Martin luther king . Don’t pretent to be as big as he is. Also don’t want to perform an act/speech about the power of testers .. The swedish ladies did better in there fun session about supertesters, than I could ever do. But I have a dream …. About testers considered as equal as other project members. I was wonder … why ….. The added value of testing is still discussed within projects. The introduction of risk based testing improved the value of testing, but we still can not answer the question about what depth of testing is the most efficience depth …. We often stop testing when we are out of budget or time … but did we achieve the best quality level? So the question is not whether we should test of not (we already convinced the project/organisation of the need for testing) … but what is the optimum balance / most optimum value of testing.
- In the top list of quality gurus next to Deming and Shewhart (Plan do check act cycle), Philip Crosby (Quality is free) - Pareto principle: by doing the job 20%, 80% of the advance of the entire job can be created. Or 80% of the problems are produced by 20% of the causes - Six sigma: performance improvement methodology. Toolbox. Six sigma levels represents 3,4 defects per million opportunities (Five sigma 233 defects) - Lean approach: produce what is needed and avoid wastes. 8 wastes: (overproduction, waiting, transport, extra processing, inventory, motion, defects, waste of human potential)
Histogram. Lower specification Limit, Higher specification limit. 68.27 percent of any normal population lies between + or – 1 sigma of the average. Sigma is the standard deviation. 95,45 between 2 sigma …. Etc. Define: scope, critical to quality, cost of poor quality, process map Measure: Identify process variables, current measurement system Analyse: Identiy causes of variation, conduct hypothesis testing Improve: identify possible solutions, verify solutions Control: Implement and sustain
Reactive QA maturity (far left block) On this level mostly corrective measures are present. Failures are found in the production-environment after going life. Next to the direct costs of correcting the failures the costs of marketing to correct the damage/decreased image are also high. The total cost of quality is much too high and can be reduced by 30-50 percent. The organisation/project only reacts on failures that occur. 2. Passive QA maturity (corrective and detective measures at a late stage) The focus within the level lies on finding the failures on the right side of the V-model during dynamic testing (often during the acceptance test). Testing is present but often only in a unstructured and ineffective way. Almost no preventive measures are present. 3. (Over)active QA maturity (far right block) Lots of preventive and detective measures are taken. Lots of test types are present and less defects are found after going life..... But testing and QA are too expensive. Finding a defect will take more time than correcting them. The damage occurred by the remaining defects is almost null. So: Too much detective and preventive measures in relation to the corrective measures. The total costs of quality can be reduced by reducing the number of preventive and detective measures (or by defining more efficient measures). 4. Balanced QA maturity The level in which the right balance between the quality measures is reached and proven via the collected metrics.
High lead time to train new testers.
Th14 - How Much Testing is Enough? Achieving the Optimum Balance 2 december 2010
Lean approach – avoid wastes Waste Examples of translation to testing Overproduction Good enough testing Waiting Minimize gaps between cause and detect of a problem Transport Deployment of software / Effective Test data and infrastructure management Extra processing Avoid extra testruns and “ Unconscious” redundancy of test Inventory Avoid to much test capacity / Clear and effective reporting Motion Monitoring the total process Defects Prevent instead of detect or correct / Standardize and Simplify human potential Effective use of human capabilities