How to Think Like a Data Scientist (No Degree Needed)

About this video

• Video Title: How to Think Like a Data Scientist (No Degree Needed)
• Channel: Stephen Petro
• Speakers: Stephen Petro
• Duration: 14:50

Overview

This video explains how to think like a data scientist by breaking down complex information into three core skills: data interpretation, model interpretation, and normative reasoning. The presenter aims to equip viewers with critical thinking tools to discern facts from misinformation and avoid manipulation, regardless of their formal educational background.

Key takeaways

• Skill 1: Data Interpretation: This involves asking critical questions when presented with statistical claims to identify potential manipulation or bias. Key questions include: Are the samples randomized and of sufficient size? Are dollar amounts adjusted for cost of living? Is income the sole measure of wealth? Are observed differences statistically significant? Is the data subject to Simpson's paradox?
• Skill 2: Model Interpretation: This focuses on understanding the underlying models used to make predictions or describe phenomena. Critical questions include: Which model is being used? What are its mathematical constraints? Is the model causal or correlational? What are the normative assumptions embedded within it? This skill helps in identifying how different models can lead to contradictory conclusions.
• Skill 3: Normative Reasoning: This is about understanding the value judgments and motivations behind data presentation and model creation. Key questions involve understanding the creator's motivations, what variables were included or excluded and why, and how the model or theory got its name, as names can carry implicit biases or assumptions.
• Avoiding Manipulation: By applying these three skills and asking the right questions, viewers can avoid being easily fooled by biased or misleading information, leading to clearer thinking and greater mastery over their own minds.