Understanding the Dynamics of Test-driven Development

B1e9f9e8d56bc254052905c2e49b5606?s=47 Davide Fucci
December 10, 2014

Understanding the Dynamics of Test-driven Development

Doctoral Symposium of the 15th International Conference of Product Focused Software Development and Process Improvement. Helsinki, Dec. 2014


Davide Fucci

December 10, 2014


  1. PROFES2014  Doctoral  Symposium   Davide  Fucci   Supervised  by  Burak

     Turhan  &  Markku  Oivo   10.12.2014   Understanding the Dynamics of Test-driven Development
  2. Better  software  in  less  time

  3. Test-­‐driven  development  in  a  nutshell

  4. Less    defects

  5. Productive   developers

  6. “A  1998  survey  found  that   54.4%  of  Icelanders  claimed

      to  believe  in  elves”   Wikipedia
  7. “40%  of  developers   employ  TDD”     State  of

     Agile  Survey  2013
  8. Studies  with  a  high  rigor   and  relevance  show  clear

     results   for  improvement  in  external  quality,   which  seem  to  come  with  a  loss  of   produc?vity     http://dx.doi.org/10.1016/j.infsof.2014.01.002.
  9. The  results  indicate   that,  in  general,  TDD  has  a

      small  positive  effect  on  quality   but  little  to  no  discernible   effect  on  productivity . http://dx.doi.org/10.1109/TSE.2012.28
  10. The  effects  of  TDD   s?ll  involve  many  unknowns.  

    Indeed,  the  evidence  is  not  undisputedly   consistent  regarding  TDD’s  effects  on  any  of   the  measures  we  applied:  internal  and   external  quality,  produc?vity,  or   test  quality http://shop.oreilly.com/product/9780596808303.do
  11. Much  of  the   inconsistency  likely  can  be   aIributed

     to  internal  factors  not   fully  describe  in  the  TDD  trials http://shop.oreilly.com/product/9780596808303.do
  12. None
  13. Do  developers  conform  to  the  TDD  cycle?   How  does

     conformance  impacts  quality  and  productivity?
  14. None
  15. Besouro:  a  tool  for  automated  TDD  recognition   https://github.com/brunopedroso/besouro

  16. Conformance   to  the  TDD  cycle   matters  only  for

      quality Conformance  does   not  matter  at  all Students   follow  the  TDD  cycle   most  of  the  time http://dx.doi.org/10.1145/2652524.2652526 http://dx.doi.org/10.1145/2601248.2601272
  17. How  about  SKILLS  and  EXPERIENCE? Not  really Professionals   conform

     to  the  TDD  cycle   most  of  the  time  
  18. How  the  TDD  cycles  look  like?

  19. None
  20. How  the  process  sub-­‐dimensions  impact  quality  and  productivity?

  21. None
  22. Dec  2014 Dec  2015 May  2015

  23. THANK YOU! davide.fucci@oulu.fi