More Science Talk
Steve: Welcome to Science Talk, the weekly podcast of Scientific American posted on May 19, 2009. I'm Steve Mirsky. This week we'll talk to high school scientists, who have done some really fascinating research with the added benefit that I could actually understand most of it. At the annual meeting of the American Association for the Advancement of Science back in February, I ran into a few dozen high school students, who were presenting their research in a big poster session. The kids had won their state science competitions, sponsored by the American Junior Academy of Sciences. As I wandered through the posters, I wound up interviewing five of the winners whose research just grabbed me. We'll hear from Sruti Swaminathan, Maia ten Brink, Alyssa Bailey, Moyukh Chatterjee and Fedja Kadribasic. First up is Sruti Swaminathan. She is a student at High Technology High School in Lincroft, New Jersey.
Sruti: I studied the effect of cell phone and Bluetooth devices on simulated driving performance.
Steve: And why did you decide to do this study and what did you find?
Sruti: I started to conduct a study on driving performances because, as a 16-year-old, I'm going to be a future driver and I wanted to know the risks that I would be putting myself in by talking on a cell phone while driving or using a Bluetooth device while driving.
Steve: And the risk you would be putting me in, as well...
Sruti: Yeah, and my also, like, fellow drivers on the roads. So what I did was I recruited 28 test subjects and had them complete three laps on the game Gran Turismo 3 for PlayStation 2; and they had a steering wheel, an acceleration [and] brake pad to simulate a real driving experience, and I counted the number of crashes they made during each lap, which was the number of times they hit the wall and also the number of times that they deviated from the course. And after collecting that raw data, I used an ANOVA, which is [a] statistical test, and I found that there was a significant difference between the three modes of conversation. So I went on to conduct three individual T-tests in between each mode of conversation.
Steve: Well back up a little bit. Three modes of conversation: somebody next to you, somebody on telephone or somebody with Bluetooth?
Sruti: Yeah. The first one was with no conversation at all, and the second one was using a cell phone, and they would be talking on the cell phone, and I would be in a different room.
Steve: They'll be talking on the cell phone holding it.
Sruti: Holding it.
Sruti: So they had one hand to drive. And I would engage them in conversation with a set of scripted questions and [they would] be questions [that] simulated, like, thought, like "What did you eat for dinner last night?" or questions like that. And the third trial, they had a Bluetooth device and earpiece, so they had both hands to drive, and I would also ask them another set of scripted questions.
Steve: Okay. So what were your results?
Sruti: My results were that there were significant differences between all three modes of conversation, and using the Bluetooth device would be the safest option for a driver who wants to engage in conversation. But of course using no devices at all and not engaging in conversation while driving is the safest option for any driver.
Steve: And actually holding the phone to your ear was the worst.
Sruti: It was the worst option, yeah.
Steve: And give me some numbers.
Sruti: So, the p value for, without any devices versus, like, with the cell phone was 0.056 x (10)-6 which is extremely lower than the alpha value of 0.05.
Steve: Okay. So all that means you have a very low risk of crashing.
Sruti: Yes. So, and the same thing for no devices versus the Bluetooth device, which is 0.002 x (10)-6, which is also extremely low. And the final comparison was between the cell phone and the Bluetooth device, and that was 0.029 which is less than 0.05, but not as significant. But you can still see that from the mean averages between all three modes of conversation, using a Bluetooth you only crashed 4.04 times in a lap, but while using the cell phone it was 5.04.
Steve: And how many times per lap with no communication?
Sruti: That was only 1.79.
Steve: Okay. So, you're still crashing. Again this is just on a PlayStation game.
Steve: But clearly the, and how many—28 subjects?
Sruti: Twenty-eight subjects.
Steve: But with those subjects in your runs, you found
out what appears to be quite a significant difference.
Steve: Maia ten Brink attends Falmouth Academy in Falmouth, Massachusetts. Maia, tell me about your study, what, it's called—the effects of...
Brink: ...Sleep depravation on cockroach olfactory memory.
Steve: And how did you wind up getting interested in the relationship between sleep deprivation and memory? Was it your own personal experiences?
Brink: Kind of. I am a teenager. I never sleep enough. I find any neuroscience stuff really fascinating and memory, sort of, sparked my interest. Memory, sleep stuff, I just thought they were interesting.
Steve: And you studied olfaction as a way to get into memory.
Brink: Yes, exactly.
Steve: Not particularly interested in how cockroaches smell things.
