the CDC recommends NOT going to the doctor if you think you have the flu
My uni is obeying this guideline. Have symptoms? Call the student health center and we'll evaluate you over the phone. If we believe you should stay home, we'll email your profs for you.
I suspect we'll see an outbreak round about midterms.
Price are higher now, but you're on the right track.
A colleague at a top-tier university called me up with the exciting news that some educational reform theorist from an Ivy League school had just visited to explain the future of higher education. His ideas included getting professional practitioners to teach courses in their field, holding classes on students' schedule, removing many residency restrictions, etc.
Congratulations, I told him, you just discovered the community college.
A few years ago an old-time participant set fire to "the man" a couple of days early. The organizers decried it as criminal vandalism and reported it to law enforcement.
If I could take a 20% pay cut to work 20% less, I'd do it in a heartbeat.
Have you ever considered academia? No, not a Tier-1 school where the 60-20-20 split is hours/week, but at a state or liberal arts college. You can get a decent 12-month salary for nine months of work. The downside is that you have to deal with people like the girl in TFA, but the summers are *very* nice.
Same here. My wife chose not to work, but she's got a solid degree and is looking at grad school just in case we need her to work. The hardest part is having to compete for goods with families on two incomes. Every one of our neighbors is a two-worker family, which means we're the only house without a sports car and/or boat/jet skis/etc.
And I would love to be the one that stays home. It's a ton of work, but I am jealous of the relationship my wife has with our kids. I have a great friend who was a stay-at-home dad for four years. His wife had a better job, so he did the Mr. Mom thing and was great at it. His kids run to *him* with their problems.
I hate that too, but I wonder how much of the draft I get off that car (or more often truck) is affecting my acceleration when I pass. It will take less power to pull even with the other vehicle than it does to actually pass them. Also, the other car works a little bit harder while you're in their slipstream (ask any cyclist), so when you leave that draft, they'll speed up slightly.
While it's true that there will be false positives, as well as false negatives, you don't convict someone, or have a lung removed, without further testing.
A couple of other examples:
1. A "99% accurate" test for banned substances in a sport where a positive test will result in a two-, four- or eight-year ban. While 99% sounds conclusive, each stage of a major cycling race (like the Tour de France) will included tests for the stage winner, the overall race leader and five other riders chosen at random. 7*21 = 147; my back-of-napkin math says that it's almost a 1-in-4 chance that there will be a false positive or negative during the race. The UCI will double-test before issuing a suspension, but the initial positive test is enough to remove someone from the race. And the second tests are not blind: They are done at the same facility with workers who new the outcome of the original test (sometime the very same technicians performing both tests). And these tests are not like litmus paper; they're more like interpreting sine curves.
2. A recent technical report for the New York State achievement tests in English/Language Arts (Grades 3-8) claimed to have correctly classified students "pass" or "fail" 90% of the time... So 10% of the students are being miscategorized: either failing when they should have passed or passing when they should have failed. In this case, there is no "additional testing."
A 90%+ accuracy in the instrument may sound impressive, but when it's drawn out over the intended breadth of its usage, and compared with the consequences of the results, it's not good enough.
<Sideshow_Bob>Conspiring to violate the Digital Millennium Copyright Act... Now honestly, what is that? Do they give a Nobel Prize for conspiracy chemistry?</Sideshow_Bob>
A professor is someone with a PhD who is tenured at the university in question.
So, a "visiting professor" is not a professor?
Actually, no, it's not. "Visiting Professorships" are a bit like "honorary degrees."
In some place (UK, I think) you're only a "professor" if you fill an endowed chair. Now, here in the US, it differs a little by institution, but typically both tenured and tenure-track faculty (assistant, associate, and full professors) are considered "professors."
But a "visiting professorship" is usually not on par with even an "assistant professorship." For example, many mid-level institutions will hire people who are working on their doctorates (usually ABD: All But Dissertation), but they're given the rank of "visiting professor" and kept off the tenure track. This can mean they have to submit for review each year, rather than every three-years for "assistant professors." Also, "visiting professors" usually don't get a vote in department matters.
Once they're hooded, their the get promoted to "assistant professor," and their tenure clock starts.
A colleague of mine just left a copy of some preliminary research on this issue. Sections A and B of a course were both given the same assignments, submitted homework was scored in both section, but the assignments only counted for credit in Section A. Homework completion rates in Section A ranged from 80% to 90%, but hovered around 20% in Section B.
Here's the kicker: Both sections also had periodic quizzes on the material. There was no difference between the sections on their quiz scores. Grading homework appeared to encourage more students to do the homework, but it did not improve their quiz scores.
Of course, this was a small sample, and more research needs to be done.
So true. I started blogging long before I knew there was a word for it (only a couple of years after the term was coined). I was in college, trying to learn PHP/MySQL, so I coded a quick site that ran a lot like Slashdot. I even put it on Sourceforge, but that project is long dead.
I realized I could blog my class notes, notes from my reading, etc. I could blog my thoughts on current events - which made me think through things much more than I had. You'd be surprised how many times I've searched my own blog for information.
I also could blog to keep in touch with family and friends scattered to the four winds.
I run my own blogging software (either homebrewed or WordPress, etc.) so I owned the data. Sure, I use Facebook, but I would never put too much effort or data into a site someone else owns.
In the end, I blog, but I've never entertained the idea on monetizing it. In fact, I've contributed code, WordPress themes, and free technology advice. So I guess I've done the opposite of trying to turn a profit.
My first CS professor told me that there are nearly infinite ways to accomplish any single task on a computer. He felt this meant computer science and software engineering were arts...
"But we don't call them art because we like paychecks."
I don't do what I do because someone's slapped a name on it. Anytime I've become involved in a "named" movement, I've quickly found big pieces with which I disagree. Now I avoid such associations.
But language compresses reality and co-opting labels that already mean something else only increases that compression. Eventually we lose meaning. Speaking only for myself, I know some people oppose the movements to which I contribute. I don't want those people marching to Washington DC with an article that says, "Yeah, we're socialists." A) It's not true. B) That label carries far too great a negative connotation (whether deserved or not) to be useful.
I've had this conversation with a sociologist from an Ivy League school. I had pointed out some similarities between Osama Bin Laden and Emmanuel Goldstein when she said, "Yes, but remember that Orwell was writing an allegory for Stalinism. It doesn't really apply to our situation."
That response, like the reviewer's comment that the book "misappropriates" the image of Big Brother, left me speechless. It wasn't until later that I realized such arguments are like saying, "Jesus was really only talking about fishing, so don't try to apply those teaching to your life," or "One hand clapping is only in reference to people with one arm," or "Aesop and Jean de la Fontaine were just talking about animals."
Part of the benefit of taking real totalitarianism and moving it into a fictional context (see also Animal Farm) is that we can generalize more freely the components of government control. There are many aspects of Stalinism we wouldn't deign to claim exist in our society, but Orwell gave us a fictional world devoid of the historical context of Stalinism. He gave us words to express our disapproval.
On the Diane Rehm Show on NPR [...] suggestions included government bailouts and subsidies [...] preserve "journalism"
So what you're saying is that journalists are desperately clamoring for government bailout to save their behinds? Shocking.
I really like DR's show, but you have to take anything technical she says with a grain of salt. She brought up "farm-raised whales" last month only to be told it was an April Fool's joke.
Newspapers clamoring for a bailout scares the bejebuses out of me. I wonder how much the kid gloves with which the "4th estate" is treating President Obama is because a) he's popular (his family moves copy), or b) because they want to be the GM and not the Chrysler when some media czar is deciding which papers get to see next year.
IQ scores have an average of 100 and a standard deviation of 15. In this case, both groups scored around 80, which is more than a standard deviation below the mean. For comparison, the difference between the groups was 3 points, or one-fifth of a standard deviation.
3 points is an extremely small difference in IQ scores. In fact, most tests aren't accurate enough to give you a result +/- 3 points for an individual. Imagine you got a score of 108, so above average, and we plot it on a number line:
No competent professional would report your 108 as if it were totally accurate. Instead, they would look in the test manual and find the confidence interval for your score, then *shade in* a region of the number line, like this:
...85-----90-----95-----100-----105XXXXX110XX---115...
^we're 95% confident your score is somewhere in this range.
IIRC, on the PPVT (the test in question) is +/- between 3 and 5 depending on age and score. Other tests, like the BOT-2 can have confidence intervals that are 15 points wide.
Yes, I know this. I edited that post a couple of times for concision. But, since you brought it up, let me explain.
Given that the distribution of the normative sample is known, one could quite easily estimate the necessary sample size for a given difference to be considered significant. This should be done *before* the research took place. The fact that these researchers stated that their no-difference findings may be due to a small sample (which was equal to about 10% of the size of the entire norming sample) hints that they *thought* there would be a bigger difference.
It may actually be quite clever: If you expect a big difference, sample narrowly so that only your expected difference would be significant, while still plausibly dismissing non-significant differences as the result of a small sample.
Also, while you are correct that the specific caveat of sample size wouldn't be applicable if they found a significant difference, there are other important limitations (e.g. external validity) that would have to be discussed. The research I have seen on the topic tends not to bring up those or anything else that could lessen the impact of their conclusions, but only when they find a difference.
I'm a teacher educator, but my background isn't in public school teaching; it's in psychometrics and quantitative psychological research. Because of this, I find my views are vastly different from the faculty around me.
When I see the discrepancy between poor-inner-city-[minority] and well-to-do-suburban-[majority], my science sense starts tingling, and I think, "I wonder if it's because of X, Y, or Z... or maybe something else. What data could I gather to establish or refute some of these connections."
In other words, I tend to think like you do.
My colleagues just scream, "Oppression!" They conduct qualitative "critical analyses," which means they gather data to back up their apriorisms (because "everyone's biased; we're just acknowledging ours and leveraging it"). In the end, the conclusion they formulated before even gathering the data is supported and the view that it is race, not socioeconomic status, becomes accepted as social science "law."
One of my graduate students showed me an article this week which compared the achievement between poor whites and poor blacks on the Peabody Picture Vocabulary Test (which is a popular whipping boy of the anti-standardized test movement). Long story short: There was a whopping 3-point difference between the two groups, and both were 20 points below the national mean. (The PPVT uses an IQ-equivalent scale, so 100 is the mean, with a standard deviation of 15.)
But the researchers concluded that "the sample (N > 200) was too small to generate any meaningful conclusions." I wonder if they would have included the same caveat had they discovered a significant difference.
A physician has a far greater probability of distinguishing bullshit from actual facts than a layperson, though it doesn't of course always hold true.
As a PhD who has spent many evening hours explaining medical journal statistics to my physician brother, I would emphasize that last part. Medical schools in the U.S. train future professionals in the practice of medicine, not research. (I'm not saying one is better than the other.) Once I realized this, I started noticing all sorts of comments from physicians that demonstrated their lack of research training.
It's a real issue with no real solution. Expecting an MD to take more stats and research coursework is really not reasonable. Giving non-medically trained PhDs authority over prescriptions isn't really an option either.
I've taken a class from him. You can sign up to follow his open learning seminar via his blog and wiki. Though his lectures are only on blipTV, he does read everyone's blog and will respond.
I'm also a professor and I find blog-based discussion to be far superior to face-to-face. A few topics require the immediacy of being in person, but many many more conversations are best when each party has the time to think between submitting responses.
However, the headline is taken WAY out of context. This is what he said:
"If universities can't find the will to innovate and adapt to changes in the world around them (what's happening in the economy, affordability, the impacts of technology and openness, etc.)... universities will be irrelevant by 2020."
Wait until this hits the MSM: "Researchers warn: Facebook lowers grades!"
The article was well written and avoided the correlation/causation fallacy that so many socialtech studies (or at least their reports) fall into.
