Digital signal processing, made to stick

[Preamble] If you ask someone to think of a sticky idea, a lot of times they’ll blurt out a slogan. “Wassssup!” “Just do it.” And, no question, these are sticky ideas. But because people tend to associate the notion of “stickiness” with things like slogans — i.e., short, punchy, cleverisms — they have a hard time imagining that stickiness could apply to really complex things. Stickiness is for marketers, not for engineers or scientists, or so goes the thinking.

Well, no. Chip and I may be largely to blame for this misperception — the word ‘sticky’ we’ve embraced is itself clever and vaguely marketingish. But sticky just refers to an idea that was understood, remembered, and changed something (opinions, behaviors, values). So the fact that someone is a practicing nuclear scientist means that, at some point, nuclear science concepts stuck. Which in turn means that a nuclear science teacher found an artful way to communicate really hard concepts.

Because of this backstory, we were thrilled to get a note from Andrew Singer, who teaches a digital signal processing (DSP) course at the University of Illinois at Urbana-Champaign. Professor Singer and I ran across each other as a result of a talk I gave on campus. He had read Made to Stick and shared some changes he had made to his course curriculum as a result. And, as you’ll see, slogans are not his bag. He deals with really complex topics that must be communicated to really smart people. After Chip and I got his note, we exchanged “Wow!” emails with each other. [/Preamble]

I teach a course on digital signal processing to juniors and seniors in electrical engineering at the University of Illinois. This is a course that describes the mathematics and theory behind applications like digital modems, HDTVs and MP3 players. Basically, whenever “signals”, that is information they you care about, are “processed” using a computer, cell phone, or anything that samples or “digitizes” the signals of interest, digital signal processing is used.

Needless to say, the theory that I need to teach the students makes this largely a math course for upper level engineering students. However, as you know, motivating students to learn more mathematics for mathematics sake (especially for non-math majors) is no easy task. I’ve struggled over the years in coming up with means to increase their interest and maintain this throughout the course so that I can supply them with the tools they need to be productive and successful engineers.

As a young faculty member, I used boundless enthusiasm and energy in my lectures, which managed to maintain their interest in my lectures, but this didn’t necessarily translate into deepening their understanding of the material. They would pay attention to what I was saying, since I embedded anecdotes of my time in industry or interjected jokes into the discourse, but in the end, the “smart students” did well, and the “not as smart” students did less well, and the results in terms of what I could discern they had learned, based on their exams and finals, was about the same. I did find over the last 10 years a number of things that did help to engage students in the learning process that did translate into broader and deeper understanding. It was after reading M2S that I saw the connection with the themes of your book and understood more broadly why these techniques worked. The problem that a professor has is a deep case of the Curse of Knowledge. Not only has it been a long time since we did not understand what we are trying to teach to students who have not yet grasped the concepts, but we have also taught it so many times, that there is a sense of “I’ve taught this 100 times, haven’t you understood it yet?” This is of course not a conscious phenomenon, but nonetheless, something that we all must battle.In the course of reading your book, I have also been re-writing the course lecture notes for this digital signal processing course and have been focussed on (using your term) “finding the core” of the course. I had come up with, over the last few years, a core set of ideas that I thought focussed on what it means to have taken and understood digital signal processing. When a student from the university of Illinois interviews at a company and says “I took digital signal processing from Prof. Singer” what are the 3 things that they need to know to both get the job *and* make the University of Illinois proud to have this graduate working in this field? By focussing on the core ideas of the course, I widdled away the extraneous details that basically served to separate the A+++ students from the A++ students, but largely fell on deaf ears on the rest of the class.Students need to understand what a mathematical model for a signal is, what happens when it is sampled, understand the concept of analog and digital frequency and how they are related, understand what happens when the digital signal is processed (in time and frequency) and what happens when this signal is then reintroduced to the analog world, through a digital-to-analog converter. This set of core ideas can be visualized in a picture, where the signals that touch the world—say a musical recording—are sampled and become a digital file, this digital file is manipulated, and then the file is played out through a D/A converter. By showing this to the class at the beginning of the term and referring back to this example, I found I could keep the class on track to the core messages I wanted them to learn. I also focussed on this core message when deciding what material to keep in the course and what should be left out. This was all before reading M2S, and now I see that I had successfully managed to get chapter 1 on my own, with a little of the notion of stories and concrete examples.

