The major problem is really poor reporting on science research. This is indeed a major problem. But it doesn't have much to do with this article, which discusses poor analysis by scientists of their own data, more or less. When a scientists can collect data, fail to understand what can be learned from it, misreport it, and have that misreporting compounded by more misreporting by journalists, then you have... well, the world as we know it.
Shave a little off the kilogram reference, everyone who measures their weight in kilos gains a little. US residents are largely unaffected, and it helps squelch stories about the American obesity epidemic. I'll bet if you turn the Secretary of Health and Human Services upside-down, 50 micrograms of metal shavings drops right to the floor.
If publishers are really contributing nothing to academic publishing, and just charge high prices and force you to sign away your rights (which I think is a fair characterization), here's a crazy idea: stop publishing through them! Set up your own journals and charge nothing or a token amount for access. If scientists are so bigoted they only deign to acknowledge work published in overpriced, unnecessary, exploitative publishers' journals, the problem is on the scientists' end. In many (probably most or all) fields, the most prestigious journals are commercial journals. Do you really want to be the first scientist on your block to stop publishing in reputable journals because you don't like their terms? It's a huge oversimplification to say the problem is on the scientists' end. These journals became prestigious back when there were no open alternatives. Now they're entrenched. It's difficult to get an entire community to change overnight. And if it doesn't change overnight, then someone has to be first to take that risk with an article they'd really like to disseminate to the community. Best case, you get a bunch of big shots to start as bunch of open journals. In fields with a lot of journals to choose from, it would take many many such efforts to make a dent.
I don't have an investment in pointing fingers, but you're giving out bad advice. In many fields, telling someone to "stop publishing through them" is equivalent to saying, "go into a different field." I have a hard time seeing the fact that few scientists take this advice as a problem with scientists. It's actually a difficult problem to solve, and one that requires a concerted effort from people who are largely ambivalent.
The situation it's supposed to be is that you patent the ALGORITHM. Indeed. An algorithm is just a specific kind of idea (but since patents happen to apply to many kinds of ideas, we usually use the more generic term).
But that's really been forgotten at this point Not sure I can endorse that. Among Slashdot readers, anyway, there's a lot of outrage over the fact that you can patent algorithms. Not sure how much outrage there is over the patenting of other kinds of ideas.
It sounds like you're confusing patents and copyrights. Setting aside the issue of whether or not software patents make sense, this is a big part of the reason why someone would want a software patent as opposed to a copyright. You can copyright the code, but not the idea. But you can patent the idea (basically), as long as the idea has been implemented (as code, in this case). This is a ridiculous dichotomy, but one we're stuck with in the US for the moment.
Like it or not, copyright law in the US reserves certain rights for authors, musicians, etc. The fact that nobody's sure how many people downloaded the song you made available doesn't make it legal, it just makes it more difficult to estimate damages. You also can't print up 1000 copies of a popular CD and leave them on a street corner.
Who's hiring these recruiters? I may have all the qualities of an expert developer, but I'll never in a million years get even a sniff if I don't have all the checkboxes in order. Recruiters don't care if your resume screams out that you can be idiomatic in a new language or system within a few weeks. They'd far prefer you have a ten year history of making the same pinhead mistakes over and over. The attitudes of recruiters reflect the desires of the company, whether they're implicit or explicit. Companies that have trouble finding expert programmers are just lazy.
You tell yourself that. But it's BS. Poker, when it comes down to it, is all about a) statistics, and b) luck. Is there a psychological component to it? Sure. But I'll bet dollars to donuts those aspects are greatly outweighed by luck and a given player's ability to evaluate the statistics on a given hand. One of the chief reasons there are winning poker players is that there are lots of players out there who are willing to bet dollars to donuts without knowing what they're talking about. There are experienced poker players who would agree with you. They're usually pretty bitter, because they can't understand why despite having learned the statistics and having played a large enough number of hands for their skill to win out, they're long-term losers to those of us who've taken the time to understand it better.
In limit games against unskilled opponents, you're right. In other games, the psychology is much more important. And in fact, if you want to do the probabilities right, you need the psychology. There's almost no hand of interest you can analyze properly without an estimate of some quantity like "the probability this bozo would make that raise in this situation." Is it statistical analysis or psychology? Is it the sugar or the stirring?
If I had mod points left, I'd mod that as funny. Yeah, well. I wasn't trying to be funny, I just didn't feel like getting into an argument by stating the obvious. For all I know this guy lives somewhere that has much better stuff than I can get here.
FYI, I did read the original article, and I've read many more on the subject (not Slashdot articles, real articles). I didn't think I would have to spell this out in such excruciating detail to make this trivial and obvious point, but I'll try again.
I am considering getting FIOS. I may get it, I may not. I'm still deciding.
It's not clear to me if the local Verizon guys will let me keep copper for my phones. They may, they may not. I've heard stories about it going both ways.
Your previous post expressed incredulity that anyone would want to keep copper ("Seriously?... Cell phones and VoIP are making POTS a thing of the past.").
Cell phones and VOIP are inadequate where I live (outside Philadelphia, which is a large-ish city in the US), and probably in many other areas, due to coverage, quality, and reliability problems.
Therefore, it's important to me to know if I can keep copper before I commit to FIOS. I don't know for sure if there's any other way to have reliable high-quality phone service in my house, and I happen to need reliable phone service with better quality than what I get over my cell phone. Reliability of non-copper Verizon phone service in our area is not yet known, reliability of copper appears to be as close to perfect as I could hope for.
I'm glad you're happy without a land line, but I have legitimate concerns about increased phone downtime if they take out the copper. Downtime with our current broadband provider has been extremely bad. They blame squirrels, I don't happen to know if the squirrels like Verizon cables too.
Cell phones and VOIP are only making POTS a thing of the past for people who have good cell or internet service in their house. I have neither, so this is a real issue for me. If I can't talk Verizon into leaving my copper alone, it'll be a real gut check.
It's a shame this guy went to the effort of creating this blog post without making any effort to involve useful metrics of how informative a test really is. The words sensitivity, specificity, and variance don't come up at all. There is a kernel of truth here, which is that you can have both noisy items (not informative about the taker's knowledge) and informative items on tests, and hard tests tend to have more noisy items. The author seems to miss the point that two-choice items at which students guess maximize the error variance. In other words, he chooses the best possible case to support his argument, even thought it's unrealistic. Five-choice guessing items contribute less, although it depends how the items are structured (if three of the choices are easily eliminated by even the worst students, then it's much closer to a two-choice item). As a thought experiment, if there were a million choices per "hard" item, they would contribute almost no variance to test scores. The article seems to make no reference to the true score variance among the test takers, which is obviously critical.
I would have liked to see an analysis of the relationship between the number of plausible choices per question and the probability of mis-ordering two test-takers (giving the less knowledgeable a higher score). That would have been a lot more informative than simply saying, essentially, "two is bad, you do the math for more -- but trust me, it's a mathematical certainty."
There is a kernel of truth here, that multiple-choice tests are often not that sensitive, and that when everyone is guessing on an item, it contributes only noise to the measure. At issue really is how much variance in the test score is explainable by knowledge. In other words, how much information is contained in the test score. An article that uses phrases like "mathematical certainty" and "complete fraud" is obligated to provide some legitimate analysis, or at least references to the literature, not just anecdotes.
When I first heard about this, I thought it was great news. Independent music labels and individual artists are still free to license their music to internet radio stations at lower prices, right? The high rates set by the CRB seem to basically mean that major labels will not have access to internet radio, while independent labels and musicians will (if they want). Although I like the occasional major label tune as much as the next guy, it's easy to see a major upside to this move. Through one channel at least, the RIAA not only loses its virtual monopoly, it loses access almost completely. As internet radio grows, this could turn out to be a huge boon to anyone who cares about independently produced music. Am I missing something?
