Process Improvement

I am keen to try the Test Process Improvement (TPI) model when reviewing testing processes and mapping out improvement programmes. I have not yet had the chance to use TPI, but my studies and research have persuaded me that it has significant advantages over the models that are better known in the UK, i.e. SEI-CMM, SEI-CMMI and TMM.

TPI is designed specifically to improve testing and to identify those improvements that would be of most value to an organisation. The other, popular models do not allow organisations sufficient flexibility, being designed to improve processes across the board without acknowledging that some processes are less, or more, important in specific cases.

Testing Processes

I am also keen to have the chance to apply nbsp;the Dutch testing process model, TMap®. This is the basis for TPI. However, the models can be used independently.

TMap offers advantages over more traditional testing approaches, in particular the flawed and dated V Model. It is readily scalable, it is conceptually sounder. and it can accommodate usability testing (and other types of non-functional testing) more easily.

Usability Testing

Ultimately, the development of high-quality, usable systems requires usability engineering techniques to be built into the development processes of an organisation. However, it would be difficult, or impractical, for many organisations to make such a radical change in practices in the short term. This would be a lengthy and potentially expensive exercise.

In the short to medium term it is possible to obtain significant and cost effective improvements by adopting cheap, limited improvements.

My preferred approach would be to introduce elements of Nielsen's Discount Usability Engineering and Constantine & Lockwood's Collaborative Usability Inspections.

This entails cheap prototyping (possibly paper-based), early direct user involvement, evaluation of alternative designs against usability criteria, and more formal inspections of interface designs at an early stage, before the design is set in stone.

Such evaluation techniques can be used to review existing applications as well as assessing new applications while they are being developed.