I had to laugh when I caught sight of this chart earlier this evening
That said, it’s still not as funny as “Figure 1” in George Yule’s classic work from 1926, “￼Why do we Sometimes get Nonsense-Correlations between Time-Series?–A Study in Sampling and the Nature of Time-Series”. Click this link – I won’t spoil it for you.
In my opinion, what’s even more dangerous is the abuse of R2. In research utilizing large data sets (most notoriously time-series), you’ll often see relationships between sets quantified in terms of this metric. Not a few times I’ve heard students say that they ran a regression and got a “good R2“. This begs the questions of what is meant by “good” and what R2 really signifies. My guidance was always to have them start by plotting the data, taking a lesson from Anscombe’s quartet. But that guidance isn’t very practical if one is evaluating thousands of time series, and it’s ultimately susceptible to the same error as the aforementioned two figures: attributing “causation” to “correlation”. The figures of Yule and chartporn.org are problematic because they look so convincing. If a correlation is what you’re looking for, you’ll be tempted to stop analyzing your data when you see that “cluster” of data points.
For this reason, I recommend working with a statistician when you’re formulating experimental designs. This requires a bit of patience on both sides, because the experimentalist needs to be able to accept input and the statistician may need to be educated about the nature of the experiment. There will probably be an initial delay in your work. But if you can establish a good relationship built on trust and respect (and hopefully, fun!), I guarantee it will pay scientific dividends!
I just saw this at Instapundit:
In 1900, some divers found the wreck of a Roman vessel off the Greek island of Antikythera. Among the other treasures remanded to the Greek government was an unassuming corroded lump. Some time later, the lump fell apart, revealing a damaged machine of unknown purpose, with some large gears and many smaller cogs, plus a few engraved words in Greek. Early studies suggested it was some type of astronomical time-keeping device – researcher Derek J. de Solla Price laid the groundwork by establishing initial tooth counts and suggesting that the device followed the Metonic cycle, a 235-month pattern commonly used to predict eclipses in the ancient world.
read the whole thing because it gets even more interesting, especially if you enjoy science, technology, engineering & math.
I recently read an excellent article at salary.com for college students (and their parents) who are struggling to choose their major.
If you are still in the process of choosing a college, you should probably have your major figured out before you sign on the dotted line. Each year at a college or university is going to cost you tens of thousands of dollars, and that debt can really cut into your quality of life if the job you get upon graduation (assuming you can get a job) has a relatively low salary. It may force you to delay marriage or starting a family, or worse, cripple you financially for the rest of your life.
The salary.com article focuses on the concept of “ROI” or “Return on Investment”. If you don’t know what this is, read the wiki. Short version: How much money will I get if I invest “x” dollars in a college education?. It had better be bigger than “x”
The article is a bit more negative than I’d like in that it focuses on “bad” majors rather than “good” ones. But if money is an object for you (or your college-aged child), I highly recommend reading it. I think it’s more constructive to apply the methodology to “good” majors to select the one that can meet your desires for the future (e.g. location, type of home you want, marriage or the decision to have kids or not, or how many!).
Many people say that what’s important is that you have a job you like. I say you will like the job that you can do well. Just be honest with yourself about what you can do, and choose wisely so that you can reach your personal goals beyond what your title will be when you graduate! Good luck!
Quantis has a blog addressing the visualization of the results of life cycle assessments (LCA). Well worth reading for those working on LCA, or anyone else looking to improve their data visualization skills. Check it out!
Today I read an article at the “Daily Beast” by Megan McArdle that really struck a nerve, entitled “The Absurd Lies of College Admissions”. It begins…
A high school student has penned an open letter to the colleges that rejected her, published in the Wall Street Journal.
“I also probably should have started a fake charity. Providing veterinary services for homeless people’s pets. Collecting donations for the underprivileged chimpanzees of the Congo. Raising awareness for Chapped-Lips-in-the-Winter Syndrome. Fun-runs, dance-a-thons, bake sales—as long as you’re using someone else’s misfortunes to try to propel yourself into the Ivy League, you’re golden.”
I highly recommend that you read the whole thing.