There's no question that fMRI researchers have an ugly history of reinventing old mistakes. But I don't know that a lack of training in psychometrics is the problem. More to the point, is it really true that "going from a scientific explanatory mode, where you have potentially large samples and budgets and cooperative subjects, to prediction of individual behavior is a huge leap?"
Well, sort of. My impression is that this has little to do with a lack of training in psychometrics, but a lot to do with the more general problem, evident to anyone who reviews the occasional fMRI article, that researchers like to make unquantified (or improperly quantified) observations. Most often the data are there, just not analyzed properly. This is really just a basic issue with the use and reporting of inferential statistics.
That said, I don't honestly see that it's a big issue here. It seems like the authors did something sort of reasonable and drew mostly reasonable conclusions (I say this without having given it the close reading I reserve for research I really care about). My sense is that the desire to overextend the results is coming more from the reporting of the article and less from the reporting in the article. In other words, it's not clear to me who needs the training in psychometrics.
Adding and subtracting is "high-level" intellectual activity, now?
Most studies of this kind of thing use simple motor tasks, which are comparatively concrete, low-level, and have a much better understood neural substrate. It depends a little on which psychology/neuroscience subculture you're talking to, but "high-level" is often used to mean something along the lines of "stuff your dog can't do." This is different from the ordinary meaning of the phrase, meaning roughly: "comprehensible only to Susan Sontag."
The authors aren't to blame for using the phrase in a way that is well-understood by their peers. But no science reporter is going to get into this level of detail. No science reporter with a deadline, anyway. Unfortunately, reporting this kind of article without explaining that "high-level" means something different to neuroscientists than it does to everyone else is liable to be at least a little misleading (in the sense that it encourages a predictable misunderstanding).
Can anyone translate the difficulty of those crossword puzzles into NYT crossword equivalents? In my view, anyone who can reliably solve an entire NYT Saturday crossword completely and without errors in 12 minutes is a freak of nature. Anyone who can't do it for a Monday crossword (give or take the occasional botched square) is an idiot. Would I have had a chance at Bletchley park if I sometimes have trouble finishing (almost afraid to admit it) Thursdays? Does that make me hopeless for an IT career (I already have a career, so I'm just asking out of idle curiosity)?
This is a terrible example. Of course you don't get arrested for breaking US laws if the US isn't involved. This guy would never have been arrested if US citizens weren't placing bets on his site. Furthermore, it's not completely clear where the crime was committed. The wire act is written in technologically outdated terms, but conceivably the crime includes the transmission of wagering information, which required two-way communications between the web site (in the UK) and the computers of customers in the US. Did he commit a crime while on US soil? No. But it was the technological equivalent of sitting ten feet past the border and lobbing molotov cocktails into the US. It's hard to fathom why it would surprise anyone that he got arrested.
In your example, of course you don't get arrested for crapping in your own country on Sunday, even if it's illegal somewhere. It doesn't involve Elbonia in any way, which is absolutely not the case with online gambling sites.
Incidentally, I love online gambling sites, and I think they should be legal in the US (but perhaps regulated). The wire act is an asinine way to pretend that they've already been outlawed. That decision has not yet been made.
Those are good subjects to learn. There are also many hundreds of other important subjects to learn, so choose carefully. I disagree that theoreticians write the most elegant, efficient code, but there's no point arguing this point here for the zillionth time. Knowing more is better than knowing less, so in addition to the many other things you will need to learn to write good code (the most important of which is the domain in which the code operates, which usually has nothing to do with computer science), you may as well learn discrete math and automata theory, among a dozen or so other core topics in computer science and numerous important topics outside the field.
The advice about English is also important (although it probably wasn't a good idea to lead into that point by misusing the word "simplistic"). If you can write a coherent, accurate paragraph on anything remotely complex, then people with money will fund your projects, and not those of the 90% of your colleagues who are borderline illiterate. (This is not specific to computer science.)
The article offers no details, so it's impossible to evaluate this work meaningfully. Based on the limited details provided, it would be really easy for anyone with even the least background in research or statistics to think up a zillion reasons to be skeptical. But research almost always sound simplistic and inept when it's subject to simplistic and inept reporting. This brief report reads like the confused ramblings of someone who's overheard a conversation in an elevator.
Fortunately, a kind Slashdot reader has posted a link to the article, and that clears up a lot of the mis-reporting. I took a brief look to check two things: what they did to avoid confounding due to other potential risk factors that might be correlated with cell phone use, and what the evidence is for increased risk on the side of phone use. It seems like they didn't do much to avoid confounding. If heavy cell phone users are also heavy drug users, technophiles, work in high-stress occupations, or whatever, it seems like it could confound the overall rates. It would be a little harder to cook up a story for the ipsi/contra difference. The evidence that the tumors are more likely on the ipsilateral side seems to be numerical, not statistical (they do the stats on one side and again on the other side, but they don't compare the two directly, unless I somehow missed it). It would be nice to know how reliable that difference is, since that seems to be the strongest link between the tumor rates and cell phone use. But that's just my first impression.
I saw a tongue-in-cheek poster at the Society for Neuroscience a few years ago, in which the authors compared portrayals of different professions in a large number of movies. Overall, the most negatively portrayed profession was murderer, and scientists were right in there at #2. The methods employed for this survey involved beer and pizza.
The average person in this country couldn't even begin to tell you what science is, what it's useful for, or what scientists do. To be fair, it's not a question with a simple answer like 42. But it's not surprising that people who make policy decisions at all levels of government know nothing whatsoever about science. It's mis-portrayed almost completely in the media, and probably mis-taught at all levels of education. Scientists are not valued by society in any meaningful way.
Any scientist whose work is in the popular press probably has a story about how their work was portrayed in a way to mislead, not inform people. Perhaps someone will repost the link to that recent insightful article about how few science reporters have any science background.
The government has been rewriting science more blatantly in environmental sciences than in other areas. But it's the other kind of rewriting that's more insidious and harmful. Necessarily, most science funding comes from the government. They decide what to fund and what not to fund. Serious scientists get input into this decision, but not the last word. What's insidious about it is that no individual scientist is doing what they do because the government told them. But since there's such an oversupply of scientists, including a healthy supply interested for their own reasons in doing the specific things the government would like, the government can shape science to whatever extent they want without there ever being a single scientist who was specifically influenced.
I don't find that so objectionable, but the article portrays it as though they're repudiating my favorite graphical email client, which really seems to overstate their real claims.
Memory deficits are a risk of epilepsy surgery. As yet there's no truly reliable way to predict post-surgical memory problems, but since surgery is generally a last resort, it's a risk the patients have to take. This kind of memory problem is also typical of Wernicke-Korsakoff's Syndrome, dramatized in the second chapter of Oliver Sacks's "The Man who Mistook his Wife for a Hat," which is often a consequence of long drinking binges (and an accompanying vitamin deficiency, I think). You don't always see the truly dense amnesia, but when you do it's striking.
Slashdot Mirror
Shave a little off the kilogram reference, everyone who measures their weight in kilos gains a little. US residents are largely unaffected, and it helps squelch stories about the American obesity epidemic. I'll bet if you turn the Secretary of Health and Human Services upside-down, 50 micrograms of metal shavings drops right to the floor.
If publishers are really contributing nothing to academic publishing, and just charge high prices and force you to sign away your rights (which I think is a fair characterization), here's a crazy idea: stop publishing through them! Set up your own journals and charge nothing or a token amount for access. If scientists are so bigoted they only deign to acknowledge work published in overpriced, unnecessary, exploitative publishers' journals, the problem is on the scientists' end. In many (probably most or all) fields, the most prestigious journals are commercial journals. Do you really want to be the first scientist on your block to stop publishing in reputable journals because you don't like their terms? It's a huge oversimplification to say the problem is on the scientists' end. These journals became prestigious back when there were no open alternatives. Now they're entrenched. It's difficult to get an entire community to change overnight. And if it doesn't change overnight, then someone has to be first to take that risk with an article they'd really like to disseminate to the community. Best case, you get a bunch of big shots to start as bunch of open journals. In fields with a lot of journals to choose from, it would take many many such efforts to make a dent.
I don't have an investment in pointing fingers, but you're giving out bad advice. In many fields, telling someone to "stop publishing through them" is equivalent to saying, "go into a different field." I have a hard time seeing the fact that few scientists take this advice as a problem with scientists. It's actually a difficult problem to solve, and one that requires a concerted effort from people who are largely ambivalent.
It sounds like you're confusing patents and copyrights. Setting aside the issue of whether or not software patents make sense, this is a big part of the reason why someone would want a software patent as opposed to a copyright. You can copyright the code, but not the idea. But you can patent the idea (basically), as long as the idea has been implemented (as code, in this case). This is a ridiculous dichotomy, but one we're stuck with in the US for the moment.
Like it or not, copyright law in the US reserves certain rights for authors, musicians, etc. The fact that nobody's sure how many people downloaded the song you made available doesn't make it legal, it just makes it more difficult to estimate damages. You also can't print up 1000 copies of a popular CD and leave them on a street corner.
Who's hiring these recruiters? I may have all the qualities of an expert developer, but I'll never in a million years get even a sniff if I don't have all the checkboxes in order. Recruiters don't care if your resume screams out that you can be idiomatic in a new language or system within a few weeks. They'd far prefer you have a ten year history of making the same pinhead mistakes over and over. The attitudes of recruiters reflect the desires of the company, whether they're implicit or explicit. Companies that have trouble finding expert programmers are just lazy.
In limit games against unskilled opponents, you're right. In other games, the psychology is much more important. And in fact, if you want to do the probabilities right, you need the psychology. There's almost no hand of interest you can analyze properly without an estimate of some quantity like "the probability this bozo would make that raise in this situation." Is it statistical analysis or psychology? Is it the sugar or the stirring?
...so google hired an infinite number of monkeys instead.
FYI, I did read the original article, and I've read many more on the subject (not Slashdot articles, real articles). I didn't think I would have to spell this out in such excruciating detail to make this trivial and obvious point, but I'll try again.
... Cell phones and VoIP are making POTS a thing of the past.").
I am considering getting FIOS. I may get it, I may not. I'm still deciding.
It's not clear to me if the local Verizon guys will let me keep copper for my phones. They may, they may not. I've heard stories about it going both ways.
Your previous post expressed incredulity that anyone would want to keep copper ("Seriously?
Cell phones and VOIP are inadequate where I live (outside Philadelphia, which is a large-ish city in the US), and probably in many other areas, due to coverage, quality, and reliability problems.
Therefore, it's important to me to know if I can keep copper before I commit to FIOS. I don't know for sure if there's any other way to have reliable high-quality phone service in my house, and I happen to need reliable phone service with better quality than what I get over my cell phone. Reliability of non-copper Verizon phone service in our area is not yet known, reliability of copper appears to be as close to perfect as I could hope for.
I'm glad you're happy without a land line, but I have legitimate concerns about increased phone downtime if they take out the copper. Downtime with our current broadband provider has been extremely bad. They blame squirrels, I don't happen to know if the squirrels like Verizon cables too.
Cell phones and VOIP are only making POTS a thing of the past for people who have good cell or internet service in their house. I have neither, so this is a real issue for me. If I can't talk Verizon into leaving my copper alone, it'll be a real gut check.
It's a shame this guy went to the effort of creating this blog post without making any effort to involve useful metrics of how informative a test really is. The words sensitivity, specificity, and variance don't come up at all. There is a kernel of truth here, which is that you can have both noisy items (not informative about the taker's knowledge) and informative items on tests, and hard tests tend to have more noisy items. The author seems to miss the point that two-choice items at which students guess maximize the error variance. In other words, he chooses the best possible case to support his argument, even thought it's unrealistic. Five-choice guessing items contribute less, although it depends how the items are structured (if three of the choices are easily eliminated by even the worst students, then it's much closer to a two-choice item). As a thought experiment, if there were a million choices per "hard" item, they would contribute almost no variance to test scores. The article seems to make no reference to the true score variance among the test takers, which is obviously critical.
I would have liked to see an analysis of the relationship between the number of plausible choices per question and the probability of mis-ordering two test-takers (giving the less knowledgeable a higher score). That would have been a lot more informative than simply saying, essentially, "two is bad, you do the math for more -- but trust me, it's a mathematical certainty."
There is a kernel of truth here, that multiple-choice tests are often not that sensitive, and that when everyone is guessing on an item, it contributes only noise to the measure. At issue really is how much variance in the test score is explainable by knowledge. In other words, how much information is contained in the test score. An article that uses phrases like "mathematical certainty" and "complete fraud" is obligated to provide some legitimate analysis, or at least references to the literature, not just anecdotes.
When I first heard about this, I thought it was great news. Independent music labels and individual artists are still free to license their music to internet radio stations at lower prices, right? The high rates set by the CRB seem to basically mean that major labels will not have access to internet radio, while independent labels and musicians will (if they want). Although I like the occasional major label tune as much as the next guy, it's easy to see a major upside to this move. Through one channel at least, the RIAA not only loses its virtual monopoly, it loses access almost completely. As internet radio grows, this could turn out to be a huge boon to anyone who cares about independently produced music. Am I missing something?
There's no question that fMRI researchers have an ugly history of reinventing old mistakes. But I don't know that a lack of training in psychometrics is the problem. More to the point, is it really true that "going from a scientific explanatory mode, where you have potentially large samples and budgets and cooperative subjects, to prediction of individual behavior is a huge leap?"
Well, sort of. My impression is that this has little to do with a lack of training in psychometrics, but a lot to do with the more general problem, evident to anyone who reviews the occasional fMRI article, that researchers like to make unquantified (or improperly quantified) observations. Most often the data are there, just not analyzed properly. This is really just a basic issue with the use and reporting of inferential statistics.
That said, I don't honestly see that it's a big issue here. It seems like the authors did something sort of reasonable and drew mostly reasonable conclusions (I say this without having given it the close reading I reserve for research I really care about). My sense is that the desire to overextend the results is coming more from the reporting of the article and less from the reporting in the article. In other words, it's not clear to me who needs the training in psychometrics.
Can anyone translate the difficulty of those crossword puzzles into NYT crossword equivalents? In my view, anyone who can reliably solve an entire NYT Saturday crossword completely and without errors in 12 minutes is a freak of nature. Anyone who can't do it for a Monday crossword (give or take the occasional botched square) is an idiot. Would I have had a chance at Bletchley park if I sometimes have trouble finishing (almost afraid to admit it) Thursdays? Does that make me hopeless for an IT career (I already have a career, so I'm just asking out of idle curiosity)?
during the last two years, they will vaporize the plant itself
This is a terrible example. Of course you don't get arrested for breaking US laws if the US isn't involved. This guy would never have been arrested if US citizens weren't placing bets on his site. Furthermore, it's not completely clear where the crime was committed. The wire act is written in technologically outdated terms, but conceivably the crime includes the transmission of wagering information, which required two-way communications between the web site (in the UK) and the computers of customers in the US. Did he commit a crime while on US soil? No. But it was the technological equivalent of sitting ten feet past the border and lobbing molotov cocktails into the US. It's hard to fathom why it would surprise anyone that he got arrested.
In your example, of course you don't get arrested for crapping in your own country on Sunday, even if it's illegal somewhere. It doesn't involve Elbonia in any way, which is absolutely not the case with online gambling sites.
Incidentally, I love online gambling sites, and I think they should be legal in the US (but perhaps regulated). The wire act is an asinine way to pretend that they've already been outlawed. That decision has not yet been made.
