To Answer Questions About Your World, You Took a Sample
One month ago, the Science Club encouraged listeners to answer questions about their world through the use of sampling. By finding a representative, unbiased sample, it’s possible to gain insight into questions about things that might be too large or complicated to observe directly. You responded, sending us dozens of examples of sampling projects on topics such as water quality and carrot length. This week, the Science Club meets again to review your work, with field ecologist Rhine Singleton adding his thoughts on sampling techniques and data collection challenges.
Ariel Zych is Science Friday’s director of audience. She is a former teacher and scientist who spends her free time making food, watching arthropods, and being outside.
As Science Friday’s director, Charles Bergquist channels the chaos of a live production studio into something sounding like a radio program. Favorite topics include planetary sciences, chemistry, materials, and shiny things with blinking lights.
IRA FLATOW: This is Science Friday. I’m Ira Flatow.
That sound means, of course, it’s time for our Science Club to meet and kick off this season’s project. And that was looking at scientific sampling. And we told you to go out and sample stuff, anything you wanted to sample. You had a month to explore your world, and now we’re back to talk about what you found.
Did you take a sample to answer some question about the world around you? Did you? Give us a call, we want to hear about it. Our number 844-724-8255. You can also tweet us @scifri. The phone number, 844-SCI-TALK.
And here to get us going to talk about some of the projects you sent in, our Science Club founding members, Ariel Zych and Charles Bergquist Welcome back.
CHARLES BERGQUIST: Thanks Ira.
ARIEL ZYCH: Hey, thanks Ira.
IRA FLATOW: Ariel is joining us from WESA in Pittsburgh. And also joining the Science Club today is Rhine Singleton, he’s an Associate Professor of Biology and Environmental Science at Franklin Pierce University in New Hampshire. Where among other things, he teaches budding ecologists how to do field work. And he is at New Hampshire Public Radio today. Welcome to the program.
RHINE SINGLETON: Thank you. It’s a pleasure to be here.
IRA FLATOW: All right Charles, kick us off on the Science Club. Remind us what people had to do.
CHARLES BERGQUIST: Well we started by saying a lot of the really interesting questions in the world, like how many hairs are there on my dog. They’re things that would really be hard to answer in a straightforward manner. You can’t count all the hairs on your dog.
So we wanted people to think first of a question about their world that they were interested in exploring. Something that might be hard to get a handle on in a straightforward, linear fashion. And then work out a way to use sampling to explore a little bit of that thing, and use that little bit to answer the larger question. Or at least approximate the larger questions.
IRA FLATOW: Yeah, draw a conclusion about the larger prac– And Ariel, how well did we do for projects coming in?
ARIEL ZYCH: Oh my goodness, there were just so many interesting things. Well, so first off the questions were all over the map. So Ms. Nolan’s fourth grade class wanted to know what was in their schoolyard, and so they actually sampled their schoolyard by the square meter and then extrapolated to the whole schoolyard to estimate. They had like 4,000 caterpillars, which is cool.
And then Doug and coworkers were really excited when they found a giant bag of cookies in the vending machine. And so they decided to spend the month sampling the bags of cookies to see if all of the cookies–
IRA FLATOW: Tough assignment, tough assignment.
ARIEL ZYCH: Yeah, right? I know. Eating cookies every week to kind of get an estimate of the number of cookies per bag. So we did very, very well.
IRA FLATOW: And what are some of your favorite, you mentioned them. But there was some sort of really interesting– there’s one about dog poop I understand.
ARIEL ZYCH: Yeah, we had a couple of dog poop heroes. Shout out to Claudia and Theresa for attacking one of my least favorite things about living in an urban environment. So they had a dog park that was overrun the dog poop. And they wanted to know, OK, what is the actual daily rate of dog poop leavage.
