> I wasn't planting a suggestion -- I was stating a fact as a point to the accusation
There you are doing it again. This time you even mentioned a source. Clearly it can affect how one may answer some of the final questions. (The only ones relevant to the purpose of the survey). Your survey is doomed.
Try again later
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
I'm going to explain something fundamental about surveying and slanting statistics. You seem intelligent enough that you'd be able to understand it, so I'll give you an idea of why I feel slightly offended about that poll as it stands.
When a poll asks a question that is biased and limits the answers to said question... it does two things. One, it gives any less informed poll taker a false idea of reality. For instance, one would infer by question #9 that the US has shown blatant disregard for the Geneva Convention at Guantanamo Bay which is wrong. This leads those less informed to give a possibly incorrect answer relative to how they would feel about the actual situation. Two, since those more informed are limited to only incorrect answers (if the question itself is wrong, then all answers would subsequently be wrong in terms of the poll takers opinion) then they'll likely simply ignore doing the poll all together (as you suggest to me) there by eliminating an entire segment of the population. This would obviously result in unbalanced data.
So, here is what I would do if I were you. If you care about the grade, a lot... then you might as well just completely BS the data yourself. This is a programming forum... I'm sure you can write a simple program that randomly does the poll one hundred times. It would likely yield data just as good (assuming you seed rand(), of course) as polling real people would (which would be inaccurate, anyway). If you care more about giving an honest and truthful poll, then go to whoever wrote these questions, tell them everything I just wrote, and then proceed to torture that person. ... but only if s/he likes it.
Last edited by SlyMaelstrom; 12-01-2006 at 06:30 PM.
Sent from my iPad®
There a few others things fundamentally wrong wit the survey.
For one, it is aggressive towards the person answering it. 3 or 4 questions concerning if we agree or not with terrorism put too much weight on what is not the real objective of the poll. For a pacifist, for instance, it will make him fell guilty about the remaining questions.
Some questions seem to scream one of the answers. Almost forcing a person to drawn the X next to it. When terms like "an eye for an eye" are used in the question itself, it's hard for anyone not to say it is wrong. But where the pacifist would have trouble before, now the war supporter will be divided, because this question seems to want to destroy his convictions on the other ones.
Your survey is aggressive and plants conflict on the minds of those answering it. They will feel unsatisfied because no matter what they answer, they will think their answers didn't reflect their opinions.
It's a bad survey, mate.
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
Observations
1. Me thinks Mario F. needs to stick to programming and leave politics to others.
2. This poll is so slanted and biased, calling it a 'poll' is ludicrous.
3. This thread topic is so lame and so over-discussed I cannot believe anyone answered it.
4. The poll creator needs to google for Loaded Questions
Stupid thread and I doubt anyone cares how we feel about the topic in here. What a waste of GD space.
Flame me later. Thx and have a nice day.
Last edited by VirtualAce; 12-01-2006 at 07:26 PM.
Later. Take care.
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
Thank you.Originally Posted by Bubba
2. Yep, it actually is biased. Not actually in the way I am biased, but they are bad questions.
3. Quite honestly, me too. You wouldn't believe how surprised I was there were 2 pages of responses when I checked it.
4. I didn't create most of the poll's questions and in fact the poll doesn't even represent what I truly wanted to poll about. We had a required number of questions per group and the groups were allowed to choose between a few topics which weren't very interesting. Most of the questions are loaded... as I said, if you don't really care, just answer the questions. If you care, then just don't answer the survey.
I realize I could quite easily write a program to generate fake poll responses, but that would be cheating and I try to avoid that
at least it will truly represent the opinion of your computer on the matterwrite a program
All problems in computer science can be solved by another level of indirection,
except for the problem of too many layers of indirection.
– David J. Wheeler
Like in.... S&M?
Yeah, these questions suck. They're like "Do you think killing should be legal?" or something.
"The Internet treats censorship as damage and routes around it." - John Gilmore
