What's Nicky Learning? + Design Doodles (Nov 2019) (Patreon)

(total reading time: 18 minutes)

Welcome to the first-ever, Patreon-exclusive WNL?+DD post!

What's Nicky Learning: emotion in learning, the physics of sound, drawing from imagination, juggling

Design Doodles: adventures with anxiety inside-out design, voice gender trainer 5000

- - - - - - - - - - - - - - - - - - - - 

🤔 Emotion in Learning

So earlier this year, I learnt my whole career was based on a lie. As one does.

I make games to help folks learn by doing. People learn all kinds of complex mechanics entirely "by doing" in games like Portal, right? And it's fun! Of course beginners learn better from pure exploration, rather than being railroaded through step-by-step instructions.

Anyway, this idea's been experimentally tested, and it's replicatably false.

But here's where this goes from disappointing to bizarre: "active" learning-by-exploring is ineffective for beginner students, but very effective for advanced students! And "passive" learn-by-step-by-step-instruction is effective for beginners, but ineffective for the advanced.

This paradox is called The Expertise Reversal Effect.

(slide from a talk I'll post online soon!)

But maybe I shouldn't be surprised. Trying to bench-press 200lb would kill a beginner weightlifter, but it's good for an advanced weightlifter. Like physical load, there's cognitive load: step-by-step instructions are a "small" load, while pure exploration is a "heavy" load.

According to Cognitive Load Theory, good education is giving a learner a fitting "load": one that's slightly beyond their comfort zone. (See: Zone of Proximal Development, scaffolding)

Oh, as for "learning-by-doing" in games like Portal, it turns out the designers add tons of hand-holding guidance – they just keep it invisible so you can feel smart.

. . .

So if you're an educator or someone who "communicates through multimedia" like me, what should we do?

Richard Mayer has applied Cognitive Load Theory to create 10 empirically-based Principles for Multimedia Learning. These principles include: cut out the irrelevant "fun" fluff, organize things step-by-step, signal the important parts, etc. This doesn't mean being dry, just clear and minimalist.

But, still...

I like the fun fluff.

Leonard Susskind's The Theoretical Minimum is a technical introduction to physics by one of the world's top physicists, yet each chapter is sprinkled with silly dad jokes. Cognitive Load Theory predicts these to be "extraneous load" that would hurt learning, but I think the dad jokes are necessary.

Consider Yerkes-Dodson's Law. For complex tasks – like learning a new thing – too little or too much stress leads to poor task performance, but there's a "just right" amount of stress for the best task performance:

Back to Susskind's books: tensor products are pretty stressful, so dad jokes help bring the stress level back down to the optimum. Same technique is used in other technical-but-fun (fun-but-technical?) stuff like 3Blue1Brown, Gödel Escher Bach, Mathematical Games, etc.

That's the big gap in the original version of Cognitive Load Theory: it says nothing about emotion... even though neuroscientists & psychologists have known for decades that the "cognition-emotion divide" is bull! (See: Lazarus 1991, Descartes' Error)

So this month, I found 4 helpful papers about emotion in learning!

Paper #1)

I'm very pleased Um et al 2012 provides empirical support for my approach to teaching: "slap a bunch of silly faces on it".

Obviously, making something more "fun" makes a learner more motivated to keep going. But this study found something stranger: making something more "fun" improved learning outcomes directly, not through motivation alone.

Three possibilities why:

1. The Yerkes-Dodson thing I mentioned earlier. 
2. Memories stick better when they're emotional. (including painful memories)
3. Being in a relaxed mood actually "opens your mind" more. (Imagine how close-minded people get when they're scared!)

But what about all the other research showing that interesting-but-irrelevant "seductive details" damage learning? And in the "educational games" practice, it's frowned upon to create "Chocolate-Covered Broccoli", coz kids can smell the bull from miles away.

The difference between the chocolate-covered broccoli and the Um et al paper is that the extra "fun stuff" is actually relevant. Thus, they don't misdirect attention – in fact they help focus it! The cute faces are only on the important biological agents, while the unimportant background & buttons stay relatively boring.

