Question 4

Some of the adaptive approaches to developing systems (like agile) are becoming more popular. Are they better or more appropriate than prescriptive approaches in most cases? Why or why not?

Many IT projects involve researching, analyzing, and then purchasing and installing new hardware and software with little or no actual software development required. However, some projects involve minor modifications to enhance existing software or to integrate one application to another. Other projects involve a major amount of software development. Many argue that developing software requires project managers to modify traditional management methods, depending on a particular product’s life cycle. 

A system development life cycle (SDLC) is a framework for describing the phase of developing information systems. Some popular models of an SDLC include the waterfall model, the spiral model, the incremental build model, the prototyping model, and the Rapid Application Development (RAD) model. 

These life cycle models are examples of a predictice life cycle, meaning that the scope of the project can be articulated clearly and the schedule and cost can be predicted accurately. The project team spends a large portion of the project attempting to clarify the requirements of the entire system and the and then producing a design. One issue with this system style is that users are often unable to see tangible results in terms of working software for an extended period.

Below are brief descriptions of the most used predictive SDLS models: 

The “waterfall life cycle” model has well-defined, linear systems analysis, design, construction, testing and support. This model assumes that requirements will remain stable after they are defined. Used when risk must be tightly controlled and restricted after requirements are defined. Works well with large-scale projects where complexity and cost are so high that the more rigid steps. Fig.1

The “spiral life cycle” model was developed based on refinement of the waterfall model as applied to large government software projects. It recognizes the fact that most software is developed using an iterative or a spiral approach rather than a linear style. The project team is open to changes and revisions later in the project life cycle, and return to requirements phase to more clarification and design. This approach is suitable for projects in which changes can be incorporated with reasonable cost increases or with acceptable time delays. Fig.1

Figure 1 (Waterfall and Spiral Model Life Cycle)

Additional predictive models are:

The “incremental build life cycle” mode provides for progressive development of operational software, with each release providing added capabilities; used by Microsoft.

The “prototyping life cycle” model is used for developing software prototypes to clarify user requirements for operational software. Heavy user interface interaction needed for development. Physical design specification simultaneously. Able to throw away or keep prototypes, depending on the project. Often used as an example for mobile applications.

The “RAD life cycle” model uses an approach in which developers work with an evolving prototype Heavy user involvement needed. Developers use additional tool aids for rapid prototyping and code generation.

In contrast to the predictive models, adaptive software development (ASD) life cycle model assumes that software development follows an adaptive approach because the requirement cannot be clearly expressed early in the life cycle. This model is also used to provide more freedom than the prescriptive approach, giving development more free-form to create components that provide functionality for specified by the business group. The projects are mission driven and component based, using time-based cycles to meet target dates. Requirements are developed using an iterative approach, and development is a risk driven and change tolerant to address and incorporate risks rather than mitigate them.

More recently, the term agile software development has become popular to describe new approaches that focus on close collaboration between programming teams and business experts. 

Depending on the need of the project, the chart below illustrates the pros and cons of the most popular model used by project managers, Agile vs. Waterfall: