Statistical process control: what you don't measure can hurt you!

Eickelmann, Nancy; Anant, A.

doi:10.1109/ms.2003.1184166

Cited by 19 publications

(8 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As pointed out by [14], in hardware manufacturing, the number of observed failures is close to the actual number of failures that occur over time. In software, this is not the case.…”

Section: Statistical Quality Controlmentioning

confidence: 69%

See 1 more Smart Citation

What can Software Engineers Learn from Manufacturing to Improve Software Process and Product?

Schneidewind

2009

IIM

View full text Add to dashboard Cite

Abstract:The purpose of this paper is to provide the software engineer with tools from the field of manufacturing as an aid to improving software process and product quality. Process involves classical manufacturing methods, such as statistical quality control applied to product testing, which is designed to monitor and correct the process when the process yields product quality that fails to meet specifications. Product quality is measured by metrics, such as failure count occurring on software during testing. When the process and product quality are out of control, we show what remedial action to take to bring both the process and product under control. NASA Space Shuttle failure data are used to illustrate the process methods.

show abstract

“…As pointed out by [14], in hardware manufacturing, the number of observed failures is close to the actual number of failures that occur over time. In software, this is not the case.…”

Section: Statistical Quality Controlmentioning

confidence: 69%

“…In Figure 7 it is shown that the cumulative failure data Y i are approximately normally distributed by virtue of the data (red) dots being close to the normal (blue) line. Third, compute the means of the cumulative failures and the target cumulative failures in Equations (14) and (15) …”

Section: Design Of Experiments [18]mentioning

confidence: 99%

What can Software Engineers Learn from Manufacturing to Improve Software Process and Product?

Schneidewind

2009

IIM

View full text Add to dashboard Cite

show abstract

“…Moreover, according to the discussions in (Jalote, 2002(a); Eickelmann & Anant, 2003) we can consider three main differences between manufacturing and software processes that have to be kept in mind in order to assure a more appropriate use of SPC in software context in terms of control charts, run test indicators, anomalies interpretation and control limits calculation.…”

Section: Spc For Softwarementioning

confidence: 99%

Statistical Process Control for Software: Fill the Gap

Boffoli¹,

Baldassarre²,

Caivano³

2010

Quality Management and Six Sigma

View full text Add to dashboard Cite

The characteristic of software processes, unlike manufacturing ones, is that they have a very\ud high human-centered component and are primarily based on cognitive activities. As so, each\ud time a software process is executed, inputs and outputs may vary, as well as the process\ud performances. This phenomena is better identified in literature with the terminology of\ud “Process Diversity” (IEEE, 2000). Given the characteristics of a software process, its intrinsic\ud diversity implies the difficulty to predict, monitor and improve it, unlike what happens in\ud other contexts. In spite of the previous observations, Software Process Improvement (SPI) is a\ud very important activity that cannot be neglected. To face these problems, the software\ud engineering community stresses the use of measurement based approaches such as QIP/GQM\ud (Basili et al., 1994) and time series analysis: the first approach is usually used to determine\ud what improvement is needed; the time series analysis is adopted to monitor process\ud performances. As so, it supports decision making in terms of when the process should be\ud improved, and provides a manner to verify the effectiveness of the improvement itself.\ud A technique for time series analysis, well-established in literature, which has given\ud insightful results in the manufacturing contexts, although not yet in software process ones is\ud known as Statistical Process Control (SPC) (Shewhart, 1980; Shewhart, 1986). The technique\ud was originally developed by Shewhart in the 1920s and then used in many other contexts.\ud The basic idea it relies on consists in the use of so called “control charts” together with their\ud indicators, called run tests, to: establish operational limits for acceptable process variation;\ud monitor and evaluate process performances evolution in time. In general, process\ud performance variations are mainly due to two types of causes classified as follows:\ud  Common cause variations: the result of normal interactions of people, machines,\ud environment, techniques used and so on.\ud  Assignable cause variations: arise from events that are not part of the process and\ud make it unstable.\ud In this sense, the statistically based approach, SPC, helps determine if a process is stable or\ud not by discriminating between common cause variation and assignable cause variation. We\ud can classify a process as “stable” or “under control” if only common causes occur. More\ud precisely, in SPC data points representing measures of process performances are collected.\ud These values are then compared to the values of central tendency, upper and lower limit of\ud admissible performance variations.\ud While SPC is a well established technique in manufacturing contexts, there are only few\ud works in literature (Card, 1994; Florac et al., 2000; Weller, 2000(a); Weller, 2000(b); Florence,\ud 2001; Sargut & Demirors, 2006; Weller, & Card. 2008; Raczynski & Curtis, 2008) that present\ud successful outcomes of SPC adoption to software. In each case, not ...

show abstract

“…The defects were found during the first 3 months of software use in a production environment Number of defects reported to the service desk during the project (Lindströn 2004) Initial number of defects (Eickelmenn and Anant 2003) Initial number of defects in the code Total defects injected (Humphrey 2000) Overall productivity and quality trends for each project (Kulpa and Johnson 2008) Productivity Programmer productivity (Kulik and Weber 2002) Productivity (McGarry et al 2002;Stutzke 2005;Becker et al 2006;SEI 2006aSEI , 2006b Productivity (Kitchenham et al 2006) Effort/software size Productivity (McGarry et al 1998) Effort/software unit Process productivity parameter (Lindströn 2004) LOC/person Delivery rate…”

Section: Number Of Defectsmentioning

confidence: 99%

Defining a catalog of indicators to support process performance analysis

Monteiro¹,

Oliveira²

2009

J. Softw. Maint. Evol.: Res. Pract.

View full text Add to dashboard Cite

To achieve high-level maturity in software process improvement, organizations must change the project management focus from the empirical assessment of process performance to quantitative management of the software process based on performance measures and statistical techniques. To this purpose, managers need to work with years of measurement data to establish control limits and conduct performance analyses. In this context, some difficulties are encountered: Which measures should be used, considering that early definition is important? Which statistical technique is more adequate in each case? How should we work with the data? We found more than 500 different measures in the literature applicable to software process performance and a number of different statistical techniques to analyze and choose from. This article aims at defining a catalog of indicators with their related measures in answer to the above questions, in a way that can help managers perform process performance analysis of the Capability Maturity Model Integration for Development (CMMI-DEV) engineering processes. This article presents how we define and use this catalog and the complete specification of two indicators.

show abstract

Statistical process control: what you don't measure can hurt you!

Cited by 19 publications

References 0 publications

What can Software Engineers Learn from Manufacturing to Improve Software Process and Product?

What can Software Engineers Learn from Manufacturing to Improve Software Process and Product?

Statistical Process Control for Software: Fill the Gap

Defining a catalog of indicators to support process performance analysis

Contact Info

Product

Resources

About