(Same with Susskind's The Theoretical Minimum – it's not filled with jokes, just sprinkled with them. There's only 1 pun per 10 pages, and each one is relevant to the learning material. More or less.)

But still, this technique somehow feels like... cheating? Lymphocytes aren't that cute in real life. Is there a way to get learners to feel positive emotions about the material without resorting to cute doodles?

Yes! Let's get curious, about an emotion called curiosity.

Papers #2 and #3)

Monkeys will solve puzzles for food. So far, so behaviourist.

But: Monkeys will solve puzzles even without a food reward. In fact, introducing food as a reward made the monkeys learn worse. Understanding the unknown is already a biologically fundamental reward in itself.

Same for us monkeys. Contrary to the "logical thinking means being dispassionate" myth, the world's top mathematicians and scientists are driven by a specific emotion: curiosity. (Also desperation over grant funding)

But what, then, causes curiosity?

Kang et al 2009 was a neuroscience study that found three things:

1. being more curious led to more activation in reward & memory neural circuits,
2. being more curious led to better recall, and
3. we're most curious when we have a medium-level of confidence in our answer.

Like the Yerkes-Dodson curve, it's another upside-down U shape. When the answer's obvious, well duh we're not curious – it's "too comprehensible". But when we have no idea what the answer even could be, we also don't care – it's "too complex".

But in-between, when you balance complexity and comprehensibility, then monkeys get curious! 

NOTE: Cognitive Load Theory can't predict a such thing as "too comprehensible"! This leads to different implications for science communication: instead of giving the information in a logical order bridging all the gaps ("blah blah blah here's the formula for orbits"), you should first reveal a medium-sized gap in the learner's knowledge ("if gravity keeps the moon near us, why doesn't the moon fall down?"), to stimulate the learner's curiosity! Show the gap, then build the bridge.

Silvia 2008 reports the same thing. I just want to include this 🌶 spicy 🌶 paragraph of his:

College textbooks are an intriguing example. The typical textbook wants to engage students’ interest, so it sprinkles each chapter with irrelevant quotes, cartoons, contrived stock photos, and random stories from the authors’ distant childhoods. But diverting attention from the text’s main points isn’t the same thing as making the text’s main points interesting.
[...]
If interest comes from seeing something as new and comprehensible, then people who want to evoke interest should try to enhance both complexity and comprehension.

That is: don't coat the broccoli in chocolate.

Paper #4)

When I started writing this post, I intended to include just the 3 papers above. But as I hunted down my citations, I found this extra gem.

You remember above, when I said:

That's the big gap in the original version of Cognitive Load Theory: it says nothing about emotion

Turns out people have recently added emotion and motivation to Cognitive Load Theory. Moreno 2010 recaps the problems with the old theory, and summarizes a new theory: Cognitive-Affective Theory of Learning.

It proposes a loop: (Bad drawing by me)

You'll learn poorly if you're not motivated, and you'll be demotivated if you're learning poorly. But, from vicious to virtuous cycle: you'll learn well if you're motivated, and you'll be more motivated if you're learning well!

So, if you make "multimedia learning", the trick is to boost both sides of the loop:

(💥💥💥 THIS IS THE SUMMARY SECTION 💥💥💥)

• 1) Improve Cognition with:
  
  • Cognitive Load Theory – remove distractions, organize info, signal what's important, etc
    
  • Zone of Proximal Development – keep the cognitive load difficulty just above the learner's comfort zone.
    
• 2) Improve Motivation with:
  
  • Relevant fun – add "fun" stuff only if it highlights the important parts.
    
  • Curiosity first – show a (medium-sized) gap, then build the bridge.

(💥💥💥 / END SUMMARY SECTION 💥💥💥)

I've made lots of chocolate-coated broccoli before. A science piece with irrelevant stories, a math explainer with strained pop-culture references, a simulation with unnecessary "ooh solve the puzzle!" game-y bits.

