Human Judgment, Decision Making, and Bias

Overview

• Humans make many judgments and decisions daily, varying in importance and consequence.
• Judgment and Decision Making (JDM) involves selecting choices from alternative actions after deliberation.
• JDM includes predicting possible consequences and evaluative reactions.
• Using data improves decision-making and judgment, central to HR analytics.

Rational Decision Making Model

• A rigorous approach ensuring better decisions, especially in high-stakes situations.
• Scientific Process in Decision Making:
  1. Formulate a hypothesis
  2. Design a study to test the hypothesis
  3. Collect data to test the hypothesis
  4. Analyze and interpret data
  5. Report findings and possibly repeat the cycle
• Steps in Rational Decision Making:
  1. Identify the problem
  2. Establish decision criteria
  3. Weigh or rank decision criteria
  4. Generate decision alternatives
  5. Evaluate decision alternatives
  6. Choose the best alternative
  7. Implement the decision
  8. Evaluate the decision

Example in HR Selection Context

• Problem: New employees not performing well
• Decision Criteria: Validity, cost, and applicant reactions
• Generating Alternatives: Structured interviews, work samples, personality tests
• Evaluation: Validation designs, cost analysis, and applicant surveys
• Choosing and Implementing: Select best predictor, e.g., work sample
• Evaluation: Cross-validate with a new sample, identify issues, and possibly redesign

Human Biases and Errors

Common Biases

1. Anchoring and Adjustment Bias: Focusing on certain information over others, possibly due to recent exposure.
2. Availability Bias: More readily available information is attended to more.
3. Escalation of Commitment Bias: Continuing down a path due to prior investment, despite conflicting evidence.
4. Hindsight Bias: Post-event, perceiving outcomes as more predictable than they were.
5. Correlation vs. Causation: Assuming co-occurring events cause each other without evidence.
6. Sampling Bias: Non-representative sampling leading to biased inferences.
7. Overconfidence Bias: Overestimating the ability to predict future events, relying too much on intuition.

Mitigation Strategies

• Use data to inform decisions and predictions.
• Be cautious and aware of predictive model limitations and potential errors.
• Ensure representative sampling in data collection.
• Apply rational decision-making models in high-stakes scenarios.

Conclusion

• Rational decision-making models help make better decisions in high-stakes settings.
• Be aware of common human biases and errors in decision-making processes.
• HR analytics plays a crucial role in improving judgment and decision-making.

Thank you.