Post M2S: The night I finished reading M2S, I literally put down the book, went over to my lecture notes for the next day’s lecture and asked: “What is the core message of this lecture?” Where is it? Why am I burying this message so deeply in mathematics? I wrote a single page with the core message for the day on it together with a catchy diagram that illustrated these key concepts. Then, I focussed on creating a set of increasingly challenging concrete examples that illustrated this key concept and developed the supporting concepts one by one. Each example that I wrote, I looked at and decided were not yet concrete enough. For example, in one case I had a signal of the form “a^n u[n]” to express a one-sided complex exponential sequence. I thought, “Why am I introducing this extraneous variable ‘a’ ” in my supposedly “concrete” example?” I replaced this with the number “1/2” instead. Additionally, I provided a story to go with each concrete example. “Suppose the number of album sales for a particular record fell off geometrically, with half as many sold each day—that is, the sales took the form 1000(1/2)^n for the nth day of sales, beginning with 1000 sales the first day, 500 the next, and so on…”

Basically, I grounded each signal in as concrete an example as I could. Then, when I wanted to describe properties of the signals or how I would manipulate them, I gave the corresponding meaning (as close as I could) in the story of the album sales. The lecture went flawlessly, and I kept them in class past the bell at the end of the hour. Since then, I’ve added “mysteries” to be solved, introduced early in lecture, with the answer only revealed at the end. I’ve included such “riddles” in homework and laboratory exercises, to tease out the student’s interest in understanding the concepts sufficiently well that they *want* to find the answer.

I don’t know what the end result will be at this point, however I know that the course text that I write will be much more inviting, more concrete and focussed after reading M2S than it would have otherwise been, and, whenever I stand up in front of the students, I am constantly going through the “SUCCESS” list, where, in my case, the last “S” is for Student.

The Screaming Man in the Four Stroke Engine

Here’s one of our favorite stories so far from the “100 books for 100 stories” contest. There are still plenty of books to giveaway, so make sure to tell your teacher friends: Email us — — a story of a lesson that stuck and we’ll ship you a free signed copy of our book. (Must be a U.S. resident and a current teacher.)

Check out this tale from Saleem Reshamwala (a few comments below the story):

When I was a middle school student in Apex, North Carolina, I took a class called “Small Engines” with a guy named Mr. Trueblood. It was basically a class in how to repair lawnmowers, and a stepping-stone class for learning how to fix cars.

Here’s the four steps in making a four-stroke engine (the one in most cars) go:

1) Piston goes down, gas and air mixture gets sucked into the cylinder
2) Piston goes up, compresses gas and air (makes gas and air mix more explosive)
3) gas explodes piston is forced down (this is the explosion that makes your car go)
4) Piston goes up (exhaust is pushed out)

I don’t think a single one of us understood that about cars before we started the class. So, Mr. Trueblood tells us, a group of middle-school boys in rural North Carolina, that he’s going to teach us the basic science of how a four-stroke engine works. We’re expecting him to go to the blackboard with the chalk. He walks out of the room.

1) He then walks back in giving a monologue as if he were a mix of gas and air that had been sucked into a car engine. “Woah, got sucked in here, it’s not so bad lots of space to move around” and he’s kind of moving around the class a bit, acting as if he’s talking to various particles around the room. It’s a little weird, and some of the boys are laughing.

2) Then he starts acting as if the back wall of the class is moving toward him. He gets really into it. Laughing nervously at first, talking about how the piston is making things get really crowded for him and the other particles. Then he briefly looks genuinely scared. He’s talking about how being this crowded in, all he wants to do is anything he can to get out.

At this point, a few of us were like, ‘Uh, what the hell is going on here’

3) He yells something about a fire coming in the side of the class, and then SCREAMS and SPRINTS toward the back of the room, yelling that he’s burning. I was kind of terrified at this point. He looked crazy. And, like I said, he’s yelling about having come into contact with flame.

4) He slams himself into the back wall, stops acting crazy, and just acts like he’s exhausted, mentions how shocked he is at the force that he was able to push the piston away with, acts like it’s coming back towards him, and then walks out the classroom door.

I can’t remember if we clapped or not, but I know we all laughed. Nervously. And it sure as hell taught the concept.

There’s a lot to love about this: Note how the teacher is trying to turn a complex process into a concrete story. He is trying to get students to experience the four-stroke engine. And the fact that he freaks them out a little is just gravy. Also note that the initial student reaction to the, er, performance is not particularly positive. Sticky ideas won’t always get instant acclaim, and yet it wins in the end — here’s a guy who still remembers the details of a class from middle school!

Oceanography, amplified

I conducted a workshop recently for high school science and math teachers. We were working together to find ways to make their lesson plans stickier. My favorite example came from a couple of teachers who were trying to revamp the oceanography unit. Below is my own paraphrasing of what they said:

“We weren’t happy with how our unit on oceanography went last year. So we’ve put a lot of energy into how to make it better. Here’s what we’ve come up with.

