human Error Taxonomy (HET) that classifies requirements phase errors was created with direct ties to the existing human error theories. Several empirical validations of the taxonomy have helped in: successfully demonstrating the taxonomy’s usefulness for understanding requirements faults and errors, and developing a formal HET-based Error Abstraction and Inspection (EAI) approach and supplementary human error investigation tools.

Degree

thesis:*
Grantor dc:publisher
North Dakota State University
Year dc:date.issued
2018

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Anu, Vaibhav Kumar
ORCID dc:identifier.orcid

Rights

dc:rights
Statement dc:rights
  • NDSU Policy 190.6.2

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/10365/28760

Chain of custody

source
Harvested from
North Dakota State University
Base URL
library.ndsu.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Anu, Vaibhav Kumar. Using Human Error Models to Improve the Quality of Software Requirements. North Dakota State University, 2018. https://hdl.handle.net/10365/28760