That disconnect between perception and reality does not necessarily mean that Facebook leads to less studying and worse grades -- the grades association could be caused by something else. However, it does raise more questions about how students spend their time outside class on activities such as Facebook, part-time jobs and extracurricular activities.
Slashdot Mirror
It's the opposite with anything Apple. The *current* version is perfect... until the next version comes out... then the older version was crap.
PS - I'm a Mac user too.
My uni is obeying this guideline. Have symptoms? Call the student health center and we'll evaluate you over the phone. If we believe you should stay home, we'll email your profs for you.
I suspect we'll see an outbreak round about midterms.
Price are higher now, but you're on the right track.
A colleague at a top-tier university called me up with the exciting news that some educational reform theorist from an Ivy League school had just visited to explain the future of higher education. His ideas included getting professional practitioners to teach courses in their field, holding classes on students' schedule, removing many residency restrictions, etc.
Congratulations, I told him, you just discovered the community college.
A few years ago an old-time participant set fire to "the man" a couple of days early. The organizers decried it as criminal vandalism and reported it to law enforcement.
The hypocrisy was thick.
Have you ever considered academia? No, not a Tier-1 school where the 60-20-20 split is hours/week, but at a state or liberal arts college. You can get a decent 12-month salary for nine months of work. The downside is that you have to deal with people like the girl in TFA, but the summers are *very* nice.
Same here. My wife chose not to work, but she's got a solid degree and is looking at grad school just in case we need her to work. The hardest part is having to compete for goods with families on two incomes. Every one of our neighbors is a two-worker family, which means we're the only house without a sports car and/or boat/jet skis/etc.
And I would love to be the one that stays home. It's a ton of work, but I am jealous of the relationship my wife has with our kids. I have a great friend who was a stay-at-home dad for four years. His wife had a better job, so he did the Mr. Mom thing and was great at it. His kids run to *him* with their problems.
I hate that too, but I wonder how much of the draft I get off that car (or more often truck) is affecting my acceleration when I pass. It will take less power to pull even with the other vehicle than it does to actually pass them. Also, the other car works a little bit harder while you're in their slipstream (ask any cyclist), so when you leave that draft, they'll speed up slightly.
Yeah, we did run away and for years my little college has been happily using a competitor's product... Until this last year when Bb bought them out.
A couple of other examples:
1. A "99% accurate" test for banned substances in a sport where a positive test will result in a two-, four- or eight-year ban. While 99% sounds conclusive, each stage of a major cycling race (like the Tour de France) will included tests for the stage winner, the overall race leader and five other riders chosen at random. 7*21 = 147; my back-of-napkin math says that it's almost a 1-in-4 chance that there will be a false positive or negative during the race. The UCI will double-test before issuing a suspension, but the initial positive test is enough to remove someone from the race. And the second tests are not blind: They are done at the same facility with workers who new the outcome of the original test (sometime the very same technicians performing both tests). And these tests are not like litmus paper; they're more like interpreting sine curves.
2. A recent technical report for the New York State achievement tests in English/Language Arts (Grades 3-8) claimed to have correctly classified students "pass" or "fail" 90% of the time... So 10% of the students are being miscategorized: either failing when they should have passed or passing when they should have failed. In this case, there is no "additional testing."
A 90%+ accuracy in the instrument may sound impressive, but when it's drawn out over the intended breadth of its usage, and compared with the consequences of the results, it's not good enough.
<Sideshow_Bob>Conspiring to violate the Digital Millennium Copyright Act... Now honestly, what is that? Do they give a Nobel Prize for conspiracy chemistry?</Sideshow_Bob>
Actually, no, it's not. "Visiting Professorships" are a bit like "honorary degrees."
In some place (UK, I think) you're only a "professor" if you fill an endowed chair. Now, here in the US, it differs a little by institution, but typically both tenured and tenure-track faculty (assistant, associate, and full professors) are considered "professors."