In the first class in the unit, we start with a mystery: Let’s say you put a message in a bottle, drive out to the coast, and throw it as far as you can into the ocean. Where will the bottle end up? We let students make their guesses. (‘The waves will bring it right back to shore.’ ‘It’ll end up in Antarctica.’ ‘It’ll sink.’ Etc.) But we don’t give an answer (since there isn’t a clear answer).

Then we began to explore this same mystery in a more dramatic form. We’ll have students read a wonderful article from Harper’s magazine. What happened is this: In January 1992, somewhere in the Pacific Ocean, a cargo ship hit a severe storm lost a container overboard which held 7,200 packages of plastic toys, including thousands of rubber duckies. Years later, we know where many of these rubber duckies ended up. In fact, many of them ended up on the same beach! By tracing the paths that these duckies swam, we learn a lot about the way ocean currents work.

Next, we let the kids do some hands-on experimentation. We’ll set up tanks of water with different salinities and different temperatures, and let them see how those variables change the water current. In essence, we are letting them create their own ocean currents.

Finally, we’ll pivot to the critical role that oceans play with global climate. We’ll start by asking them: What determines the weather of a city, like New York City? Inevitably, students say it depends on the latitude of the city ‘ the closer to the equator the city is, the warmer it is, and the closer to the poles it is, the colder it is. There is much truth to that, but there are huge discrepancies: For instance, New York City and Madrid are at roughly the same latitude. Yet it snows every winter in NYC and doesn’t snow in Madrid. What’s the difference? That paves our way to talk about the way that ocean currents influence climate.

We will be trying this ‘new & improved’ sequence this fall, and we’re hopeful it will make the unit much more vivid for students!”

Tanzania post-mortem

Chip and I had an incredible time in Tanzania. So, let me start by saying this: You know that feeling you get when a full explanation would take 25,000 words, but you’ve only got 500, and you’re afraid to oversimplify, but you’re also afraid to give a vague “neat experience” summary, and this causes a cycle of anxiety and actually deters you from writing anything…? So, yeah, that’s the feeling I’ve got.

Let me start at the end: In conjunction with an amazing team of collaborators (more on them in a sec), we cooked up 2 campaigns that will likely be pilot-tested within 6 months. One campaign is intended to encourage Tanzanians to get tested for HIV. (The push for testing was recently given a huge boost by the Tanzanian president, who courageously made a well-publicized visit to a clinic to get tested. The number of tests spiked immediately afterward. And, as a further tangent, his action is a classic situation in which external credibility is more effective than idea stickiness.) The core of the “get tested” campaign is this: That getting tested is an act of solidarity with your fellow Tanzanians. It’s good for the country.

In the prototype posters we hatched, you see pairs of dissimilar Tanzanians (such as a grandma and granddaughter, or a Masai and a manager) who are clutching hands in a show of unity. Around their wrists are colorful bands that signify that they’ve been tested (a la LiveStrong). And the headline is, roughly, “We are Tanzanians.” [For those who have read MTS, we were consciously making an identity appeal with this campaign, similar to Don’t Mess With Texas. See the Emotion chapter. So many AIDS-related campaigns appeal solely to consequences — live longer, reduce your anxiety, it’s quick and easy, you can prevent harm for others, etc. We felt that an identity appeal could provide a stronger motivation.]

The other campaign is intended to combat the common phenomenon of cross-generational relationships. In the stereotypical case, an older male with money and status finds himself a young mistress (often as young as 18 or 19) and maintains a long-term-ish “sugar daddy” type relationship with her. These relationships provide a route for HIV to spread between dissimilar populations, which creates a nasty epidemiological problem. Now, certainly this kind of relationship exists in the Western world. But there’s a counterforce: Social taboos. If a 50 year-old guy is hitting on an 18 year-old, our Pathetic Alarm goes off. Nasty looks are offered. People whisper about the lecher behind his back. But there’s no real equivalent of this taboo in Tanzania — there’s no common language of scorn or disgust or distaste to invoke against these situations.

So, in the second campaign, we introduce a villainous character named Fataki (Swahili for “explosion” or “fireworks” — i.e., keep your distance from this dude). Fataki is a wealthy, powerful guy who is irredeemably sleazy. Comically sleazy. He hits on anything that moves — schoolgirls, strangers, other mens’ wives. In the campaign — which would include posters and radio spots — we’ll try to make people laugh at Fataki’s shamelessness, and also, importantly, at the repeated rejections he receives (think Wile E. Coyote). As the tagline of the campaign will read, “He’s constantly hitting on women. He won’t wear a condom. He won’t get tested. DON’T LET YOUR LOVED ONES GET MIXED UP WITH A FATAKI.”

Our hope is that “Fataki” will enter the lexicon. Our dream is that a year or two from now, we could return to Tanzania and overhear someone in a bar, saying about someone else, “He’s such a Fataki.”

Working on these campaigns was one of the most fun and satisfying moments since the book launch. And what made it fun and satisfying was the company we had: Our hosts, Pam and Mike, from USAID, the branch of the State Department that would lead the launch of these campaigns. Bob (affiliated with Johns Hopkins), who helped organize and lead our work. Plus a half-dozen AIDS-education experts who’ve been working in the region and were willing to share their knowledge and experience. And a hell of a Tanzanian creative team — approximately a dozen scriptwriters, producers, artists, and actors — who helped us develop and produce the campaign mockups, which in turn gave us something concrete to present to the decision-makers. (The process of producing creative work, with conversations constantly translated between English and Swahili, is probably worth another blog post.)

Okay, I’m gonna shut up now. Thanks to those of you who sent encouraging notes to us. We’ll continue to post updates to track the progress of the campaigns.

A tale of two bottled-water stories

I previously blogged about Charles Fishman’s insightful and thoroughly sticky piece on the bottled water industry. His main point: Our embrace of bottled water “is not a benign indulgence.”

Then I got this note from Mojo Mom, who I’m a big fan of:

Today I heard NY Times reporter Julia Moskin interviewed [about bottled water] on NPR’s The Splendid Table and I learned two interesting things:

She had written a long article about bottled water last year, “Must Be Something In the Water” (2/15/06) and covered much of the same ground as Charles Fishman in Fast Company.

But her article was much less sticky: she buried the lede about the environmental concerns in the middle of the article, and didn’t use compelling concrete or emotional details. She covered broad set of ideas, about why we like bottled water, what it tastes like, etc. Fishman made a much stickier, emotional narrative (going to Fiji and finding out that native people didn’t have access to water, when we Americans are buying their water and shipping it 8000 miles across the world when we have clean water coming out of our tap. Ouch.)

This is a great case study for people interested in sticky communication. Go ahead, follow the NYT link and plunk down the money for the archived article. It’s worth it. What you’ll find is that there is absolutely nothing wrong with Moskin’s article. It’s well-researched, interesting, well-written. But you’ll also notice that there is something different about Fishman’s piece, and it’s precisely that difference that is likely to change behavior.
Moskin’s piece doesn’t suggest it has qualms about bottled water until the 7th or 8th paragraph — and even then, it’s a qualm about whether sugary water (like Propel) counts as “water.”
Fishman’s piece leads off with an image that encapsulates the core idea of his piece (that our embrace of bottled water is absurd and costly):
The largest bottled-water factory in North America is located on the outskirts of Hollis, Maine. In the back of the plant stretches the staging area for finished product: 24 million bottles of Poland Spring water. As far as the eye can see, there are double-stacked pallets packed with half-pint bottles, half-liters, liters, “Aquapods” for school lunches, and 2.5-gallon jugs for the refrigerator.
Really, it is a lake of Poland Spring water, conveniently celled off in plastic, extending across 6 acres, 8 feet high. A week ago, the lake was still underground; within five days, it will all be gone, to supermarkets and convenience stores across the Northeast, replaced by another lake’s worth of bottles.

(Note, by the way, that the statistics are utterly unsticky — it’s only the image of vastness (a plastic lake) that sticks.)

Both writers discuss the environmental costs of trucking water around (in particular from Fiji). Here’s Moskin:

“This month the Earth Policy Institute, an environmental association based in Washington, published a research paper outlining the global issues raised by bottled water. ”Water is very heavy, and moving large quantities of it, for example, 8,000 miles from Fiji to New York, takes considerable resources,” said Janet Larsen, the institute’s director of research. ”Nearly a quarter of all bottled water around the world crosses national borders to get to its market. Bottled water is not a global environmental crisis in itself, but it is an issue of global equity and of human rights; we believe clean water is a basic human right.””

Here’s Fishman:

The label on a bottle of Fiji Water says “from the islands of Fiji.” Journey to the source of that water, and you realize just how extraordinary that promise is. From New York, for instance, it is an 18-hour plane ride west and south (via Los Angeles) almost to Australia, and then a four-hour drive along Fiji’s two-lane King’s Highway.

Every bottle of Fiji Water goes on its own version of this trip, in reverse, although by truck and ship. In fact, since the plastic for the bottles is shipped to Fiji first, the bottles’ journey is even longer. Half the wholesale cost of Fiji Water is transportation–which is to say, it costs as much to ship Fiji Water across the oceans and truck it to warehouses in the United States than it does to extract the water and bottle it.

