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

Maximilian Duesberg - The Data is Clear - But Humans are not

Maximilian Duesberg - The Data is Clear - But Humans are not

What we all love about data is it's objective nature. It's not moody, it doesn't create drama, it is clear in the most basic sense. However, until robots will take over the world, it's still humans who will be sourcing, manipulating and interpreting this data. Unfortunately, humans are cognitively biased in a systemic way. This flawed perception of data leads to misinterpretation which can, depending on the severity of the situation, have dramatic consequences. In this talk we will discuss the most common biases in data analytics, which tactics work to avoid these biases and eventually how looking at data correctly can even make you happier in life.


August 07, 2023

More Decks by MunichDataGeeks

Other Decks in Science


  1. Agenda Cognitive Biases Examples of Biases in Data Analytics Tactics

    to avoid Biases Digression – How the right look at data can make you happier
  2. What are Cognitive Biases? Mental shortcuts that our brains use

    to process information and make decisions quickly. Help reducing complexity and protect from cognitive overload. Automatic Subconscious
  3. Availability Bias “The tendency to rely on information that comes

    readily to mind when evaluating situations or making decisions.”
  4. Technical Availability Bias – Website Tracking • Low cookie consent

    doesn’t allow valid hypothesis on whole population of users. • Patterns in cookie consent further skews data. • Opt-in users not representative group.
  5. Cognitive Availability Bias – TV Campaign 0 0.005 0.01 0.015

    0.02 0.025 0.03 0.035 0.04 0.045 0.05 Pro 7 N24 DMAX Comedy Central Pro 7 MAXX VIVA Sky Response Rate
  6. How to Avoid Availability Bias • Be aware of the

    effect. • Avoid quick decisions. Awareness. • Promote in group brainstormings. • Seek input from others. Diversity.
  7. One of the most holistic Biases in Data Analytics. Can

    influence all parts of research. Research question. “Do we have more sales on weekends?” vs. “Are there certain dates with increased sales?” Interpreting the data. Subconsciously looking out for data that supports sales on weekends are stronger. Creating follow up hypothesis. Which business decisions should be taken to leverage the (false) conclusion?
  8. How to Avoid Confirmation Bias • Awareness • Proactively seek

    out for evidence disprooving your hypothesis. • Engage a “Devil’s Advocate”. • Promote open discussions about findings.
  9. Summary - How to Avoid Bias in Data Analytics Structural

    • Constant reminders • Open company culture that encourages discussion. • Regular training Awareness • Acknowledge existence of bias. • Raise awareness in peers. Diversity • View data from diverse viewpoints. • Engage with people from different backgrounds.
  10. Negativity Bias “Tendency to pay more attention to negative information

    than to positive information. Here, more weight is given to negative experiences over neutral or positive experiences.”
  11. Avoid Negativity Bias • Change the filter and review the

    data. • Proactively focusing on positive events. ➢ Gratitude Journal • Improves wellbeing and optimism significantly. • Other findings: • Test groups exercised more. • Improved sleep. • Less visits to physicians. • Lower blood pressure. • Less inflammation biomarkers.