• Claxon (unregistered)

    You can hardly call the words "Please" and "Hold" a language!

  • s. (unregistered)

    I wonder how screwed up their percentage calculation algorithm is to miss the target sum of 100% by 2.

    The task of rounding the percents on options so that they make sense is not trivial, I had this task not long ago and the simple concept of 2-line "divide by sum, multiply by 100, round" had to be replaced with good 50 lines of code to make the percent add up to 100%, never give more percent to an option with less votes and so on. But I wonder what kind of screw-up did happen here, seriously.

  • (cs)

    Yes, but you have to remember that those 75% are in the minority.

  • s. (unregistered) in reply to Claxon
    Claxon:
    You can hardly call the words "Please" and "Hold" a language!

    My first bet was "Wow, they provide access using DTMF signals or a modem connection for communication with the system!"

    This would actually be an amazing feature allowing to use the dial-in service with embedded devices and external automated services, say "transfer your heart monitor results right to the physician", "query the database and get a few row of records automatically" etc.

  • (cs)

    What happens if you only speak English? You select "No Preference" and hope your doctor isn't limited to speaking Russian?

  • (cs) in reply to s.
    s.:
    I wonder how screwed up their percentage calculation algorithm is to miss the target sum of 100% by 2.

    The task of rounding the percents on options so that they make sense is not trivial, I had this task not long ago and the simple concept of 2-line "divide by sum, multiply by 100, round" had to be replaced with good 50 lines of code to make the percent add up to 100%, never give more percent to an option with less votes and so on. But I wonder what kind of screw-up did happen here, seriously.

    37% of all statistics are made up on the spot!

  • Bluemoon (unregistered) in reply to akatherder
    akatherder:
    What happens if you only speak English? You select "No Preference" and hope your doctor isn't limited to speaking Russian?

    If you read the small print behind the drop-down it says "in addition to english"

    I am sure glad that you can also select 'sign language' as a language. I hope that docter also has a phone for the deaf and speech-impaired to communicate with the doc.

    Bluemoon

  • Beeblebrox (unregistered) in reply to Claxon
    Claxon:
    You can hardly call the words "Please" and "Hold" a language!

    Well, those are two different sounds, so that's enough for you to transmit binary data, eh?

    Horribly inefficient, yes, but it'd work. You'd need to make sure that they support encodings other than UTF-8, though.

  • (cs)

    Maybe the first one is simply implying that the given person can make it, but only after a few tries.

  • MadeItUp (unregistered) in reply to ParkinT

    I thought it was 42%

  • SlyEcho (unregistered)

    C++? AT&T assembler? Those actually are languages.

  • (cs)

    I actually thought the last one was funny because of the death penalty/moratorium combination.

    I probably confused it with 'mortuarium', the Latin (and also Dutch) word for mortuary.

    Never mind.

  • JM (unregistered)

    What do you think of herding emus?

    1. It's stupid.

    2. Herding emus may be an appropriate action in some circumstances, but definitely not in others. Which circumstances are appropriate should be determined by a broad consensus among people who are knowledgeable on this subject, and due democractic support from the lay public.

    3. It's OK.

  • Doug (unregistered) in reply to Claxon

    I think C++ might be the AT&T language they speak of.

  • s. (unregistered) in reply to SlyEcho
    SlyEcho:
    C++? AT&T assembler? Those actually are languages.

    This gives me a clue as what this WTF might originate from.

    "Grab me a list of all languages." "Yeah, here you are, ls /usr/share/locale." "What is this? Hey, you, match a human name to each of these." "Humm, locale 'C'. What locale is 'C'? Ah!"

    and so the "default locale language" made it into the final product.

  • (cs) in reply to s.
    s.:
    I wonder how screwed up their percentage calculation algorithm is to miss the target sum of 100% by 2.

    The task of rounding the percents on options so that they make sense is not trivial, I had this task not long ago and the simple concept of 2-line "divide by sum, multiply by 100, round" had to be replaced with good 50 lines of code to make the percent add up to 100%, never give more percent to an option with less votes and so on. But I wonder what kind of screw-up did happen here, seriously.

    5.49% => 5 % 18.49% => 18 % 75.49% => 75 %

    which adds up to 99.47%.

    0.53% think it's a stupid question. So it doesn't count.

  • JD (unregistered)

    Although a pretty messed up description, "AT&T Language" is probably one of the best options for many non-English persons. "AT&T Language" is a service by which you can have access to an interpreter over the phone - this includes interpreters of 140 different langauges.

  • (cs) in reply to akatherder
    akatherder:
    What happens if you only speak English? You select "No Preference" and hope your doctor isn't limited to speaking Russian?
    No, "No Preference" means you don't care; you speak all of them and your health care professional can speak to you in whatever is his native language rather than having to use English.
  • s. (unregistered) in reply to jimlangrunner
    jimlangrunner:
    5.49% => 5 % 18.49% => 18 % 75.49% => 75 %

    which adds up to 99.47%.

    0.53% think it's a stupid question. So it doesn't count.

    It would add up to 99.47, not to 100.0 this way, meaning the usual, non-WTFy 1% error, not the exceptional 2% they managed to make on 3 options.

    Likely they didn't use simple round(). They probably truncated the number:

    5.75% => 5 18.75% => 18 75.50% => 75

    or such.

  • billswift (unregistered) in reply to jimlangrunner

    5.49 => 5 18.49 => 18 75.49 => 75

    99.47 => 98

  • (cs) in reply to Claxon
    Claxon:
    You can hardly call the words "Please" and "Hold" a language!

    You've obviously never had to deal with them on a technical level. I work at a telecom call center, and it took quite a while to figure out where the PBX meets the SMAS with the extended d-marc inbetween. And let's not forget the Hoot'n'Hollar and ring-down circuits. Just be careful not to pull your bridge coils, or else your levels are going to be long!

  • Former AT&T employee (unregistered)

    For those of you who aren't aware, there really is an AT&T language. It sounds a lot like a cross between an Indian dialect with the cries of millions of unhappy customers.

  • charon (unregistered) in reply to billswift
    Though Dion was a bit rusty on the physics for this question, he's pretty sure he'll be able to do fine using process of elimination...

    ...until he realized that the 5 tries are shared for all the 30 questions

  • Robert Synnott (unregistered)

    I am reminded of an old BBC sketch show; one sketch had a conversation between a lecturer in business at Low Brow University and a talk show host. It went as follows.

    ... Lecturer: The vast, vast, HUGE majority of people, who voted for Mrs. Thatcher... Host: The 42%? Lecturer: Yes, them.

  • Keith Twombley (unregistered)

    I love these types of posts. The only way it'd be better is if the URL for the pictures didn't contain a quote which seems to break a lot of IM clients' URL finding algorithms.

  • (cs)

    For the first one, the "attempts remaining" could be for the entire test. He'll surely get this one question right but he'll have fewer attempts left for the rest of them.

  • (cs) in reply to Robert Synnott
    Robert Synnott:
    I am reminded of an old BBC sketch show; one sketch had a conversation between a lecturer in business at Low Brow University and a talk show host. It went as follows.

    ... Lecturer: The vast, vast, HUGE majority of people, who voted for Mrs. Thatcher... Host: The 42%? Lecturer: Yes, them.

    I love examples like this, because it shows the fallacy of statistics.

    42% is not a majority of the whole. 42% could be a majority of the votes as the other 58% could be split evenly among 5 other candidates giving them each 11.6%.

    Without context, statistics mean nothing. So in the conversation who is right? Funny enough, they both are.

  • Grant (unregistered)

    Did you notice that English was not even a choice for the doctor? They must have the same doctor as me!

  • AdT (unregistered) in reply to JM
    JM:
    What do you think of herding emus?

    (snip)

    Excellent observation, JM. I think because of answer #2, this poll qualifies as a complex question. It's discomforting that the visualization incorrectly and overwhelmingly supports this answer, one could almost think this was by intent. Though if so they might have faked the numerical results as well.

    Regarding the discussion about the 98% total: It's not unusual for the partial results in a poll to total less than 100%, at least outside the Internet, as "unexpected answers" are often included in the reference total (like "I don't know" or "F*** off" :-) ). So in this respect, I take PBS to be innocent until proven guilty of dyscalculia. ;-)

  • Loren Pechtel (unregistered) in reply to s.
    s.:
    I wonder how screwed up their percentage calculation algorithm is to miss the target sum of 100% by 2.

    The task of rounding the percents on options so that they make sense is not trivial, I had this task not long ago and the simple concept of 2-line "divide by sum, multiply by 100, round" had to be replaced with good 50 lines of code to make the percent add up to 100%, never give more percent to an option with less votes and so on. But I wonder what kind of screw-up did happen here, seriously.

    A simple truncation of the decimals will do this.

    5.66%, 18.66%, 75.67%.

    This could be caused by display formatting while the internal calculations are correct.

  • (cs) in reply to s.
    s.:
    I wonder how screwed up their percentage calculation algorithm is to miss the target sum of 100% by 2.

    The task of rounding the percents on options so that they make sense is not trivial, I had this task not long ago and the simple concept of 2-line "divide by sum, multiply by 100, round" had to be replaced with good 50 lines of code to make the percent add up to 100%, never give more percent to an option with less votes and so on. But I wonder what kind of screw-up did happen here, seriously.

    Don't they normally just write "Percentages may not add up to 100 due to rounding?" I've never heard of trying to massage the numbers to always equal 100.

  • (cs) in reply to KattMan
    KattMan:
    42% is not a majority of the whole. 42% could be a majority of the votes as the other 58% could be split evenly among 5 other candidates giving them each 11.6%.

    Without context, statistics mean nothing. So in the conversation who is right? Funny enough, they both are.

    No, 42% of anything is never a majority of that thing. If 42% of the votes were for Thatcher, she did not receive a majority of the votes. The word you're looking for when you say "a majority of the votes" is plurality.

  • Pi (unregistered)

    Oh god. The thing with the True/False question is that if you answer it wrong, you get 0 points. I had to work with that (&*#!^ thing in my physics class.

  • IV (unregistered)

    I have no idea where some of you learned math/politics. 42% cannot ever be a majority unless <84% vote. If the other votes are split up, and 42 is the highest, that is a plurality.

    Captcha: haero (thus proving I may not be a robot)

  • MeRp (unregistered) in reply to Someone You Know
    Someone You Know:
    KattMan:
    42% is not a majority of the whole. 42% could be a majority of the votes as the other 58% could be split evenly among 5 other candidates giving them each 11.6%.

    Without context, statistics mean nothing. So in the conversation who is right? Funny enough, they both are.

    No, 42% of anything is never a majority of that thing. If 42% of the votes were for Thatcher, she did not receive a majority of the votes. The word you're looking for when you say "a majority of the votes" is plurality.

    Well, in a way, 42% could be a majority, not just a plurality. If, in an election, only 42% of the general populace voted, but they all voted for a single candidate, then that candidate would have received the largest majority possible (100%), but would only have been endorsed by 42% of the populace.

  • Salty (unregistered) in reply to Former AT&T employee
    Former AT&T employee:
    For those of you who aren't aware, there really is an AT&T language.

    Yes, it uses iPhonetics.

  • Salty (unregistered) in reply to s.
    s.:
    I wonder how screwed up their percentage calculation algorithm is to miss the target sum of 100% by 2.
    Choice 1: 5% Choice 2: 18% Choice 3: 75% No opinion: 2%

    Boss: "Nobody cares about the undecided/no opinions. Drop them from the table."

  • Dion (unregistered) in reply to apetrov87

    Hey, nope, each question has its own number of attempts, always five. There's only about 1-2 multiple choice questions for each assignment. This seems like a case of bad CONSTANTS!!

    private const int MAXIMUM_ATTEMPTS = 5 neone?

  • Dion (unregistered) in reply to Pi

    Really? I got it right after changing my answer.

  • (cs) in reply to MeRp
    MeRp:
    Well, in a way, 42% could be a majority, not just a plurality. If, in an election, only 42% of the general populace voted, but they all voted for a single candidate, then that candidate would have received the largest majority possible (100%), but would only have been endorsed by 42% of the populace.

    Correct, although my point was that what KattMan was describing was a plurality, not a majority—and that 42% of the votes is obviously never a majority of the votes.

  • Laie Techie (unregistered)
    Lecturer: The vast, vast, HUGE majority of people, who voted for Mrs. Thatcher... Host: The 42%? Lecturer: Yes, them.

    I think it's obvious: The vast majority of the 42% who voted for Mrs. Thatcher had something in common which the Lecturer could not say because the host interrupted.

    It's funny that my captcha would be populus on a question regarding percentages and majorities.

