S5E09 Regularized Variable Selection Methods

In this week’s episode, Patrick and Greg play with some of the basics of probability in the context of some classic, fun, and often counterintuitive examples.  Along the way they also discuss arguments with relatives, a feel for the roulette wheel, Xeroxing your butt, “The coin has spoken.”, Quantitude BooqQlub, the Bellagio Fountains, Clooney and Pitt look-alikes, the Flippier, Rosencrantz and Guildenstern, Monty Hall, Ferraris and goats, the birthday problem, how to carve an elephant, and pick-6 lotteries. 

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode Greg and Patrick are honored to visit with Yi Feng, a quantitative methodologist at UCLA, as she helps them understand classification and regression tree analysis. She describes the various ways in which these models can be used, and how these can serve to inform both prediction and explanation. Along the way they also discuss looking pensive, drunken 3-way interactions, Stephen Hawking, parlor tricks, Cartman, validation, dragon boats, anxiety, spam filters, hair loss, audio visualizations, overused tree analogies, rainbows & unicorns, rain in Los Angeles, and Moneyball.

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode Greg and Patrick talk about Simpson’s Paradox: what it is, examples of where it occurs in real life, and why we might not really need to think about it as a paradox at all.  Along the way they also discuss Apple Vision, The Simpsons predicting the future, sitting too close to the TV, low levels of radiation, Paul the octopus, Blinky, shtuff, duck duck duck, Jonathan the tortoise, batting averages, bad at fractions, Cal Berkeley admissions, GDP and happiness, exercise and heart attacks, and Double Super Secret whatever.

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode Greg and Patrick take a walk down memory lane to rediscover classical test theory,  although they revisit this through the lens of modern latent variable models. They describe how these classical methods are actually highly restricted latent measurement models and they explore how these restrictions can be relaxed and even tested. Along the way they also discuss weird dates, free Tupperware, yellow fatty beans, advice cookies, in bed, scalloped tiles, scales at NASA, important footnotes, working in your garden, being buck naked, Dark Side of the Moon, squarshing, deserving variance, and when the cops show up.

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode Patrick and Greg launch a new occasional series called Stuff You Should Know. The topic for today is regression to the mean: what the heck is it, how does it arise in every day life, and what can we do about it. Along the way they also discuss the space-time continuum, Kai Ryssdal, inflation, witches of MacBeth, the hidden curriculum, oh dang, sh*t (ummm...STUFF) you should know, SAT prep courses, the triumph of mediocrity, angstroms, blobs. reinventing history, and being patronizing. 

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode Greg and Patrick talk about confidence intervals: symmetric and asymmetric, asymptotic and bootstrapped, how to interpret them, and how not to interpret them. Along the way they also mention tire pressure gauge mysteries, conference travel reimbursement, phases of the moon, gyroscopic effects, baseball walk-of-shame, why people hate us, settling out of court, confidence tricks, Mack JcArdle, Shakespearean means, lipstick on a pig, the cat rating scale, the Miller's Tale, hot pokers, inverse hyperbolic tangents (duh), and Quantitude out-takes. 

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode Patrick and Greg have great fun talking about meta-analysis with Paschal Sheeran, a social psychologist from the University of North Carolina at Chapel Hill. He describes what meta-analysis is, what it offers, and how to apply it in your own work. Along the way they also discuss taking off into the north vs the south, flipping people off, being subtle, metaphysics, say my name, most effective tactic available, fleeing Ireland, struggling to fit in, falling off a bus, taking the piss out of your own work, opposable thumbs, descriptive descriptions, sleepwalking reviewers, the wild west, the importance of cake, living the dream, and the five stages of meta-analysis. 

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode, marking the fifth Quantitude Holiday Celebration, Greg and Patrick argue about their favorite holiday movies, including whether Die Hard counts as one or not; they then proceed to discuss several statistical ideas that also seem to Die Hard. Along the way they talk about so much gibberish that we don't even bother listing it here.

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In this week's episode Greg and Patrick explore alternative parameterizations of the SEM-based latent curve model to capture various forms of nonlinearity, some that are approximations and others that are exact. Along the way they also discuss Swifties, remastering your life, bull testicles, the world's worst RA job, Yerkes-Dodson law, show a little ankle, the St. Louis Arch, bachelorette parties, deck screws, DIY-ing a model, being a little too quiet, complete nonsense, blasting your pecs, haters gonna hate, the worst day ever, Frankenspline's monster, being left off at the third floor, and looking for a new cohost. 

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com

In today’s episode Greg and Patrick talk about regularization, which includes ridge, LASSO, and elastic net procedures for variable selection within the general linear model and beyond. Along the way they also mention Bowdlerizing, The Family Shakespeare, disturbance in the force, McNeish on his bike, Spandex, C’mon guys wait up, the altar of unbiasedness, Curranizing, shooting arrows, stepwise goat rodeo, volume knobs, Hancockizing, always angry, getting slapped, betting a chicken, mission from God, hypothetico-deductive porpoising, and letting go of truth (which you can’t handle anyway).

Stay in contact with Quantitude!

• Twitter: @quantitudepod
• Web page: quantitudepod.org
• Merch: redbubble.com