What a terrible inflammatory title for an article. There are many measures of employment/unemployment because each one of the available ones is incomplete. The one cited is useful and historically fairly consistent so it is possible to make some meaningful comparisons from year to year. At least it is more useful than many other cited numbers like the Dow.
As an example, here are 6 measures for the USA: https://www.bls.gov/news.release/empsit.t15.htm
Given the large number of people posting along the "only an idiot would make a mistake on this sort of ballot" theme, need to understand better how interfaces and expectations interact with human performance.
Voters don't enter the booth expecting the ballot to be difficult to understand or with a warning of "there might be a trick hole". If you think it going to be straightforward, you don't exhaustively analyze what's before you. Many of these voters had voted numerous times before, and based on their experience, they would expect the second hole on the right to correspond to their choice. It's quite possible the right half of the book was not even used in their previous experience.
As an experiment, try looking at the sample ballot on CNN and cover up the right half in order to model someone who doesn't expect the right half of the ballot book to contain information or who's not interested in third party candidates. The Gore/Lieberman section encompasses two holes. It's easy to see how people might punch either the second hole or both holes.
GUI programmers should be aware of the "intuitive interface" that gives the wrong intuition problem. If you use one of these interfaces, you'll find you are bound to make mistakes, no matter how clever you are.
Additional information gathered from an NPR call-in yesterday. (I don't know if they have been confirmed.)
The actual ballots were different from the sample ballots given out ahead of time, adding to the probability of mistakes.
At least one woman, realizing that she had accidently voted for Buchanan instead of Gore, was not allowed to a new ballot. Instead, the election working took the ballot from her hand.
Note: Instructions on sample ballot said that if a voter made a mistake while punching out a ballot, he/she should tell an election officer and the voter would receive a new ballot.
It's absolutely amazing that a something like this could affect the election. I hope that someone investigates this.
I programmed quite a bit under NeXTStep and thought it was quite good as a Unix, at least coming over from using Suns and DEC workstations.
NeXTStep 3.0 came with a compiler, cc, which was based on gcc with modfications for Objective C support. (The objective C code was submitted back to gcc project.)
It was quite easy to compile any GNU tools that were missing--I had a substantially easier time on the NeXT than on SGIs.
Slashdot Mirror
What a terrible inflammatory title for an article. There are many measures of employment/unemployment because each one of the available ones is incomplete. The one cited is useful and historically fairly consistent so it is possible to make some meaningful comparisons from year to year. At least it is more useful than many other cited numbers like the Dow.
As an example, here are 6 measures for the USA: https://www.bls.gov/news.release/empsit.t15.htm
As an experiment, try looking at the sample ballot on CNN and cover up the right half in order to model someone who doesn't expect the right half of the ballot book to contain information or who's not interested in third party candidates. The Gore/Lieberman section encompasses two holes. It's easy to see how people might punch either the second hole or both holes.
GUI programmers should be aware of the "intuitive interface" that gives the wrong intuition problem. If you use one of these interfaces, you'll find you are bound to make mistakes, no matter how clever you are.
It's absolutely amazing that a something like this could affect the election. I hope that someone investigates this.
I programmed quite a bit under NeXTStep and thought it was quite good as a Unix, at least coming over from using Suns and DEC workstations. NeXTStep 3.0 came with a compiler, cc, which was based on gcc with modfications for Objective C support. (The objective C code was submitted back to gcc project.) It was quite easy to compile any GNU tools that were missing--I had a substantially easier time on the NeXT than on SGIs.