But a "visiting professorship" is usually not on par with even an "assistant professorship." For example, many mid-level institutions will hire people who are working on their doctorates (usually ABD: All But Dissertation), but they're given the rank of "visiting professor" and kept off the tenure track. This can mean they have to submit for review each year, rather than every three-years for "assistant professors." Also, "visiting professors" usually don't get a vote in department matters.
Once they're hooded, their the get promoted to "assistant professor," and their tenure clock starts.
A colleague of mine just left a copy of some preliminary research on this issue. Sections A and B of a course were both given the same assignments, submitted homework was scored in both section, but the assignments only counted for credit in Section A. Homework completion rates in Section A ranged from 80% to 90%, but hovered around 20% in Section B.
Here's the kicker: Both sections also had periodic quizzes on the material. There was no difference between the sections on their quiz scores. Grading homework appeared to encourage more students to do the homework, but it did not improve their quiz scores.
Of course, this was a small sample, and more research needs to be done.
So true. I started blogging long before I knew there was a word for it (only a couple of years after the term was coined). I was in college, trying to learn PHP/MySQL, so I coded a quick site that ran a lot like Slashdot. I even put it on Sourceforge, but that project is long dead.
I realized I could blog my class notes, notes from my reading, etc. I could blog my thoughts on current events - which made me think through things much more than I had. You'd be surprised how many times I've searched my own blog for information.
I also could blog to keep in touch with family and friends scattered to the four winds.
I run my own blogging software (either homebrewed or WordPress, etc.) so I owned the data. Sure, I use Facebook, but I would never put too much effort or data into a site someone else owns.
In the end, I blog, but I've never entertained the idea on monetizing it. In fact, I've contributed code, WordPress themes, and free technology advice. So I guess I've done the opposite of trying to turn a profit.
My first CS professor told me that there are nearly infinite ways to accomplish any single task on a computer. He felt this meant computer science and software engineering were arts...
"But we don't call them art because we like paychecks."
Too right.
I don't do what I do because someone's slapped a name on it. Anytime I've become involved in a "named" movement, I've quickly found big pieces with which I disagree. Now I avoid such associations.
But language compresses reality and co-opting labels that already mean something else only increases that compression. Eventually we lose meaning. Speaking only for myself, I know some people oppose the movements to which I contribute. I don't want those people marching to Washington DC with an article that says, "Yeah, we're socialists." A) It's not true. B) That label carries far too great a negative connotation (whether deserved or not) to be useful.
I remember a McDonalds in Bordeaux, France, that printed the code for the bathroom on the receipt. It kept the vagrants and hoodlums out.
I've had this conversation with a sociologist from an Ivy League school. I had pointed out some similarities between Osama Bin Laden and Emmanuel Goldstein when she said, "Yes, but remember that Orwell was writing an allegory for Stalinism. It doesn't really apply to our situation."
That response, like the reviewer's comment that the book "misappropriates" the image of Big Brother, left me speechless. It wasn't until later that I realized such arguments are like saying, "Jesus was really only talking about fishing, so don't try to apply those teaching to your life," or "One hand clapping is only in reference to people with one arm," or "Aesop and Jean de la Fontaine were just talking about animals."
Part of the benefit of taking real totalitarianism and moving it into a fictional context (see also Animal Farm) is that we can generalize more freely the components of government control. There are many aspects of Stalinism we wouldn't deign to claim exist in our society, but Orwell gave us a fictional world devoid of the historical context of Stalinism. He gave us words to express our disapproval.
Kind of an anti-Newspeak.
Google "euphemism treadmill".
I really like DR's show, but you have to take anything technical she says with a grain of salt. She brought up "farm-raised whales" last month only to be told it was an April Fool's joke.
Newspapers clamoring for a bailout scares the bejebuses out of me. I wonder how much the kid gloves with which the "4th estate" is treating President Obama is because a) he's popular (his family moves copy), or b) because they want to be the GM and not the Chrysler when some media czar is deciding which papers get to see next year.
IQ scores have an average of 100 and a standard deviation of 15. In this case, both groups scored around 80, which is more than a standard deviation below the mean. For comparison, the difference between the groups was 3 points, or one-fifth of a standard deviation.
