Estimation models to effective guage the accuracy of testing efforts?



  • Are there separate Estimation models to effective guage the accuracy of testing efforts? Keyword-driven based estimation model Data-driven based Modular-driven Hybrid-driven Component-based Please clarify



  • No doubt there are academic papers that look at least some of those test efforts, but you should be reluctant to generalize them to your own work. Here are a couple of some reasons why: There are lots of confounding factors that impact the accuracy of an estimation model, e.g. adherence to the methodology under study, experience level of the testers who perform the estimation, experience level of the testers doing the actual testing, quality of the system under test, and amount of churn during the test process. Any worthy academic paper will include a section on potential confounders and their potential impact on the conclusions, but you should not assume they list them all, or are even aware of them all. There are implicit biases in academic literature, e.g. publication bias (a paper that confirms what we already believe is less likely to be published than a paper with a surprising result). If you want to generalize a study to your own organization, ask yourself whether their organization is sufficiently like your own.


Log in to reply
 

Suggested Topics

  • 2
  • 2
  • 2
  • 2
  • 2