And so they actually cleaned up and counted the number of poops left every day for a whole week in their dog park. And then also counted and got an estimate for the other dog parks in the area. Estimate that in an average year, about 32,000 dog poops go uncleaned in their neighborhood. Which is a pretty telling sample estimate. I know, I thought that was incredible.
IRA FLATOW: Should put that in a newspaper.
ARIEL ZYCH: Yeah, right? I know. I think that could inspire some change. There were also some of those questions that were just kind of quirky and curious. So Lilly wanted to know what like a tolerable number of R2D2 sounds were.
So she actually watched some of the Star Wars series and sampled the number of R2D2 sounds that he makes in a typical Star Wars film, and then tried to figure out what is like the going R2D2 speed rate. And it’s about 54.66 sounds per movie, which I thought was kind of interesting.
We also had a ton of citizen science projects. So folks that have been doing citizen science. Like Craig has been sampling the Charles River for over 20 years. Like Laura and her family, who went out with Save Our Streams and sampled aquatic micro invertebrates and macro invertebrates to figure out if their stream was healthy, which it was. They got a clean bill of health thanks to all of the stone flies and caddisflies they found.
And we also had a lot of scientists who were just doing their thing. Scientists use sampling all the time, and a lot of them were courteous enough to share some of their scientific sampling with us, which is really excellent.
IRA FLATOW: Speaking of scientists, Rhine Singleton, you saw some of the questions our listeners were tackling. And you teach early ecology students field techniques. What are some of the key rules that you need to go out, if you’re sampling, going out into the field. What are the rules of thumb here?
RHINE SINGLETON: Yeah, well we could boil it down to three, which I think give a pretty good start. And maybe I’ll just list them, and then we could talk about each one in a bit more detail if you want. But I would say the first thing is to define the population you plan to sample from. And the hair on a dog would be a great example. Nobody’s ever going to count all those hairs, but you could define that population and take a sub-sample.
The second rule of thumb is to take care to get an unbiased representative sample. Really key, and I think we could talk a little bit more about that in a second if you want. And then my third rule of thumb would be to keep in mind the sampling effort that it’s going to require to do your study. And I think a lot of your listeners did a wonderful job. I had a lot of fun browsing your web page with those posts.
IRA FLATOW: Our number 844-724-8255. We have some calls coming in from people who did some sampling. Let’s go to Orlando. Hi Kevin.
KEVIN: Hi, Ira. How we doing?
IRA FLATOW: Fine, go ahead.
KEVIN: I’m an appliance repairman over here in Orlando. And actually I took a sample over maybe like the last year or so of my career. And literally every person’s dryer smells the same. No matter what kind of dryer lint you have, no matter what kind of dryer sheets you use, any kind of product. I’ve worked in thousands of homes, and every single dryer smells exactly the same.
IRA FLATOW: That’s a big sample, Kevin! Have you told Maytag about this? Or maybe they know.
KEVIN: No not yet. I’m pretty sure they’d know though.
IRA FLATOW: All right, that truly really unique. Thanks, Kevin, for calling. Wow, Rhine What do you think about that?
RHINE SINGLETON: Well I do have a response. That seems to violate one of the fundamental rules of the natural world. I’m constantly telling my students that there’s variability everywhere you look. And apparently I hadn’t yet looked in dryers, so I guess we have an exception to that rule now.
IRA FLATOW: Charles how well did our listeners do on these things? we heard from Ariel talking about some of the school kids.
CHARLES BERGQUIST: Right. I mean there was a lot of creativity out there. The approaches to the questions were really, really amazing. Probably I think if we look back to Rhine’s three rules, the one that people had most trouble with was the representative sample, the unbiased sample.
I think some of our submissions, while they were really great, definitely had concerns with sample size and whether or not they were truly picking a representative sample of the population. It’s hard to say something about the entire universe of carrots when you measure five carrots.
IRA FLATOW: Oh. Rhine, how many carrots would you need?