I don’t agree with everything this high school student wrote – for instance, the type of thing she describes is more characteristic of some Ivies than others – but there is a lot of this at certain Ivies and one in particular that I won’t name.
For that matter, I don’t agree with everything Megan McArdle added. But I think there is an important thing for any high school student to consider: What do you want to get out of college, and do I need a degree from an “elite” school to get it? For instance, do you want to own a home? Get married? Raise kids and be able to pay for their weddings or college degrees should they be interested in going? Do you want to create things? Will people pay for the things that you create?
And will the label on your diploma affect any of these greater goals? Because after four years, nobody cares where you got your degree. An engineer with a B.S. from Princeton or Cornell will work side by side with engineers from UT Austin, Penn State, Rutgers, RPI and Georgia Tech, make the same amount of money and have the same career track.
If you don’t have those greater goals that allow you to calculate a “return on investment” for your college – and more importantly, major of choice – you may have bigger long term problems than concocting fantastic stories about extracurricular activities. Keep your eyes on the prize!
Deterministic versus stochastic modelling in biochemistry and systems biology will be available on March 28, 2013. If you plan to offer a course on bioinformatics, computational biology, or systems biology in the fall semester or quarter of 2013, you’ll have plenty of time to review the text and get an order in to your campus bookstore.
The book is well structured for a single-semester course, with about a week devoted to the material in each chapter. If you have questions about the contents of the book and want to know how you can use it with your existing course plan, please don’t hesitate to contact me.
I’m pleased to announce that my book with Paola Lecca and Ferenc Jordan (of the Center for Computational and Systems Biology (COSBi)) is now available for pre-order!
Paola, Ferenc and I wrote Deterministic Versus Stochastic Modelling in Biochemistry and Systems Biology to target advanced undergraduate students and graduate students. The book bridges biology, physical chemistry and computer science – filling the gaps that may be experienced by students of classical disciplines as they enter the exciting interdisciplinary field of systems biology. The book includes “hands on” examples including Matlab examples for students with limited programming or simulation experience.
A few weeks old, but the facts don’t change much in that time: Why does U.S. fail in science education? (from the Pittsburgh Post Gazette). I highly recommend that you read the whole thing. Here’s the highlight:
One of the things that strikes me about these responses is that they may be attributed to marketing: astrology is obvious, but the belief that antibiotics target viruses is, to my impression, a result of marketing approaches taken for household cleaners. Then, it may be a consequence of a lack of knowledge: perhaps many people don’t realize that viruses, fungi and bacteria are completely distinct forms of life?
Let’s assume my hypothesis is correct: perhaps it works both ways? Consider the sound vs light question. About 14% thought sound moves faster than light. Why such a small fraction? I’d guess that it’s because just about every sci-fi show has spaceships that travel “faster than light”. Who even thinks about the speed of sound anymore? As I watched “The Right Stuff” recently on Netflix, I couldn’t help but think “Oh, how quaint” as the test pilots were romanticizing the sound barrier.
The article proposes that the so-called “science gap” results in part from motivational differences, patience in learning, etc. That may be true. But it seems to me that people are likely to believe in astrology, etc. regardless of their motivation or scientific training if they’re accepting the marketing.
So there are really two issues: erroneous beliefs due to marketing of one variety or another, and science education, with an undefined overlap. If you want better numbers in the chart above, you need better science marketing, not just better teachers or students.
Fascinating and timely story about the Chernobyl disaster, 25 years ago today: Japan can learn from Chernobyl nuclear reactor disaster
Franciscan University of Steubenville physics professor Alexander Sich did extensive research at the site of the Chernobyl nuclear reactor disaster.
Alexander Sich’s long-standing goal has been to dispel myths surrounding the 1986 Chernobyl nuclear power plant accident.
Japan’s Fukushima Daiichi nuclear power plant accident in March revived many of those myths, including claims the Soviet Union ended the Chernobyl meltdown in Ukraine by burying exposed nuclear fuel under concrete and other materials.
Mr. Sich knows firsthand that didn’t happen.
Read the whole thing!
I love creative graphical representations of data and am a bit of Tufte disciple, I have to admit. If you are too, and you use Linkedin, try this out: http://inmaps.linkedinlabs.com/.
(Click for the full image)