Reasons for software effort estimation error: impact of respondent role, information collection approach, and data analysis method

Selecting Best Practices for Effort Estimation

“…Third, in 14 cases (rows 8,9,12,13,15,16,17,18,21,22,23,26,27, and 28) the WRAPPER discarded, on average, five to six variables and sometimes many more (e.g., in rows 9, 13, 22, and 27, the surviving models discarded nearly half the variables). This result, together with the last one, raises concerns for those that propose changes to business practices based on the precise COCOMO numerics published in the standard COCOMO texts [3], [4].…”

“…Before we can compare any effort estimation methods (be they model-based or expertbased), we must first tame the large deviation problem. For more on expert-based methods, see [2], [6], [9], [16] Biases in the model. This study uses COCOMO data sets since that was the only public domain data we could access.…”

Is This Cost Estimate Reliable? -- The Relationship between Homogeneity of Analogues and Estimation Reliability

“…The threshold can be given by some standard evaluation criteria such as Estimations reliability is usually assessed by exploring reasons or risk factors for estimation error, and then checking whether the target project has the potential risks. Risk factors have been investigated by questionnaire-based approaches [17], [25], [20] and by statistical analysis [11], [7], [10]. For example, Lederer and Prasad [12] reported that one of the most important risks is having "overlooked tasks" in measurement.…”

Organizational Strategies and Social Interaction Influence in Software Development Effort Estimation

“…As a result, projects with inaccurate resource allocation will generally experience serious schedule delay or cost overrun. [12][13][14][15][16] The Standish Group estimates that in 2013 the worldwide yearly spending for software projects was $750 billion. The United States accounted for about 40% of this number or about $300 billion.…”