RHINE SINGLETON: Well we’ve got to define what population of carrots we’re interested in. And I am pretty sure there’s several varieties out there. So–
IRA FLATOW: How do you know when your sample size is big enough, Rhine?
RHINE SINGLETON: That’s a great question. And statisticians love to debate that. There’s this thing called a power analysis that you can do in statistics that, in part, depends on how big a difference you’re trying to detect. So that is hard to boil down to a rule of thumb. But there’s a lot of fun math you can do to try to calculate what your ideal sample size should be.
IRA FLATOW: Ariel during the course of the project you ran a few Twitter polls as demonstrations, right?
ARIEL ZYCH: Yes we did. Yes we did. So we figured, OK, like what better environment to get a giant sample size than the Twittersphere. So we put it out to all of our followers, some 600,000 people, and asked them if they could curl their tongue.
And wanted to see what rate of response we got and also what the rate of tongue curlage is among our audience. This used to be thought of as a heritable Mendelian trait. No more, we know now that it can be learned, and its plastic, and probably controlled by a lot of stuff.
But we got this response rate of like 2,000, 3,000 people, and 81% of them could curl their tongue. And I was like oh, this is awesome, let’s share our results. But I asked around to some friendly statisticians who hang out with Twitter as a collection tool, and they said no. This is actually a crazy opportunity to evaluate whether your poll is biased.
And in fact, our poll was really biased because people are more likely to respond if they can curl their tongue. People might try to learn how to curl their tongue before responding. People might be embarrassed if they can’t.
We also realized pretty quickly that, yeah, Science Friday listeners, you’re exceptional guys. Like we know that, we know that they’re like scientific minds, right? They’re not an unbiased sample in and of themselves. So we can’t even infer a whole lot about like American tongue curling rates based on our sample, because Science Friday followers are a special breed of Twitter followers.
IRA FLATOW: Wow. That’s good news and bad news.
ARIEL ZYCH: It is. We did get some pictures, also some questions about what constituted a tongue curl. So there’s a lot of things that you can do to poll with Twitter more effectively. But yeah, sampling is not easy, actually. It can be pretty challenging.
IRA FLATOW: Rhine, there’s a lesson right there.
RHINE SINGLETON: There is a lesson. And I loved that example, I got a real kick out of it. But I thought to myself, well OK, what if we defined the population as the Twitter followers of Science Friday. Then we’d perhaps want to know the answer to the question.
But the self reporting bias is a challenge. It’s a little different than the challenges I experience. Sometimes I find myself with students in a shrubby thicket, and we’ve got to get an unbiased representative sample. And it’s very tempting to go towards that slight clearing in the thicket so you don’t get your eye or your ear poked with those sharp undergrowth. But we got to dive right into the thicket on our hands and knees and get that representative sample.
So my challenges are a little different than the self reporting bias that you guys experienced, but I could really sympathize.
IRA FLATOW: When you say one of your three rules is keep in mind sampling effort, what did you mean by that?
RHINE SINGLETON: Well that’s the biggest category. So that’s kind of a grab bag to boil this down to three. But on a basic level, it’s how much time is it going to take you to get enough meaningful information to proceed with the study. And if you’re just doing something for fun, measuring one or two individuals could be pretty cool and rewarding, but it may not be very meaningful.
I experience this kind of problem with my students when they want to study the bear, or the moose, or the coyotes or bobcat that come through our campus. And you know they quickly realize it’s going to take many, many hours to possibly even find sign of one of those, although it is possible. So to design a study that’s doable, you’ve got to keep in mind constraints. How much time do you have, how much funding might it take, those kinds of things.
And then if you’re comparing different areas, you may need equal sampling effort in different areas that you’re comparing. So there’s a lot there.
IRA FLATOW: I’ll bet. We’re talking about our sampling project on our Science Club this hour on Science Friday from PRI, Public Radio International. Here with Ariel Zych, and also with Charles Bergquist and Rhine Singleton.
