FRC Presentation Teaching Quantitative Reasoning Jason Hecht (ASB) November 4, 2015 1 Despite My Best Efforts…Top 7 Things Students “Learn” in Econometrics… 1. How to juggle numbers without moving (or thinking). 2. Theory is a “special case” of reality. 3. Correlation ≠> Causation (but my R2 is still 0.99!) 4. Always rationalize a contradictory result, but… 5. Never support or highlight my own insight. 6. Always put an illogical result to good use! 7. You will never see the words “heteroscedasticity” as a clue in a NYTimes crossword puzzle. 2 Venue: Econometrics Term Paper Each student must choose a different problem, issue or topic that either: 1. Updates/replicates an existing model 2. Expands/modifies an existing model Students must submit drafts of each section of the paper: i) Introduction, ii) Literature Review, iii) Model Specification & Preliminary Estimation, iv) Problem Diagnosis & Correction, v) Conclusion & Implications. 3 Objectives 1. 2. 3. 4. 5. 6. The joy of modeling a hunch! To learn how to learn from a “wrong” statistical result. Correlation ≠> Causation, but econometrics may provide the “best evidence” of a statistical association. Learning econometrics ≈ Learning to play the drums: => “4-limb” intellectual interdependence: logical thinking, estimation skills, inferential skills, writing skills. Learning how to “write to the numbers” even when they don’t support your hypotheses. Forensic learning: what the hell was I thinking & doing five weeks ago?! 4 How do I teach reasoning… 1. “Math anxiety” coach & symbol interpreter: you learned all this in 9th grade!!! 2. “Slow pitch” the algebra: work backwards from the data. 3. Create interesting in-class exercises by challenging “self-evident” propositions. 4. Closely edit their writing and force rewrites. 5. Use an entire class for one-on-one consults. 6. Be accessible!!! Instantaneous e-mail response! 5 Teaching challenges… 1. Interpretation of a statistical result or test. 2. How to write coherently and cogently. 3. How to pattern their paper after a previous study (“learning by mimicry”). 4. Why the algebra works but the numbers don’t. 5. Statistical subtlety: all roads should not lead to the highest R2 6. Why do I lose 1/4 of my class by the 1st exam? 6