Impressive!

You're right - I was speaking about professionally constructed testing, which is all field tested, etc. Without that, you're certainly right that you can't be confident that you're measuring what you think you are. For instructor constructed tests, they usually try to do something along the lines of what I described, for the same reasons, but without the rigor.

You're correct, "Reading comprehension fail".

The school board member was saying that the 4th and 8th grade tests were fine, but that the 10th grade test was too hard, as evidenced by many students who did fine on the 4th and 8th grade tests then failing the 10th grade test. The school board member then took the hard 10th grade test and failed it, which he presented as evidence that the 10th grade test was too hard.

The newspaper then printed examples from the 4th and 8th grade test, which were, as the board member said, appropriate to 4th and 8th graders.

The mistake you're making is not noticing that the questions were not from the 10th grade test, which is the one that the board member said was too hard. By missing that, you're thinking that the board member failed a test of easy math questions, which would have been embarassing if it had happened.

The fact that almost everyong posting on Slashdot missed the key facts due to the lazy/misleading writing puts the blame squarely on the writer. But I blame the journalist more than the readers - either the writer didn't realize that providing irrelevant "example questions" undermined the entire point of the story, or the writer was intentionally trying to ignore the facts and instead make the administrator look like an idiot.

Read the article more carefully. The test that the school board member said was too hard was the 10th grade test. The questions in the article were from the 4th grade test. Not only that, but the school board member said that they reason that he thought that the 10th grade test was too hard was that the kids were doing well on the 4th grade and 8th grade tests, then were failing the 10th grade test.

It was misleading of the newspaper to present the easy, 4th grade questions as if they were the questions that the school board member thought were too hard. It's disappointing that the writer (apparently) provided absurdly easy "example" questions, to make the school board member look like an idiot. Or perhaps the writer was just too dumb to realize that printing the wrong questions mattered.

It's also a bit disappointing how many slashdot readers didn't catch the newspaper's error.

Compound question such as the "axis" question are actually pretty hard because it requires people to remember quite a few concepts that are rarely useful once out of school. Which axis is X and which is Y? What is mirroring around an axis? What do positive and negative coordinates mean? What does (3,7) mean? Which of 3 and 7 in (3, 7) X and which is Y? And so on. Of course, if you know all of that, then the answer is fairly obvious - visualize an X/Y coordinate space, plot the point, mirror it around an axis, and there you are. Or remember that mirroring around an axis chances the sign of one of the numbers but not both.

Given that the administrator took the 10th grade test, I wish that they'd provided examples from the 10th grade test, instead of the 4th and 8th grade tests. Unless you read carefully you'd think that the administrator was an idiot. But if you read about this issue, you'll find that the administrator was actually saying that he thought that the 10th grade test was too hard, because the kids did well on the earlier grade tests (i.e. those test were fine) but were failing the 10th grade test (i.e. that specific test was too hard). So the point of the story should be that the 10th grade test in particular is too hard, not that the administrator was an idiot. The newspaper printed examples from the lower grade tests, which were properly calibrated for younger kids, is misleading, because the reader is given the impression that the administrator got easy math questions wrong, and was an idiot. So either the writer was too stupid to realize that he was misreprenting the facts, or he was intentionally changing the story to suit an agenda. Neither is impressive.

"That's why making multiple choice tests (and grading them) is so frigging difficult to do very well. To do it completely perfectly you need to be able to predict all possible incorrect interpretations and be sure that none of your "wrong" answers are "right" in a way that you would want to give points for."

Tests are better planned than you think. When you construct a (good) test, all of the answers are put there BECAUSE they tell you something specific about the person taking the test. That's why on four answer questions you'll usually see that one answer is right, one answer is absolutely wrong (i.e. the test taker was guessing wildly) and the other two are the answers that the test taker would arrive at if they didn't understand something.

This can be done for two reasons.

First, it allows test takers who understand the subject well enough to eliminate some of the answers a better chance of getting the right answer, which (indirectly) gives students partial credit for partial knowledge.

Second, test can be scored with different values for different 'wrong' answers. For example, 'right' might be worth 5 points, 'wrong' might be worth 0 points, and the 'close' answers might be worth 2 points, explicitly giving students partial credit for partial knowledge.

And if the testing system is really smart, it can analyze the right and wrong answers and give better guidance to the instructor so that they know to provide specific guidance to the student. For example, if someone repeatly subtracts instead dividing, perhaps they're confused about what the division symbol means, so they can get help with that specifically. Or, as someone else in the discussion pointed out, if they read the division symbol as "+" then perhaps they need glasses. Most scoring systems don't do this, but some do. :-)

"If the 2000 accounts existed and different devices are now connecting to them then there is something fundamentally wrong with the software."

Keep in mind that the 2,000 new mobile devices are being used IN ADDITION TO however those people read email. If they stopped using their desktops and just used iPads, the load on the system would be unchanged. Adding 2,000 new clients to any mail system will consume the system's capacity, and if the system doesn't have that capacity it'll be in trouble. There's no magic capacity fairy.

Solyndra's investors weren't particularly Obama donors - the Waltons (i.e. Walmart) were major investors, and they're hardly Obama fans. Keep in mind also that Solyndra was started and was fast-tracked for funding under a DOE program started under Bush, and Obama's [ep[;e actually slowed things down, did more due diligence, and put more protections in place around the loans that ended up saving us money by pulling the plug on the company. Despite Issa's partisan spinning, this isn't something to blame Obama on - any time the government sets up a fund to promote businesses, some of those businesses will succeed and some will fail, and Solyndra failed because China radically dropped the price of solar cells, wiping out Solyndra's market. The real problem isn't that the US government set up a fund to encourage solar development, it's that the US started years later than China, and with a much lower level of investment, so China is beating us. The answer isn't to give up, it's to compete harder.

The Aptera looks like it does for a reason - its primary goal is efficiency, which is how they got over 200 MPG.

But to do that, they had to not waste energy pushing the car through the air. So they made it aerodynamic, so it looks like an airplane rather than the traditional "box on wheels." And their initial target was a two seater, which is most efficient (because most driving is 1-2 people, and with a two seater you're pushing around less mass).

A year ago (apparently) marketplace realities kicked in. That is, while sedans are less efficient, people prefer buying them because it's useful to be able to carry more people when you need to. So the marketplace for sedans is much larger than two seaters, making it a much smarter business to be in. But since they didn't get their funding, we'll never know how that would have played out.

Though I would love to see what a truly efficient sedan might look like.

"Imagine what they could have done with the $700k they would have saved by choosing a tablet other than an iPad."

The iPad is under $500, so it costs the same or less than any other decent tablet. Are you saying that there's a tablet that costs $150 that's comparable to the iPad? That is pretty hard to imagine. Don't forget to include the management costs - iPads are extremely easy for an enterprise to manage, because they integrate nicely into Exchange (e.g. you can define mail policies on your Exchange server, and iPads do what they're told - encrypt, require password lock, etc.). Android doesn't do this properly yet. That leaves the RIM Playbook, which aside from sucking has the same list price as the iPad. I guess you could save some money buying discontinued products that are being dumped, but that's not a great enterprise hardware strategy. :-)

If you want to complain about the project, complain that they didn't plan for adding one more ActiveSynch server so they had capacity to support their users. Given educational pricing, the software is nearly free, and even an overpriced server would have been a trivial percentage of the project budget.

They don't need 5-10 servers.

Keep in mind that they're not talking about adding 2,000 mailboxes, just adding 2,000 devices to access existing mailboxes. So they don't need more storage, just more server compute capacity. If I had to guess, it might be as simple as them running ActiveSynch on a single, under-resourced server (or VM) as a POC, and they didn't expect (or prepare for) the increased demand of 2,000 more tablets. Should be easy to fix. Though inevitably they're trying to do a dozen other things, and it'll take three months to do the paperwork to get the approval to buy a new server and get it deployed. Remember,

If I had to guess, the issue isn't the specific protocol, it's that the number of mail clients doubled. That is, if they have 1,000 employees, each reading mail from a desktop computer, and each employee gets an iPad that they use in addition, they went from 1,000 mail clients to 2,000 mail clients, which would require them to scale the mail server to support it. If I had to guess, the iPads turned out to be much more popular than expected, greating demand that they were unprepared for.

The point of unions isn't that they render everyone angels, it's that it creates an organization that can negotiate in favor of worker's interests to balance the organization that already exists to support management's interests. So an IT workers' union could impose checkpoints in a process such that the workers could make sure that adaquate resources, training, tools, etc., were provided to allow the workers to be successful without working insane hours compensating for poor planning or resourcing. Yes, a good management team ought to be thinking of such things, but the software industry's track record is poor enough (only 10% of IT projects deliver what's required on time and budget) that giving the IT workers more leverage doesn't seem like a bad idea.

"Unions work best for the health and safety of their workers. Anything beyond that is mob rule."

Add in "and are properly equipped and trained and resourced to do their job successfully". For example, air traffic control unions negotiated limits on how many hours controllers could be forced to work, and when they unions were broken and controllers were forced to work so many hours, with no breaks for even going to the bathroom or eating meals, endangering passenger's lives. And when teachers' unions negotiate limits on the numbers of students in classes, so teachers can actually teach students effectively.

Or do you think that the MBA who runs a company knows how best to do people's jobs, not the people who actually do the jobs?

Given that the tablets appear to be so popular that they're swamping the mail server, they're demonstrably not "tablets no one uses",

Nice .sig though.