Brink: No, and I ended up working with cockroaches only because as a school project in the state of Massachusetts, we have regulations on working with vertebrates. It's really difficult to get past to work with like mice or rats, and it's really impossible to work with humans. So I was just looking around for an invertebrate species that, you know, goes through periods of rest or sleep, and there aren't that many. There is cockroaches, some species of bees and also the
re is fruit fly.
Steve: And so you're working with the species, the American cockroach.
Steve: And the American cockroach is pretty big, it can be maybe, what, an inch-and-a-half long?
Brink: Yeah. It's about two knuckles length.
Steve: Two knuckles!!!
Brink: Yeah. And they have wings, unlike the Madagascar cockroach, but they can jump about 6 or 7 inches in your face.
Steve: So how did you do this work, [and] where did you get the roaches?
Brink: I had to order the roaches from Carolina Biological Laboratories, and I did all the work in my home. I do live in Woods Hole, so it's [a] big scientific community.
Steve: Oh! [A] lot of scientists in Woods Hole.
Brink: But I ended up not working with a mentor. I just worked in my own home and sort of enlisted my mother to help me, just [be] an extra pair of hands. I basically, just was very careful [about] making sure the organism stayed always within, like, two containers, so that if they jumped out of one, I could really smack them down quickly. I made a glove box. I tried to make it as easy as possible to work with them. But it was actually very difficult method to carry out and I had to make a lot of really fast decisions in the middle of the night about changing number[s] of replications or simplifying a lot. I mean, I started out with a very ambitious plan to test and train
it at different times in their circadian rhythm and to just use a lot more replicates and then, you know, it became apparent pretty quickly that they're pretty quick buggers and just getting one alone to train, to associate the taste and the smell was very difficult. So I had to just make a lot of quick decisions like that.
Steve: So what were you actually doing with these critters?
Brink: Well, I was working with 45 cockroaches, and so I trained half of them, according to their natural preference. I was using two odors, vanilla and peppermint, and they prefer vanilla and they dislike peppermint; and two tastes as kind of reward and punishment, sucrose solution as a reward and saline solution as a punishment.
Steve: Salt as a punishment, sweet as a reward.
Brink: So half of them I trained against their natural preference, so associating peppermint with sucrose and vanilla with saline, and half according with [what] their preference already was. And then in each of those groups I sleep deprived half of those, and I allowed the others to rest normally.
Brink: Now how did you sleep deprive roaches?
Steve: I constructed kind of a shaking table, and I kept them awake shaking and with loud noise. It was made up like a rock tumbler and a tray and they were duct-taped down there. Some other studies, when they have the equipment, they usually will blow CO2 into the faces of the cockroaches, because it causes a natural predatory response and they will stay awake. They'll blow every two seconds to the cockroaches. So it's like being, you know, poked awake continually.
Steve: So what did you find?
Brink: Well, there was a pretty significant effect on memory retention. I was able to look at not only the memory retention, but how successful my training was. Obviously I had a lot of trouble making sure that each cockroach was, sort [of], getting exactly the same amount of solution, and there was a lot of, just, sort of, methods problems that really could be fixed with a couple of more hands and more planning time. Unfortunately, because I was on the 24-hour schedule to deprive them exactly 24 hours, if I was going to make a quick decision, I had to do it then, otherwise order a whole new batch of cockroaches. But I did find that their memory was decreased after sleep deprivation [a] relatively significant amount in both the cockroaches trained according to and against their natural preference.
Steve: Can you explain approximately how much their memory was affected?
Brink: The group trained against their natural preference—because I'm looking at different types of behavior to try to decide whether that means they are sleep deprived or not; so I was looking at three things. A standing event, which was when they would go over to the odor source and stand on top of it as if they kind of wanted to hang out by that odor source. And then the other was a tasting event, where they would actually associate the memory of the sucrose on their legs and begin to lick their legs when they were near to an odor source. So I was using this to see behavior criteria. So in terms of standing events, if the non-rest-deprived cockroaches, for the group trained against their natural preference, about 50 percent of the time, their standing events occurred near peppermint; and then after sleep deprivation that went down to about 30 percent, so a 20 percent drop. And it's a smaller difference for those trained according to their natural preference, and I'm not quite sure why that is, but it probably has to do with, obviously if they already preferred vanilla, sort of adding a reward, may or may not have really made a significant difference.
Steve: Very cool. Do you want to be a psychologist or a neuroscientist?
Brink: Yeah, that's definitely an area I want to continue in. We had a breakfast this morning with AAAS scientists and I hung out by the neuro table the whole time. It was really a wonderful experience.