The original Cognitive Load Theory taught me to strip the chocolate off my broccoli. But finding out that emotion directly improves learning? That taught me I need to season & roast those veggies, to bring out the natural tastiness that's already in the broccoli. As the monkeys showed us, there's a natural motivation already in understanding the unknown.

There's the junk food of bad edu-tainment, and the Nutra-Loaf of dry academic papers. The trick to sticking to a healthy diet is to cook stuff that's both delicious and nutritious.

- - - - - - - - - - - - - - - - - - - - 

(I'm planning to publish the above as a standalone essay – what parts were clear, or not? Any other links/papers I should read? If you're an educator, are the summary tips useful? Thank you!)

- - - - - - - - - - - - - - - - - - - - 

🔊 The Physics of Sound

While doing research for Voice Gender Trainer 5000, I realized I don't understand how voice works.

Thankfully, upon further study, I realized I don't understand how sound works.

Problem #1: When you hear your roommate's dubstep through the wall, you hear low-frequency bass better the high-frequency wubs. Why? The explanation "because longer wavelengths can wiggle through the wall better" may make sense for electromagnetic waves, but not for physical air waves. And "because higher frequencies have higher energy" makes anti-sense. If I punched a wall, a higher-energy punch would go through better than a lower-energy punch!

Problem #2: Why does a paper megaphone amplify sound? I thought: "because it directs the sound better". But if that's true, why are megaphones cones, not tubes? Tubes direct sound better! And wait, if a small-to-big cone (megaphone) amplifies sound, why does a big-to-small cone (ear, stethoscope) also amplify sound?!

Problem #3: While trying to find the answer to #2, I learnt that a tube wouldn't work because the sound wave would reflect off the small open end. In fact, if you time the reflections just right, you can get a "standing wave" inside an open tube. Ok. So. How the heck can a sound wave, travelling through air, bounce off... the exact same air?!

After a week of research and/or screaming, I found all these brain-breaking questions have the same answer:

"Acoustic Impedance."

...

Buggered if I know what that actually means. Let alone why "acoustic impedance" should be different for different frequencies, or different for the same air just outside an open tube. (This website suggests an analogy to electrical impedance, which I also don't understand. Also, Z = p/U ≠ z = p/u? Wow physics notation sucks.)

These What's Nicky Learning? posts aren't always going to have answers.

Sometimes, they're just going to reveal a medium-sized gap in my current state of knowledge. And as explained earlier: medium-sized gap → curiosity → learning.

So if any of you can help me understand acoustic impedance, my broken brain would be eternally grateful!

- - - - - - - - - - - - - - - - - - - - 

✏️ Drawing From Imagination

As you saw in Adventures With Anxiety, my drawing skill is currently between "enhanced stick figure" and "a red wolf that 50% of players think is a fox".

I've read Drawing on The Right Side of the Brain before, and – sketchy pop-neuroscience aside – it's one of the best how-to books I've ever read! Look at how one student of the author's improved over five days:

The typical beginner's problem is that you draw what you know, rather than what you see. For example, you "know" that a chair has a straight back, a square seat, and four legs that all touch the ground. So you draw that...

...and it looks like crap. But if you force yourself to "forget what you know" – and the book offers tips like focusing on just the edges, or just the shapes, or just the negative space between the shapes, or even drawing upside down – you get...

...something decent! So if you want to draw from real life, I highly recommend this book. 👍

However...

It's not that helpful if you want to draw from imagination – where you don't have a real-life reference to see, and you must draw what you "know".

So that's what I've been trying to learn: how to draw human bodies, semi-realistically*, from imagination, without reference.

(* I'm specifically inspired by Emily Carroll and Rob Guillory!)

Q: Why learn that?