That is not the only environmental cost embedded in each bottle of Fiji Water. The Fiji Water plant is a state-of-the-art facility that runs 24 hours a day. That means it requires an uninterrupted supply of electricity–something the local utility structure cannot support. So the factory supplies its own electricity, with three big generators running on diesel fuel. The water may come from “one of the last pristine ecosystems on earth,” as some of the labels say, but out back of the bottling plant is a less pristine ecosystem veiled with a diesel haze.

Each water bottler has its own version of this oxymoron: that something as pure and clean as water leaves a contrail.

Note a few things about Fishman’s work. First, instead of talking about the distance (8,000 miles) from Fiji to the U.S., he puts it in more human terms (a very long plane ride). Instead of talking about the big costs (in raw-dollar terms) of shipping, he makes a useful comparison — it costs as much to ship this water as it does to bottle it from its pristine source. And he also understands that this is an emotional issue. Can you drink Fiji for its purity when you know that the bottling plant “is veiled with a diesel haze”?

More on emotion. Fishman has a keen eye for gut-twisting irony:

And in Fiji, a state-of-the-art factory spins out more than a million bottles a day of the hippest bottled water on the U.S. market today, while more than half the people in Fiji do not have safe, reliable drinking water. Which means it is easier for the typical American in Beverly Hills or Baltimore to get a drink of safe, pure, refreshing Fiji water than it is for most people in Fiji.

This works because it’s tapping into your sense of identity. How can we drink fancy water from an exotic locale, shipped with grotesque expense to our doors, when the people who live in that locale are drinking dirty water? That doesn’t feel like us — that’s not the kind of people we are. Fishman has put the social justice into bottled water.

One final comparison. Both writers discuss the trivial costs of tap water compared to bottled water. Here’s how Moskin does it:

Neau, a nonprofit organization based in Amsterdam, has one goal (raising money for drinking-water projects in third world countries) and one product: an empty blue plastic bottle, for about $2, with a glossy logo and a flier inside explaining that profits are donated to the foundation’s water projects. The buyer is expected to fill the bottle with tap water. ”Two thousand liters of tap water cost less than one liter of Spa,” a popular Dutch mineral water, Mr. Liauw said. Ethos Water, an American company that sponsors similar drinking-water projects, was bought by Starbucks in 2005; 5 cents for each bottle sold is donated to water charities.

Here’s how Fishman handles the same point:

In San Francisco, the municipal water comes from inside Yosemite National Park. It’s so good the EPA doesn’t require San Francisco to filter it. If you bought and drank a bottle of Evian, you could refill that bottle once a day for 10 years, 5 months, and 21 days with San Francisco tap water before that water would cost $1.35. Put another way, if the water we use at home cost what even cheap bottled water costs, our monthly water bills would run $9,000.

Wow. Again, what he’s doing is putting statistics in human scale. The person quoted in Moskin’s piece goes with “2000 to 1” as her talking point. Fishman grounds the ratio in human experience. (Plus, the mental image of someone spending 10 years refilling a bottled water to “catch up” is fittingly absurd).

And here’s one for the road. In the book, we talk about the way vivid details build credibility. Here are two vivid details unearthed by Fishman:

– In the town of San Pellegrino Terme, Italy, for example, is a spigot that runs all the time, providing San Pellegrino water free to the local citizens–except the free Pellegrino has no bubbles. Pellegrino trucks in the bubbles for the bottling plant.

– At the height of Perrier’s popularity, Bruce Nevins [the man who brought Perrier to America] was asked on a live network radio show one morning to pick Perrier from a lineup of seven carbonated waters served in paper cups. It took him five tries.

The point of this post is not to have a “Moskin vs. Fishman” smackdown or something. And it is not fair to compare the persuasiveness of the two pieces, because Moskin is not writing a persuasive piece — she’s a reporter. (Nor did she have anywhere near the time or budget that Fishman did to pursue his story.) But what I care about is what we can learn from studying two sets of communications that spotlight the same issue.

Why does Fishman’s stick? Moskin samples various points related to bottled water; he hammers on his core point. (Simplicity) She presents useful statistics from experts; he makes those statistics real by grounding them in human terms. (Concreteness, Credibility) She presents a buffet of expert testimony; he finds the emblematic stories (Story). Both of them do a good job of shocking us into paying attention (Unexpectedness).

Only Fishman made me change my behavior.

(As a parting note, I have an unhealthy fascination with these side-by-side comparisons. If you come across pairings like this, please send ’em along.)