Steve: Alyssa Bailey goes to Central Lee High School in Donnellson, Iowa.
Bailey: Well, I am really wanting to be a pharmacist when I get older. So I wanted to do something in the medical field to see if I was really interested in it. So to get started, I just took different pain relievers, and I dissolved them in a control, which would be ...
Steve: Well, tell me what was your goal? What did you want to find out?
Bailey: Really all I wanted to find out to see which would dissolve the quickest, therefore give you better relief faster.
Steve: Your assumption was that whichever pain reliever dissolved the fastest would wind up knocking out your headache the fastest.
Steve: Okay. Do we know if that's a legitimate assumption?
Bailey: I really wanted to do further testing because I know that absorption rates into your blood won't be the same. So knowing that aspirin dissolves faster, I wanted to test the different absorption rates into [your] blood to see if it would actually absorb faster, because who knows, that one might be the slowest to absorb.
Steve: Right. But this study is about which ones dissolve—the rates of dissolving of the various ones. So which ones did you test, and what did you find?
Bailey: I tested aspirins, Tylenol and ibuprofen and then I tested like a store brand and a name brand. Ibuprofen seemed to take longer to dissolve, Tylenol was about in the middle and then aspirin was really fast to dissolve. They took about 15 seconds to dissolve.
Steve: So this study seems like you could do it at home in your kitchen.
Bailey: Yep, I did it right in my dining room.
Steve: How long did it take?
Bailey: It took quite a while because I had to, I did a bunch of trials with all of them, so I would wait for it to dissolve, and then I would have to carry it to my kitchen, clean it out, make sure it was all nice and clean before I did it again.
Steve: The glassware you mean?
Bailey: Yeah. I used beakers. I put a 100 milliliters of the stomach acid solution in that.
Steve: Where did you get that?
Bailey: I made that with hydrochloric acid and distilled water and I made that a pH of 1.
Steve: Now how did you measure when the dissolving was complete?
Bailey: I ran a couple of test runs first to see how they would dissolve. There was a liquid pill that I tested and once that dissolved, the coating, it was all dissolved, it was that liquid, but all the other ones were more of particles; and so I just made sure there wasn't one solid pill left that they were all, the particles were pretty much equal in size.
Steve: I see, so you [would] just eyeball the beaker and make sure that you might not have dissolved it so that it was invisible, but you wanted to make sure that it was granular and completely uniform in granularity.
Steve: And so the next step in this ongoing research project is going to be the uptakes.
Bailey: Yeah. I really want to test the blood absorptions, but blood is kind of [tricky] to work with in labs and get into the science fair. So we were trying to work our way around that and see what else I can do.
Steve: How many headaches did you get working on this?
Bailey: I didn't get any, but a lot of people would stop and ask me if I had any of the pain relievers left over.
Steve: Moyukh Chatterjee also goes to High Technology High School in Lincroft, New Jersey.
Chatterjee: Well, basically for a while now, I've been interested in efficiency in industry. I really want to make things more efficient; I'm the kid who works with solar power and wants to work on solar panels. So basically what I did is I was working on a previous project, I was trying to develop a project where I wanted to create a, sort of, a magnetic plastic. And I did some research and I found that there was a lot of high level chemistry involved and I just did not have that knowledge. But while working and doing the research for that, I found a very interesting concept, where a person had put two refrigerator magnets together to sort of make a worm gear. And I was looking at that and I was trying to understand how that worked, and it didn't make any sense.
Steve: Again, you said to make a worm gear.
Steve: So explain what a worm gear is.
Chatterjee: Basically what you have in a worm gear is you have a normal gear called the spur gear, basically the normal gear you see everyday. And what you have connected to that is the worm gear, but a worm gear is like a cylinder with a spiral gear around it. So it rotates almost perpendicular to the axis of the gear.
Steve: So you have like a bicycle chain but you combine that with an Archimedes screw.