3 points is an extremely small difference in IQ scores. In fact, most tests aren't accurate enough to give you a result +/- 3 points for an individual. Imagine you got a score of 108, so above average, and we plot it on a number line:
No competent professional would report your 108 as if it were totally accurate. Instead, they would look in the test manual and find the confidence interval for your score, then *shade in* a region of the number line, like this:
IIRC, on the PPVT (the test in question) is +/- between 3 and 5 depending on age and score. Other tests, like the BOT-2 can have confidence intervals that are 15 points wide.
So, no, 3 points is not a big difference.
Yes, I know this. I edited that post a couple of times for concision. But, since you brought it up, let me explain.
Given that the distribution of the normative sample is known, one could quite easily estimate the necessary sample size for a given difference to be considered significant. This should be done *before* the research took place. The fact that these researchers stated that their no-difference findings may be due to a small sample (which was equal to about 10% of the size of the entire norming sample) hints that they *thought* there would be a bigger difference.
It may actually be quite clever: If you expect a big difference, sample narrowly so that only your expected difference would be significant, while still plausibly dismissing non-significant differences as the result of a small sample.
Also, while you are correct that the specific caveat of sample size wouldn't be applicable if they found a significant difference, there are other important limitations (e.g. external validity) that would have to be discussed. The research I have seen on the topic tends not to bring up those or anything else that could lessen the impact of their conclusions, but only when they find a difference.
Thanks for keeping me honest on it.
I'm a teacher educator, but my background isn't in public school teaching; it's in psychometrics and quantitative psychological research. Because of this, I find my views are vastly different from the faculty around me.
When I see the discrepancy between poor-inner-city-[minority] and well-to-do-suburban-[majority], my science sense starts tingling, and I think, "I wonder if it's because of X, Y, or Z... or maybe something else. What data could I gather to establish or refute some of these connections."
In other words, I tend to think like you do.
My colleagues just scream, "Oppression!" They conduct qualitative "critical analyses," which means they gather data to back up their apriorisms (because "everyone's biased; we're just acknowledging ours and leveraging it"). In the end, the conclusion they formulated before even gathering the data is supported and the view that it is race, not socioeconomic status, becomes accepted as social science "law."
One of my graduate students showed me an article this week which compared the achievement between poor whites and poor blacks on the Peabody Picture Vocabulary Test (which is a popular whipping boy of the anti-standardized test movement). Long story short: There was a whopping 3-point difference between the two groups, and both were 20 points below the national mean. (The PPVT uses an IQ-equivalent scale, so 100 is the mean, with a standard deviation of 15.)
But the researchers concluded that "the sample (N > 200) was too small to generate any meaningful conclusions." I wonder if they would have included the same caveat had they discovered a significant difference.
As a PhD who has spent many evening hours explaining medical journal statistics to my physician brother, I would emphasize that last part. Medical schools in the U.S. train future professionals in the practice of medicine, not research. (I'm not saying one is better than the other.) Once I realized this, I started noticing all sorts of comments from physicians that demonstrated their lack of research training.
It's a real issue with no real solution. Expecting an MD to take more stats and research coursework is really not reasonable. Giving non-medically trained PhDs authority over prescriptions isn't really an option either.
I've taken a class from him. You can sign up to follow his open learning seminar via his blog and wiki. Though his lectures are only on blipTV, he does read everyone's blog and will respond.
I'm also a professor and I find blog-based discussion to be far superior to face-to-face. A few topics require the immediacy of being in person, but many many more conversations are best when each party has the time to think between submitting responses.
However, the headline is taken WAY out of context. This is what he said:
"If universities can't find the will to innovate and adapt to changes in the world around them (what's happening in the economy, affordability, the impacts of technology and openness, etc.)... universities will be irrelevant by 2020."
Cited from his blog
Wait until this hits the MSM: "Researchers warn: Facebook lowers grades!"
The article was well written and avoided the correlation/causation fallacy that so many socialtech studies (or at least their reports) fall into.