Those are good subjects to learn. There are also many hundreds of other important subjects to learn, so choose carefully. I disagree that theoreticians write the most elegant, efficient code, but there's no point arguing this point here for the zillionth time. Knowing more is better than knowing less, so in addition to the many other things you will need to learn to write good code (the most important of which is the domain in which the code operates, which usually has nothing to do with computer science), you may as well learn discrete math and automata theory, among a dozen or so other core topics in computer science and numerous important topics outside the field.
The advice about English is also important (although it probably wasn't a good idea to lead into that point by misusing the word "simplistic"). If you can write a coherent, accurate paragraph on anything remotely complex, then people with money will fund your projects, and not those of the 90% of your colleagues who are borderline illiterate. (This is not specific to computer science.)
Your right! It's hard to bare all the miss steaks people right on this sight, the new firefox will help a lot!
The article offers no details, so it's impossible to evaluate this work meaningfully. Based on the limited details provided, it would be really easy for anyone with even the least background in research or statistics to think up a zillion reasons to be skeptical. But research almost always sound simplistic and inept when it's subject to simplistic and inept reporting. This brief report reads like the confused ramblings of someone who's overheard a conversation in an elevator.
Fortunately, a kind Slashdot reader has posted a link to the article, and that clears up a lot of the mis-reporting. I took a brief look to check two things: what they did to avoid confounding due to other potential risk factors that might be correlated with cell phone use, and what the evidence is for increased risk on the side of phone use. It seems like they didn't do much to avoid confounding. If heavy cell phone users are also heavy drug users, technophiles, work in high-stress occupations, or whatever, it seems like it could confound the overall rates. It would be a little harder to cook up a story for the ipsi/contra difference. The evidence that the tumors are more likely on the ipsilateral side seems to be numerical, not statistical (they do the stats on one side and again on the other side, but they don't compare the two directly, unless I somehow missed it). It would be nice to know how reliable that difference is, since that seems to be the strongest link between the tumor rates and cell phone use. But that's just my first impression.
I saw a tongue-in-cheek poster at the Society for Neuroscience a few years ago, in which the authors compared portrayals of different professions in a large number of movies. Overall, the most negatively portrayed profession was murderer, and scientists were right in there at #2. The methods employed for this survey involved beer and pizza.
The average person in this country couldn't even begin to tell you what science is, what it's useful for, or what scientists do. To be fair, it's not a question with a simple answer like 42. But it's not surprising that people who make policy decisions at all levels of government know nothing whatsoever about science. It's mis-portrayed almost completely in the media, and probably mis-taught at all levels of education. Scientists are not valued by society in any meaningful way.
Any scientist whose work is in the popular press probably has a story about how their work was portrayed in a way to mislead, not inform people. Perhaps someone will repost the link to that recent insightful article about how few science reporters have any science background.
The government has been rewriting science more blatantly in environmental sciences than in other areas. But it's the other kind of rewriting that's more insidious and harmful. Necessarily, most science funding comes from the government. They decide what to fund and what not to fund. Serious scientists get input into this decision, but not the last word. What's insidious about it is that no individual scientist is doing what they do because the government told them. But since there's such an oversupply of scientists, including a healthy supply interested for their own reasons in doing the specific things the government would like, the government can shape science to whatever extent they want without there ever being a single scientist who was specifically influenced.
I don't find that so objectionable, but the article portrays it as though they're repudiating my favorite graphical email client, which really seems to overstate their real claims.
Memory deficits are a risk of epilepsy surgery. As yet there's no truly reliable way to predict post-surgical memory problems, but since surgery is generally a last resort, it's a risk the patients have to take. This kind of memory problem is also typical of Wernicke-Korsakoff's Syndrome, dramatized in the second chapter of Oliver Sacks's "The Man who Mistook his Wife for a Hat," which is often a consequence of long drinking binges (and an accompanying vitamin deficiency, I think). You don't always see the truly dense amnesia, but when you do it's striking.