Ariel, who ended up participating in this? Who contributed? All sorts of people? Or–
ARIEL ZYCH: Pretty much every kind of sampler you could imagine. We didn’t have any animal samplers, but we did have many people who sampled things about animals. We had a whole bunch of people from, like I mentioned, the scientific field. We had a bunch of normal humans who are just interested in coin collecting, who are interested in water.
We had a bunch of schools participate, including a school from South Korea, which was incredible. You want to talk about sampling effort, they were sampling things like how many grains of sand might there be on a beach and estimating it.
RHINE SINGLETON: Whoa.
ARIEL ZYCH: So, I know, they took on some pretty serious mathematical challenges, not to mention ones that required an extraordinary amount of patience. But yeah, I mean it was really the whole gamut.
CHARLES BERGQUIST: Yeah I want to give a special shout out to Mrs. Reed’s class at Donner Trail Elementary in Truckee, California.
STUDENTS: We’re from Donner Trail School. We sampled the Lahontan cutthroat trout.
MRS. REED: And how many do you have?
MRS. REED: Is that all of the Lahontan cutthroat trout?
MRS. REED: No, what do you have?
STUDENT: A sample.
MRS. REED: A sample of them, yeah.
CHARLES BERGQUIST: So there, kindergartners through fifth grade did a study with the Headwaters Science Institute, which is an educational outreach environmental group out there. And they looked at 150 of these cutthroat trout. And so we thank them for their efforts.
IRA FLATOW: You know you don’t realize how much of an influence on kids something like that may have, Ariel, do you? I mean just going out, and something they’ll remember for the rest of their lives and understand a little bit more about how science works.
ARIEL ZYCH: I think that’s absolutely true. I mean and for me what was so mind opening, and I think what was mind opening for the people who participated in this project is that you really can glimpse the unfathomable. This is like how humans understand things, right?
We know that we like vanilla ice cream because we’ve sampled a whole bunch and we generally like vanilla ice cream, right? It’s something that is so human to do, and yet when you take it to that scientific level and you start like stepping a little further out, you can explore incredible things. Like whole mountainsides, or whole ecosystems. So it’s a lot of fun, we really had a great time.
IRA FLATOW: Are we cooking up another project? Are we just thinking about it now, Charles?
CHARLES BERGQUIST: We’re thinking about it. We’ll be back in the fall with something fun for people to try, so definitely stay tuned for that.
IRA FLATOW: Rhine, any suggestions about projects that you–
RHINE SINGLETON: Oh gosh, well it’s an amazing world out there. I say go out and sample.
IRA FLATOW: Continue sampling. I
RHINE SINGLETON: Continue sampling. One of the beautiful things is when you get out there and sample, you tend to notice things that you never would’ve noticed otherwise. So I encourage everybody to do that.
IRA FLATOW: All right, and Ariel, we’ll be putting our thinking caps on for the next project.
ARIEL ZYCH: Please do. And if you’ve got good ideas that you want to share with us, you can share them at Science Club at sciencefriday.com, you can email us. And check out the website. There’s so many great samples out there that you guys can see at sciencefriday.com/scienceclub.
IRA FLATOW: There you have it. For this Science Club edition. Science, SciFri education manager Ariel Zych. Contributing producer Charles Bergquist of the Science Club, Rhine Singleton, associate professor of biology and environmental science at Franklin Pierce University in New Hampshire. Thank you all.
ARIEL ZYCH: Thanks as well.
CHARLES BERGQUIST: Thanks, Ira.
IRA FLATOW: Have a great weekend. BJ Leiderman composed our theme music. Our thanks to our production partners at the studios at the City University of New York.
And on the web this week, remember when you bought your first Kindle? And how cool it was to read an ebook that looked like paper? Well the guys who invented that display, called electronic ink, were just inducted into the inventors hall of fame. Yeah they should have been, right?
And you can read the whole story about the ink, and the Kindle, and the Hall of Fame at a link at sciencefriday.com/ink.