what keeps me up at night
play

WHAT KEEPS ME UP AT NIGHT Intro Stats in the 21 st Century Data - PDF document

Intro Stats in the 21 st Century 10/22/2016 WHAT KEEPS ME UP AT NIGHT Intro Stats in the 21 st Century Data Scientists teaching our course Students think that Statistics is irrelevant for their lives and work Students think that


  1. Intro Stats in the 21 st Century 10/22/2016 WHAT KEEPS ME UP AT NIGHT Intro Stats in the 21 st Century • Data Scientists teaching our course • Students think that Statistics is irrelevant for their lives and work • Students think that Statistics is Dick De Veaux essentially univariate Williams College USCOTS 15 • We continue to change the course around the edges May 28, 2015 - With Resampling the exception 5/28/15 USCOTS 15 THE PROBLEM WHAT DO I WANT THEM TO TAKE AWAY? • The Idea that Statistics is • We teach the wrong stuff - Relevant - Intutive • We teach it the wrong way - Cool • We teach it in the wrong order - A Powerful Method for Solving Problems in the World • Armed with a Healthy (and Informed) Skepticism for • I don’t have any answers – but I’ll keep asking the questions - Data Quality - Models - Inference and Conclusions 5/28/15 5/28/15 HOW TO GET TO THE COOL DIAMONDS STUFF? • Start with it • Four C’s: • Introduce models early - Carat, Color, Cut, Clarity - Motivate univariate questions from more complex models - Socratic method • What to leave out - Mathematics of Sampling distributions • We just need the concept - Getting too under the hood with methods - They don’t need to “produce” everything – especially in the first course 5/28/15 5/28/15 2016‐DeVeaux‐USCOTS‐Opening‐Slides.pdf 1

  2. Intro Stats in the 21 st Century 10/22/2016 COLOR PRICE BY COLOR HOW MUCH IS A FIREPLACE HERE’S WHY WORTH? The Data - 1729 randomly selected homes from the county records of Saratoga New York about 2005. - Conclusions are only about the Northeast… Starting Univariate PRICE BY FIREPLACE? No 740 $174,653 Yes 988 $239,914 Difference $65,620 2016‐DeVeaux‐USCOTS‐Opening‐Slides.pdf 2

  3. Intro Stats in the 21 st Century 10/22/2016 WAIT A MINUTE Relationship with size WHAT IF WE FIT TWO LINES? WHAT IF WE LET THEM GO? Difference now is $5567 THE TITANIC WHY NOT? Ticket Class F Crew First Second Third Total M | Lived 212 202 118 178 710 Survival Died 673 123 167 528 1491 Adult 3 1,2,C Child Total 885 325 285 706 2201 Survivors Non-Survivors 2 or 3 1 or Crew 3 1 or 2 46% 93% Class Crew 1st 14% 27% 100% Crew First 23% 33% Second Third You’ve cracked the resampling egg, why not enjoy the omelette? 2016‐DeVeaux‐USCOTS‐Opening‐Slides.pdf 3

  4. Intro Stats in the 21 st Century 10/22/2016 WHERE ARE WE? • Introduce models early - Motivate univariate questions from more complex models - Not the other way around! • Use complex questions and models to drive topics • Statistics is more than a collection of tools - Which ones are really important? • How much should we ask them to “produce”? 2016‐DeVeaux‐USCOTS‐Opening‐Slides.pdf 4

  5. Intro Stats in the 21 st Century Dick De Veaux Williams College USCOTS 15 May 28, 2015

  6. WHAT KEEPS ME UP AT NIGHT • Data Scientists teaching our course • Students think that Statistics is irrelevant for their lives and work • Students think that Statistics is essentially univariate • We continue to change the course around the edges - With Resampling the exception 5/28/15 USCOTS 15

  7. THE PROBLEM • We teach the wrong stuff • We teach it the wrong way • We teach it in the wrong order • I don’t have any answers – but I’ll keep asking the questions 5/28/15

  8. WHAT DO I WANT THEM TO TAKE AWAY? • The Idea that Statistics is - Relevant - Intutive - Cool - A Powerful Method for Solving Problems in the World • Armed with a Healthy (and Informed) Skepticism for - Data Quality - Models - Inference and Conclusions 5/28/15

  9. HOW TO GET TO THE COOL STUFF? • Start with it • Introduce models early - Motivate univariate questions from more complex models - Socratic method • What to leave out - Mathematics of Sampling distributions • We just need the concept - Getting too under the hood with methods - They don’t need to “produce” everything – especially in the first course 5/28/15

  10. DIAMONDS • Four C’s: - Carat, Color, Cut, Clarity 5/28/15

  11. COLOR

  12. PRICE BY COLOR

  13. HERE’S WHY

  14. HOW MUCH IS A FIREPLACE WORTH? The Data - 1729 randomly selected homes from the county records of Saratoga New York about 2005. - Conclusions are only about the Northeast…

  15. Starting Univariate

  16. PRICE BY FIREPLACE? No 740 $174,653 Yes 988 $239,914 Difference $65,620

  17. WAIT A MINUTE

  18. Relationship with size

  19. WHAT IF WE FIT TWO LINES? Difference now is $5567

  20. WHAT IF WE LET THEM GO?

  21. THE TITANIC Ticket Class Crew First Second Third Total Lived 212 202 118 178 710 673 123 167 528 Survival Died 1491 885 325 285 706 2201 Total Survivors Non-Survivors Class Crew First Second Third

  22. WHY NOT? F M | Adult 3 1,2,C Child 2 or 3 1 or Crew 3 1 or 2 46% 93% Crew 1st 14% 27% 100% 23% 33% You’ve cracked the resampling egg, why not enjoy the omelette?

  23. WHERE ARE WE? • Introduce models early - Motivate univariate questions from more complex models - Not the other way around! • Use complex questions and models to drive topics • Statistics is more than a collection of tools - Which ones are really important? • How much should we ask them to “produce”?

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend