Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Math Education Microsoft

Microsoft's Math-Challenged STEM Education Contest 96

theodp writes "As noted earlier, Microsoft is tackling the CS education crisis with a popularity contest that will award $100K in donations to five technology education nonprofits that help make kids technically literate. Hopefully, the nonprofits will teach kids that the contest's voting Leader Board is a particularly good example of what-not-to-do technically. In addition to cherry-picking the less-pathetic vote totals to make its Leader Board, Microsoft also uses some dubious rounding code that transforms the original voting data into misleading percentages. Indeed, developer tools reveal that the top five leaders in the Microsoft STEM education contest miraculously account for 130% of the vote. Let's hope the quality control is better for those Microsoft Surface voting machines!"
This discussion has been archived. No new comments can be posted.

Microsoft's Math-Challenged STEM Education Contest

Comments Filter:
  • by harlows_monkeys ( 106428 ) on Sunday July 28, 2013 @02:11PM (#44407347) Homepage

    A leader board shows the TOP competitors. That's the point of a leader board. It is not "cherry picking" to only show the top.

    The rounding is not dubious. They are rounding to 10% increments because that is the resolution of the progress bars.

    The "percent-10", "percent-50", and so on that the "developer tool" is showing are the classes of the progress bars. There is a style correspond to each in main.css, and that determines the length of the progress bar. The style sheet provides "percent-0", "percent-10", ..., "percent-100".

  • by plopez ( 54068 ) on Sunday July 28, 2013 @02:29PM (#44407457) Journal

    I would rather see them pay the billions they owe Washington state end the US government. The money would be better spent. Rather than a few non-profits getting a pittance, the money which could make a real difference would be available. NGOs are horrifically inefficient. See Haiti as an example.

  • by msauve ( 701917 ) on Sunday July 28, 2013 @02:43PM (#44407537)
    You're absolutely correct. Despite the variable names in the code, the leaderboard itself isn't presented as anything other that a simple "Who's in the lead?" graphic. It accurately reflects the top 5 vote getters (which makes sense for a contest which will have 5 winners) in proper order, and gives a rough indication of their relative popularity. The contest promises "We'll donate $50,000 to the organization with the most votes, $20,000 to the second highest, and $10,000 to each of the three runners up." The graphic clearly and accurately reflects the current status, and allows the view to match the prize with the contestant. It's not like they get a monetary amount based on their percent of the vote.

    It's not "math challenged," the poster is just a "hey look at me, I'm smarter than Microsoft coders" troll.
  • by Anonymous Coward on Sunday July 28, 2013 @02:52PM (#44407579)

    It looks like the only real complaint here is the "misrepresentative" bars. Those can be blamed on lazy development. There are a couple clues to this:

    1: They didn't bother to minimize their CSS
    They're apparently using the dev version of their CSS to serve on the site: http://www.azuredevs.com/css/main.css
    One could maybe argue they're offering some fun for "developers" who show up. But my money is on a misconfiguration or some generally shoddy web development.

    2: The debug notations provide a clue to the source code
    Aside from the huge waste of bytes sent to users, using debug code in production provides hints to the actual source. In this case, the percent-bar classes read like this: /* line 1479, C:\Projects\Microsoft\MWA\Repo\mwa\AzureDev\css\main.less */
    figure.progress-bar.percent-70 span {
        width: 70%;
    } /* line 1484, C:\Projects\Microsoft\MWA\Repo\mwa\AzureDev\css\main.less */
    figure.progress-bar.percent-80 span {
        width: 80%;
    }

    As you can see, each class is 5 lines apart, which means that they probably look more or less like we see here. If they were using a loop structure (non-trivial in LESS, but do-able), they could have easily generated 25 different versions of the bar. It would also have meant that each iteration would have been generated from the same line of code. So they're not doing that. And since someone had to hand-code each of the 10 version of the bar, they were probably a bit lazy about it.

    3: Another clue from the CSS: The developer probably doesn't work for Microsoft.

    If you look at the file paths from the debug CSS, you'll notice the source is stored under "C:\Projects\Microsoft\..." If you work for Microsoft and are developing Microsoft software on a Microsoft device, isn't it a bit redundant to have a "Microsoft" directory in your Projects folder? The only way that makes a lot of sense is if the person who wrote the code didn't actually work for Microsoft.

    So, my conclusion: Microsoft farmed out their website to someone who was either under-skilled, over-worked, under-motivated, or some combination of those. The result isn't of top quality. Go figure. Next story please. :-D

    So yeah, I think what we're seeing here is just bad/lazy web development. Another clue from the

  • by theodp ( 442580 ) on Sunday July 28, 2013 @03:17PM (#44407705)

    C'mon, this could be the poster child for Wikipedia's Misleading graph [wikipedia.org] article. :-)

    Other commenters pointed to the .css file, which contains a "min-width: 10%;" statement that adds to the distortion. Below the Leader Board, you'll note that even those nonprofits with essentially 0% of the vote have progress bars that suggest they have 10% of the vote. Guess it looks better than showing that pretty much nobody cared to vote for them (e.g., the Microsoft-backed STEM Education Coalition [stemedcoalition.org] has 13 votes), and they're way out of contention.

  • by nbauman ( 624611 ) on Sunday July 28, 2013 @05:17PM (#44408293) Homepage Journal

    Why doesn't Microsoft pay its taxes, so that we can run the schools, libraries and support services for kids to grow up to be programmers or anything else they want?

    We're a wealthy country. We should be able to raise money among ourselves and decide among ourselves how we want to spend it. (It's called taxes.) I think most of us would want to spend the money on free public schools, including free college (like the countries we compete with, including the countries those HB-1 immigrants come from). I don't think many people here want their children to graduate college $50,000 in debt, or to drop out of college because they can't afford it. (The Gates Foundation, BTW, was a member of ALEC, which did so much to cut our taxes and destroy low-cost public university education http://en.wikipedia.org/wiki/List_of_members_of_the_American_Legislative_Exchange_Council#Former_corporate_members [wikipedia.org])

    We don't need billionaires making these decisions for us, instead of paying taxes so we can decide ourselves.

    http://www.reuters.com/article/2011/07/27/us-microsoft-tax-idUSTRE76Q6OB20110727 [reuters.com]

    Insight: Microsoft use of low-tax havens drives down tax bill

    By Lynnley Browning

    FAIRFIELD, Connecticut | Wed Jul 27, 2011 7:07pm EDT

    (Reuters) - If you want to know why tax from surging corporate profits isn't making much of a dent in the United States' crippling budget deficit, a glance at Microsoft Corp's recent results provides some clues.

    Things were rosy in the giant software company's just-ended fiscal fourth quarter, which produced record sales of nearly $17.4 billion, a 30 percent increase in after-tax profit, and a 35 percent gain in earnings per share.

    But for the Internal Revenue Service and foreign tax authorities, things weren't so rosy. Microsoft reported only $445 million in taxes in the U.S. and other foreign countries, just 7 percent of its $6.32 billion in pre-tax profit....

  • by nbauman ( 624611 ) on Sunday July 28, 2013 @06:21PM (#44408549) Homepage Journal

    Correct. That's what Paul Farmer http://en.wikipedia.org/wiki/Paul_farmer [wikipedia.org] said. Farmer did more than anybody else in the world to improve Haiti's health care system.

    Farmer said that, to run a health care system, you need an overall plan, and that requires a central government. You can't have volunteer charities from the US and elsewhere parachute in for 6 months to do their thing. You need the government to decide what the priorities are. Maybe some church group wants to come in for 6 months and hand out eyeglasses, which is all well and good. But their urgent problems are infant mortality, maternal mortality, diarrhea (which is the main cause of infant mortality), and sheer starvation. Somebody has to come up with a strategy to assign priorities, and the free market isn't good at that. (The free market has already assigned its priority as taking care of the needs of rich people.)

    Unfortunately, Farmer (who spends half the year in Haiti and half in Harvard) said that the Clinton Administration was trying to drive Aristide, the (elected) president, out of office, so the U.S. prevented funds from going directly to the Aristide government, but sent them to the NGOs, some of whom were run by Aristide's rivals. You wind up with warehouses full of (say) enough mosquito nets for 10 years, when hospitals don't have essential drugs like morphine to give people who have their limbs amputated, or drugs for people with cancer.

    Haiti is a classic case of a government that, for all its faults, could have run its health care system better than NGOs.

  • by black3d ( 1648913 ) on Sunday July 28, 2013 @06:53PM (#44408707)

    I hope you realise, that simply believing "crazy racist stuff you heard somewhere" (likely, from racists) doesn't make it true. There are (far) more whites than blacks on death row in Florida. http://www.dc.state.fl.us/activeinmates/deathrowroster.asp [state.fl.us]

    Most people executed in Florida have been white. http://www.dc.state.fl.us/oth/deathrow/execlist.html [state.fl.us]

    There have been plenty of people executed in Florida for killing blacks and other non-whites:
    Richard Henyard [wikipedia.org]
    Mark Schwab [wikipedia.org]
    David Alan Gore
    Manuel Pardo
    And many more.. http://en.wikipedia.org/wiki/Capital_punishment_in_Florida [wikipedia.org]

    The ironic thing here is, simply blinding believing that other people are being racist due to a set of circumstances which aren't actually occuring, is highly suggestive of you, in fact, being racist. You're already set in a prejudiced view and don't care about the facts. BTW, George Zimmerman is innocent. Deal with it.

If you want to put yourself on the map, publish your own map.

Working...