  • (cs) in reply to Someone You Know
    Someone You Know:
    MeRp:
    Well, in a way, 42% could be a majority, not just a plurality. If, in an election, only 42% of the general populace voted, but they all voted for a single candidate, then that candidate would have received the largest majority possible (100%), but would only have been endorsed by 42% of the populace.

    Correct, although my point was that what KattMan was describing was a plurality, not a majority—and that 42% of the votes is obviously never a majority of the votes.

    I stand corrected, but to the layman, plurality means nothing because they never studied statistics and don't understand the difference here. To the layman, saying they got the majority of the votes when it is 42% and everyone else got barely over 10% just seems like the right thing to say, no matter how wrong they are in reality, they are correct in what really matters, meaning Thatcher got more votes than anyone else.

  • Firethorn (unregistered) in reply to Salty
    Salty:
    Boss: "Nobody cares about the undecided/no opinions. Drop them from the table."

    That would work, of course, if they happened to ask me after a particularly bad day: "Our current death penalty statutes down't go far enough! They need to execute more people!"

    I wonder what area this was, that 75% of people opposed the death penalty, because in regions which have the DP, the majority of people support at least some form of DP.

  • Robert Synnott (unregistered) in reply to KattMan
    KattMan:
    Robert Synnott:
    I am reminded of an old BBC sketch show; one sketch had a conversation between a lecturer in business at Low Brow University and a talk show host. It went as follows.

    ... Lecturer: The vast, vast, HUGE majority of people, who voted for Mrs. Thatcher... Host: The 42%? Lecturer: Yes, them.

    I love examples like this, because it shows the fallacy of statistics.

    42% is not a majority of the whole. 42% could be a majority of the votes as the other 58% could be split evenly among 5 other candidates giving them each 11.6%.

    Without context, statistics mean nothing. So in the conversation who is right? Funny enough, they both are.

    Nope, it explicitly says the majority of PEOPLE (meaning people who voted, in this case). You're right, though; in that election, the Conservatives won with ~42%, with the rest mostly split between Labour and the Liberals.

    Laie Techie:

    I think it's obvious: The vast majority of the 42% who voted for Mrs. Thatcher had something in common which the Lecturer could not say because the host interrupted.

    If the comma after 'people' wasn't there, then that would be the case. You can't just randomly go around ignoring commas, though.

  • (cs) in reply to Laie Techie
    Laie Techie:
    Lecturer: The vast, vast, HUGE majority of people, who voted for Mrs. Thatcher... Host: The 42%? Lecturer: Yes, them.

    I think it's obvious: The vast majority of the 42% who voted for Mrs. Thatcher had something in common which the Lecturer could not say because the host interrupted.

    I would agree if the conversation hadn't been described as being part of a comedy sketch, meaning that the error is almost certainly intentional.

  • Adrian (unregistered)

    I used the same physics software as the first example; first, and I have to say this whenever I talk about that software, IT SUCKED. Hard. That out of the way, the number of attempts is for each question. It can be set by the teacher; some classes I used it for had 5 attempts, one class had 20. However, in all the classes I took, if you only had x options then your score was based on how many tries it took you out of x to get it right. Thus, process of elimination never worked for me, in any of the classes.

  • Zock (unregistered) in reply to Someone You Know
    Someone You Know:
    If 42% of the votes were for Thatcher, she did not receive a majority of the votes. The word you're looking for when you say "a majority of the votes" is plurality.

    Nope. We're talking about Thatcher here, thus the term is "simple majority." ;)

  • zoips (unregistered)

    Hey, I've used that online physics homework thingy. It sucked. Either way he's still going to get 0% for the question even though it's letting him pick the correct answer.

  • Linguists Aren't Cunning (unregistered) in reply to Claxon

    Actually, it's a much richer language than that. They have many words for "obscure calling plans where you hope to pay us a tiny bit less by following arcane rules but actually pay us a lot more" but no word for "refund" or "service credit."

    Still, it's quite easy to translate the calling plans into English once you know that they all mean "pay us more money." Too many people get hung up on a word-for-word translation and miss the entire point of the writing.

  • Mikey2 (unregistered) in reply to Claxon

    Maybe they're talking about Joybubbles http://ascii.textfiles.com/archives/000461.html being able to do this http://en.wikipedia.org/wiki/2600_hertz .

Leave a comment on “Process of Elimination”

Log In or post as a guest

Replying to comment #:

« Return to Article