A: Because the human form is a fascinating subject, and learning how to draw it forces you to appreciate the intricate beauty that's always present in the everyday

A #2: So I can draw porn

Here's my better drawing attempts:

And some worse attempts:

If you want to draw from imagination, here's The Trick, as far as I can tell:

💥 Draw from the inside-out. 💥

Remember, the problem with drawing from imagination is you have to draw what you know. So if want to not suck, you have to know a lot. So, I'm using Spaced Repetition flashcards to help me remember body part proportions & connections! (Did you know your foot is as long as your forearm?)

I'm learning from tutorials all over the web, but one resource that's helping me a lot is Proko.

Proko makes great tutorials for artists at all skill levels, but there's one tutorial of his I want to share today... because it's got the piece of advice I know I should've been taking from the start, but have been avoiding:

Deliberate Practice.

You may know "deliberate practice" as "that academic paper that was misinterpreted by Malcolm Gladwell into the you-need-10000-hours-of-practice-to-become-good bestseller myth".

"Regular practice" is what I've been doing – drawing random stuff, and either 1) not noticing my mistakes, or 2) noticing loads of mistakes but not focusing on any one specific mistake.

Deliberate practice is the art of lingering on your mistakes.

For example, in the above tutorial, Proko identified one specific mistake of his – a stiff gesture – and re-drew his drawings, fixing just that one mistake. He identified a medium-sized gap in his skill, and set to bridge it.

That's how to best improve at any skill, and what I need to do next: focus on medium-sized gaps. If I ignore my mistakes I'll never improve, but if try to fix all my mistakes at once, I'll get overwhelmed, throw my hands up, and declare "I am not good at it."

🎱 Juggling

I am not good at it.

- - - - - - - - - - - - - - - - - - - - 

Are you also trying to learn how to draw? What tutorials have helped you? Which artists do you look up to?

- - - - - - - - - - - - - - - - - - - - 

[/end: What's Nicky Learning?]

[start: Design Doodles!]

- - - - - - - - - - - - - - - - - - - - 

🐺 Adventures With Anxiety: Inside-Out Design

Stop me if this sounds familiar:

You want to create a character for a story. So you start with a bunch of traits: pink hair, leather jacket, fluent in 17 languages, parents died in a tragic laundry accident, etc...

...and you get a character that feels hollow. All surface, no core.

That's because your character was designed from the outside-in. Consider the Joker in Suicide Squad (2016). The character had all the right surface traits – insane, violent, clown makeup – but in the end, not that memorable.

Contrast that to the Jokers from The Dark Knight (2008) or Joker (2019), who were designed from the inside-out. (Hey, like drawing from imagination!) Both these Jokers had very clear values & life philosophies – heck, they even say them out loud:

The Dark Knight: “When the chips are down, these... these civilized people, they'll eat each other. See, I'm not a monster. I'm just ahead of the curve.”

Joker: “For my whole life, I didn't know if I even really existed. But I do. And people are starting to notice.”

Everything else – their actions, their dialogue, even the style of clown makeup – follows directly from these cores. That's what makes these characters feel much more real... and terrifying.

(For all the controversy around Joker, the criticism isn't that the character lacks motivation – it's that his motivations are too real)

I learnt the "inside-out" approach from Robert McKee's Story, and it's the SECOND-most valuable storytelling principle I know! And I do mean principle, as in "F=ma" or "Energy is conserved", not just a nifty trick.

So, when I wrote Adventures With Anxiety, I started with the characters' core values & beliefs first! From my sketchbook (plz forgive messy handwriting):

And the conflict between their values & beliefs drives the whole story.

(I'll admit a mistake: the Party Host character wasn't "organic", because I never gave them a coherent motivation. I wrote them from the outside-in: "I need a sexy human antagonist so here we go")

Oh, and the FIRST-most important principle of storytelling I know?

🎥 Therefore & But 🎥 (5 min video)

Bad story: "this happens and then that happens and then this happens"

Good story: "this happens therefore that happens but this happens".