Chatterjee: Yeah, pretty much. And so they made a magnetic worm gear, and for that you would have to have magnetic fields be completely perpendicular to each other and still interact, and I had no idea how that could work. But I realized that if you made a special array called a Hallback array—in a Hallback array basically you have the north-south domains of the magnet rotate 90 degrees for every new magnet. You could put two of those together and have the magnetic fields line up such that you could create a conventional gear, except their gear faces would not touch. And so basically what you could do then is have a much more efficient gear because you'd minimize all that heat and vibrations and noise in the gear system. So what I did is I took refrigerator magnets which were already pre-fabricated in the special array, and I made a linear magnet and a circular magnet. And I tested that against a traditional gear system that I purchased online. And basically I did some kinetic energy calculations where I put the gear rack on an air track to minimize friction between the gear rack and its surface, and the gear on an axel which was mounted on two [metallic] washers to minimize friction there, and basically slipped the gear rack underneath the gear and saw how fast the gear rack was going before it made contact with the gear, how fast it was going afterwards and how fast the gear rotated after contact. And so basically I was able to calculate kinetic energies for the gear rack before and after contact and the gear after contact. So I had two output kinetic energies and one input kinetic energy, and I could calculate efficiencies for both systems.
Steve: And what did you find?
Chatterjee: Basically, my magnetic system, which I engineered, was about 30 percent more efficient than a conventional system. So, I saved 30 percent of the heat loss and energy that came out of a traditional gear system.
Steve: When you say 3 percent, you mean, for example 10 versus 7, but that's a huge difference because that's more like 30 percent.
Chatterjee: Yeah. Exactly like 30 percent of the energy lost by [a] normal system was saved by the system that I made. So I mean it worked out. I was really [happy]. Looking at the data, you wouldn't expect [it], but once I did the analyses, it worked out.
Steve: Do you expect this is going to find some application out there?
Chatterjee: I really...
Steve: It probably already is being tried in industry, but...
Chatterjee: I really hope so. I mean, really, I found a couple of papers on magnetic gears used in
the industry, but they were ridiculously complex. They had iron magnets arranged around a permanent magnet in the center with, like, cylindrical poles everywhere. It was ridiculous diagrams; and I just wanted to go bottom up and take the simpler approach. I mean, really what I want to do in the future is work with these magnets, and with the use of stronger magnets or a more effective sort of array, keep it really simple and just come up with something that is industry compatible, so I could have a cheap gear system, which you can buy, and which is significantly more efficient than the traditional system.
Steve: Very cool. So what's in your future—mechanical engineering, electrical engineering?
Chatterjee: I'm actually looking at two very different fields. I'm looking at nanotechnology and a sort of a nanotechnology application to engineering and efficiency in the field. So I would love to work on nanotech and solar technology. But then again I'm also looking at [biomed and] biomedical engineering, because I really love, sort of, the artificial organ synthesis they are doing with protein scaffolds and on the other hand sort of using genetic recombination to produce insulin and drugs from bacteria.
Steve: Fedja Kadribasic is a student at West Boca Raton Community High School in Boca Raton, Florida.
Kadribasic: I looked at how factors that affect tornado formation have changed over time; because it's a very big problem for scientists to look at how actual tornado numbers have changed over time. Every time a new technology is introduced to study tornadoes, the number of tornadoes that are documented increases. So scientists are not sure if the increase in tornado numbers are due to a technology or due to an actual climate change. I tried to look at the situation from a different approach by looking at certain factors that affect tornadoes and seeing how they have changed over time. Namely these factors are CAPE, which is Convective Available Potential Energy, which is the maximum amount of energy that a rising parcel (which [is] an arbitrary amount of mass) can have as it's rising through the atmosphere; wind shear, which is the difference in wind [vectors] at the top and bottom of the atmosphere; and specific humidity, which is the ratio of water to the total mass of a moist system.
Steve: So we're looking at basically the environmental conditions and the amount of available energy.
Kadribasic: Well, all three of the factors I talked about right now affect tornadoes. One is like the amount of energy that's available to it, another is the amount of, what the wind shear, which is the difference in wind, and the amount of humidity. All three are required for tornadoes to form. The only thing is, it's like, I did two parts of the study, the first of which was to see whether these factors are actually greater in tornadic versus nontornadic areas, because once I established that, I can look whether or not there was a change over time.
Steve: So you must have found that they were greater because your study does have two parts.
Kadribasic: Yes, that's correct. For all three of the factors, the difference was very, very significant with the p values being extremely small; all except for wind shear, with the only reason being was that the control area that was used is mountainous. So in a mountainous area, the wind shear happens to be very large because the amount of the wind at the top versus at the bottom of the mountain can be a very large difference. But since I did establish that these factors are greater in tornadic areas, I could now look at these factors to see if they have changed over time. I looked at averages for all three of them during the month of May, and I ran a test called a linear regression t-test to see how these average factors in the month of May had changed over time. I was only able to use the years 1980 to 2006 though, because that is how far the data goes on these average values. Even though I did not get a statistically significant change, the p values I did get, are so very small, and since my sample size is also very small, we have to look at this information a lot more if we are going to see whether or not there has been [in] fact [a] change over time.
