A Guttman Scaling of CMM Level 2 Practices: Investigatingthe Implementation Sequences Underlying Software EngineeringMaturity

  • Authors:
  • John G. Bilotta;John F. McGrew

  • Affiliations:
  • Charles Schwab and Co., Inc., San Francisco, CA 94104, USA;SBC Services, Inc., Pleasanton, CA 94588, USA

  • Venue:
  • Empirical Software Engineering
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

The Capability Maturity Model specifies several hundredkey practices that must be implemented as a team moves from Level1 to Level 5 of the model. However, the model does not specifywithin each level the optimal sequence in which to implementthe practices. Level 2 contains 121 such practices grouped undersix key process areas (KPAs) which are further subdivided intofive common feature areas (CFAs). Although the KPA/CFA structurehas a logical fit with the organizational structure of very largesoftware development teams, it does not correspond to the realityof small teams. Using Level 2 audit data collected on 10 smallsoftware development teams, the authors try to determine whetherthe data itself can point to a more appropriate implementationstrategy for small teams. The data is analyzed using Guttmanscaling techniques (scalogram analysis). The results indicatethat there is a single underlying, orderable dimension at Level2 which lays out a step-by-step path upward from Level 1. Theorder of the items is found to map well to the familiar Plan-Do-Check-Actcycle widely used by project managers to organize and controlwork efforts. The extracted scale can be used as an assessmenttool to provide management with a quick snapshot of a team‘scurrent position relative to Level 2.