The annual social attitudes survey has thrown up interesting questions about the UK's Trident nuclear weapons system.

How strong is public support for the SNP's policy of removing the weapons from the Clyde?

The party has long argued that a majority of Scots want to be rid of them, and a Campaign for Nuclear Disarmament survey, widely quoted, found that 60% in Scotland oppose replacing Trident.

However, the figures from NatCen Social Research suggest that, if Scotland were independent, 41% would favour keeping the deterrent, while 37% want it removed. Meanwhile, a majority (63%) of English and Welsh people questioned think the nuclear deterrent should be moved elsewhere if Scots vote Yes in September.

The independence issue might be skewing attitudes but the reasons for this are opaque. Some Scots might feel there is a jobs dividend which comes with Trident and that an independent Scotland might need the economic benefit of housing the weapons. That is open to question.

English respondents may simply feel that having weapons stationed on "foreign" territory would be counter-intuitive. But it is worth noting that the figures in Scotland echo a poll last year that recorded 43% of Scots favouring the retention of Trident and 39% wanting it removed.

It should be noted that the referendum is about whether and to what extent Scots should govern their own affairs, not about policy. But the timing is a challenge to First Minister Alex Salmond. Meanwhile, the survey challenges another SNP policy. Mr Salmond's vision of an independent Scotland is built on increased migration, with plans to raise net migration from an average of 22,000 to 24,000 a year.

The survey found a hardening of attitudes across the UK on what constitutes "Britishness" and an apparent concern about the impact of migration. However, it is hard to escape the feeling that ignorance lies behind much of the concern. For instance, one in four believes migrants come to the UK to claim benefits.

The survey's own authors point out that, despite the UK's points-based migration system, 42% thought it did not have one and 13% did not know about it. The number who think immigrants increase crime rates is up from 37% 10 years ago to 43% at present. With crime rates falling, this does not add up.

It is interesting that those who are more positive about the impact of immigration are the better educated and the better off. Do they feel less threatened? Or do they understand better how the system works? As 54% of Londoners believe immigration is a good thing, perhaps it is the latter.

Politicians must engage with public concerns, but what these findings call for is not a knee-jerk reaction.

Replacing Trident is increasingly hard to justify, whether within the UK or in an independent Scotland.

But the decision should be based on a mature analysis of cost and benefit (the latter justifiably coming under ever greater scrutiny).

Decisions about migration must always be based on fact, not myth.