One. That all depends on how the question was phrased!
Two. I assume, Sven, that you listened to the BBC this evening. Both BBC1 and R4 led with the story that "the Americans are asking for another 2000 soldiers and we will probably meet that request"! Twenty minutes later, you start a thread on the second or third news item!
More than likely it is a selfish motive rather than anything else. With the economy the way it is, they don't want UK taxpayer's money spent on another country when it could be used to pay their benefits.
Another option is that they think that Afghanistan is an 'illegal war', failing to understand that it's a very different situation to Iraq, not least given that UNSCRs 1368, 1373 and 1378 are amongst the most bellicose the UNSC has ever issued... It's quite remarkable how many people have managed to conflate Iraq and Afghanistan as both being opposed by the UNSC, and that may well influence a proportion of those polled.
The chances of the British public having a realistic grasp of the situation based on most of the coverage they'll encounter is, I fear, rather slim.
I suspect that the question was asked to get the answer, the majority would want free money, free beer and free sex with Kylie Minogue. Just because the majority want it doesn't mean they should get it or that they're right.
Most poll companies are fairly scrupulous about ensuring that their poll selections are a good demographic spread; their proffesional reputation and therefore livelyhood depends on it. They are also usually fairly good about not loading the questions. The size of pool is determined by the organisation that commissions, and pays for the poll.
Just because this, mostly military audience, doesn't agree with the findings of the poll doesn't mean it isn't an accurate reflection of public opinion. IMHO, the Government has done a lamentably poor job of persuading the people of the necessity for this war. If they had a choice, I am sure the politicians would not talk about the war at all - there are no votes in it.
1000 people doesn't really give a huge spread. It totally depends on where the poll was carried out, the age and sex of the people etc.
The main reason why polls are carried out is to provide stats to back up somebodies arguement. If ICM wanted to prove that the war isn't popular they would go to an area where they know they are going to get the responses that they are looking for.
The full report containing the repondents' background information (gender, age, financial status, etc.) can be downloaded in pdf from the bbc site. It looks fairly balanced.
Problem is: the poll itself doesn't answer why the UK should or shoudn't withdraw it's troops from Afghanistan. It just shows a graph with the majority of the respondents answering 'should' to a certain question.
You might be surprised that 1000 is not a lamentably small group at all. Statistics show that 95% of the time the result you get from a random sample size of 1000 will be within 3% of the 'correct' answer from sampling the whole 60m population. Stratified sampling that the polling organisations tend to use does even better. Google 'sample size'....