Steve: So you are basically looking at the factors influencing the tornado formation...
Steve: ...and the change in those factors, we're not sure yet whether the tornado increase is a function of environmental conditions changing or our ability to detect them.
Kadribasic: That's basically [what] I tried to do, to see if there has been any change, if there has been any change in these factors, which could mean that there is a change in climate, which could then finally mean that there has been a change in tornado numbers. So it's a very indirect approach, but one that is very applicable to meteorology.
Steve: What was the biggest thing you learned trying to do this work?
Kadribasic: Well, probably the biggest thing I've learned is that it takes a lot of time to collect all the data. Even though there wasn't actual data collecting per se, like you would do in, like, in some other kind of experiments, something in biology or in chemistry, the data that need to be collected, which was from, like, different databases took such a huge amount of time. It took a couple of months actually, just to collect the data. The test themselves took a very little time, but it's just, you have to have, like, a scale. Just, like, research takes really a very long time sometimes, and I suppose you spend years working on it and sometimes they won't even get anything out of it. So [I] just had to keep that in mind when doing research.
Steve: I'm sure any of the graduate students listening can relate to that. So are you interested in becoming a professional atmospheric scientist or where do you see yourself?
Kadribasic: Well, meteorology is pretty interesting, but I would rather pursue some other aspect of the physical sciences; probably thinking physics or mathematics, those are pretty interesting to me. I might pursuit meteorology, but that might be just from, like, [a] standpoint of thermodynamics or something, not like actually looking at weather data, although meteorology is interesting; you know, I never know.
Now it's time to play TOTALL....... Y BOGUS. Here are four science stories; only three are true. See if you know which story is TOTALL....... Y BOGUS.
Story number 1: The astronauts who just serviced the Hubble telescope will testify before congress in Washington on May 21st.
Story number 2: The Maleo is a bird about the size of a chicken but its eggs are about five times larger than chicken eggs.
Story number 3: Cigarette smoke may prevent allergies by lowering the immune reaction.
And story number 4: Good evidence that people were keeping cats as pets much earlier than in ancient Egypt, archaeologists [at] a 9,500-year-old site in Cypress found a house cat and a human buried in adjacent graves.
Time is up.
Story number 4 is true. The cat and the human were buried together and in the same orientation. That's according to a 2004 report from the Paris National Museum of Natural History, which is cited in the new article, "The Evolution of House Cats" in the June issue of Scientific American. You can read it free at our Web site. Just go to www.ScientificAmerican.com/sciencemag.
Story number 2 is true. The chicken-sized Maleo lays a huge egg, which it buries in the sand at its home on the Indonesia[n] island of Sulawesi. The chicks hatch, climb out of the sand and are already able to fly and fend for themselves, probably because they're already fully grown in the giant eggs. Last week the beach the Maleos use to nest became protected habitat. It was bought by a local NGO called Wildlife and Wildlands Conservation, which works with the New York‚Äìbased Wildlife Conservation Society. Here courtesy of WCS is the actual call of the Maleo (voice of Maleo). Sounds like it was laying the giant eggs and saying "Why?‚Äù
And story number 3 is true. Smoking does plenty of bad stuff, but a new study finds that cigarette smoke decreases the allergic response in mice, anyway. Smoke inhibits activity of so called mast cells, which drive the immune system's response to allergens. The report appears in the journal Clinical and Experimental Allergy. Obviously taking up smoking to cure allergies is a bad idea. Maybe you could just run through a smoke filled room.
All of which means that story number 1, about the Hubble's service astronaut team testifying in D.C. on May 21st is TOTALL....... Y BOGUS. They will be testifying before the Senate Appropriations Subcommittee on Commerce, Justice and Science, but they will be doing it from the orbiting shuttle, which isn't scheduled to land till the next day. What's the difference between outer space and a congressional hearing room? Why the hot air of course.
Well that's it for this edition of Scientific American's Science Talk. Check out www.ScientificAmerican.com for the latest science news, including our In-Depth Report on how to build a better engine, which we're gonna have to do with new gas mileage standards coming. For Science Talk, I'm Steve Mirsky. Thanks for clicking on us.
High school scientists Sruti Swaminathan, Maia ten Brink, Alyssa Bailey, Moyukh Chatterjee and Fedja Kadribasic, all winners of state competitions sponsored by the American Junior Academy of Sciences, talk about their research. Plus, we'll test your knowledge of some recent science in the news.