Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Thinking Fast and Slow for Testers: Cognitive Distortions Affecting Testing

Thinking Fast and Slow for Testers: Cognitive Distortions Affecting Testing

In his book Thinking Fast and Slow, Nobel winner Daniel Kahneman introduces two mental systems, one that is fast and the other slow. Together they shape our impressions of the world around us and help us make choices. System 1 is largely unconscious and makes snap judgments based upon our memory of similar events and our emotions. System 2 is painfully slow, and is the process by which we consciously check the facts and think carefully and rationally. However, System 2 is easily distracted and System 1 is wrong quite often. System 1 is easily swayed by our emotions. Examples he cites include the fact that pro golfers are more accurate when putting for par than they are for birdie (regardless of distance), and people buy more cans of soup when there's a sign on the display that says "Limit 12 per customer." In this session we will, through interactive games, learn how our minds can affect our abilities to investigate, observe, and recall events. We’ll discuss some strategies to minimize our minds erroneous impacts to those abilities.

Joseph A. Ours

April 16, 2014
Tweet

More Decks by Joseph A. Ours

Other Decks in Business

Transcript

  1. Introducing Joseph Ours Director of Quality Assurance and Testing Services

    More than 15 years of IT experience spanning several industries and roles Successfully managed multi-million dollar initiatives while introducing innovative solutions for Fortune 1000 companies, resulting in improved quality, increased client satisfaction, cost avoidance, on- time delivery, and establishing a true team culture @justjoehere [email protected] www.linkedin.com/in/josephours
  2. Bug Hunting Skills Investigate Search Probe Explore Interrogate Observe Detect

    Discover Monitor Study Recall Think of Recollect Summon Cite
  3. But each observations informs and changes our expectations based on

    the differences in expectations Working Together What we see or don’t see Is often based on what we expect to see And is stored in memory based on closest associative other memory And affects what we see in the future and what we tell others we observed
  4. Remember I said your mind is replete with biases Don’t

    take my word for it, let’s go through each key area and perform some exercises to see biases in action
  5. Did you guess? 2 4 6 8 10 12 N=n*2

    linear sequence n=(1,2,3,4,5,6)
  6. Did you consider? 2 4 6 10 14 22 N=n*2

    Prime sequence n=(1,2,3,5,7,11)
  7. Improve your investigative skills Change your search pattern Search for

    alternate answers and/or paths Defend your answer against antagonist Ensure gaps are not unduly filled in
  8. Recall Skills Get Ready to Pick a Card from Those

    Shown Memorize the Card Let the Fun Begin!
  9. Improve your recall skillls Acknowledge your recall CAN be faulty

    Record as you go, keep a journal Or use a recording tool Chunk Tests into groups 5-7 Reduce cognitive load
  10. Recap • Our biases affect all aspects of testing •

    What we look for, what we observe, and what we miss • How we see it • How we recall it
  11. Minimizing Bias Investigation • Change your search patterns • Look

    for alternate answers • Defend your answer against antagonist • What your fill-ins Observation • Focus/De-focus • Time box your activities • Listen to intuition • Don’t over rely on documentation, think Recall • Limit what you have to recall • Use tools to assist • Record/Journal as you go • Reduce cognitive load Acknowledge Biases Exist
  12. Q&A

  13. Bias with Metrics! You have – Urn with 1 red

    and 10 white marbles – Urn with 8 red and 100 white marbles Which urn will you use to try draw a red?
  14. Law of Small Numbers A false conclusion based on a

    logical assumption with insufficient evidence • Metrics • Frequency of occurrence of a defect • Actions taken to reproduce a defect • Test Approach • Test sample size is too small to validate functionality • Equivalency class is more reliable than it is • Pairwise testing covers more than it does Where testers may encounter this
  15. Meta Example • See a crime (car running from cops,

    crashes into another car), hear others give their accounting – You believe as others • An investigator “leads” a question (how fast was car going when smash/bump) – Your memory is altered • Telling friends of crime, suggests car driver was drunk – despite driving straight on road • Conveying driver was guilty – based on stereotype of runners