For a character-driven story, the Therefores and Buts should all be caused by the characters' core values. Example from Adventures With Anxiety: (I outlined Act II by writing & re-arranging story beats on cut-up index cards. Black ink is human's point of view, red ink is anxiety wolf's. My shorthand is ∴ = therefore, ! = but)

(This principle is so strong, I even use it in my non-fiction, as Tony Zhou does in the above video! "Therefore & But" is the difference between a listicle and an essay with actual flow.)

But Adventures With Anxiety is an interactive story. Therefore I have to connect multiple possible story-paths with Therefores and Buts! Here's how I diagrammed them:

(A fan made a much better flowchart of the game's story-paths here)

. . .

Y'know, I keep hearing authors say "my characters write themselves," and I think, "what controlled substance are you on?"

But apparently, this is common! According to a 2003 academic survey, 92% of the interviewed writers experience "their characters writing themselves", with published authors having stronger & more frequent experiences like that.

I never understood this until writing my game! I know this sounds weird, but sometimes I'd write dialogue, and the wolf in my head would tell me, "no I'd say it like this instead." I'm not being metaphorical. My anxiety-wolf honestly felt like a co-writer.

If you're not a fiction writer, maybe another way to understand this:

Physics simulations. (Like in a videogame!)

There's the misconception that "computers can't surprise you, they can only do what you program them to". You don't need to look at AI or procedural generation to debunk this – even a physics simulation can shock its own creator. (Famous example: Edward Lorenz tried simulating the weather, and it shocked him so much, he founded a new field called Chaos Theory)

For a physics sim, you just need to define:

1. Forces: gravity, magnetism, etc
2. Causality: how each step of the sim affects the next
3. An unstable starting state

And the simulation will "animate itself". Likewise, for a story, you just need to define:

1. Forces: your characters' core values + beliefs 
2. Causality: therefores and buts
3. An unstable starting state

And the characters will "write themselves". They'll surprise you. They'll shock you.

They may even teach you something new.

- - - - - - - - - - - - - - - - - - - - 

(I might also publish the above as a standalone essay! What's clear or not? Any extra resources I should look at? If you're a writer, how helpful do you find the Inside-Out and Therefore/But principles?)

- - - - - - - - - - - - - - - - - - - - 

🗣 Voice Gender Trainer 5000

"Eat your own dogfood." I personally want to train my voice, so I'm going to use my own tool as I design it!

Here's my messy "design doc", which looks like horror movie wall-scribblings:

Two principles I'm trying to apply in my design:

1. Cognitive Load (explained in the Emotion in Learning section)
2. Deliberate Practice (explained in the Drawing from Imagination section)

– Cognitive Load –

Bad design: Just dump all the controls on the user at the start. This is standard software interface design, and it overwhelms beginners.

What I want to try: Designing the interface like a videogame. Starts with just a few controls, then gradually "unlocks" as you progress! (With option to skip & unlock all) Honestly, I don't know why more software doesn't already do this.

– Deliberate Practice –

Bad design: Just give people a bunch of tips, and hope they actually do them and improve their skills.

What I want to try: Giving people loads of feedback, so they can identify their own specific mistakes, their medium-sized gaps, and bridge them. There'd be feedback through audio (your voice vs recordings), through visuals (your formant frequencies vs target frequencies), and over time (voice recording from Day 0 vs Day N).

This tool will also ship simultaneously with a (video?) explainer of sounds, spectrograms, and voice. I have a rough outline – I just need to tie it altogether with Therefore & But, for a nice logical flow.

- - - - - - - - - - - - - - - - - - - - 

😌 WHEW 😌

Okay wow that was a lot. And yet, all these different things had common themes:

• Cognitive Load
• Medium-sized gaps
• Curiosity
• Designing from the Inside-Out
• Deliberate Practice
• Therefore & But

And so on.

I hope this post was valuable to you – especially if you teach stuff through multimedia, are learning to draw, or want to write stories. Let me know what you found helpful, what concepts were still confusing, and what other resources you recommend I check out next!

Let's get curious together.

💖,
~ Nicky

[/ END: THE INAUGURAL WHAT'S NICKY LEARNING? POST]