The survey was conducted by a profe

The survey was conducted by a professional firm of telephone interviewers under the guidance of the author. This approach was adopted for reasons of efficiency and to raise the response rate from a limited sample size. The telephone interviewers also aimed to ensure that questions were answered as completely as possible. Full interviews were obtained with 50 individuals representing a response rate of 25 per cent. There was no reason to suspect bias due to non-response, as discussed in the Appendix. All interviews were completed within four weeks of Spring 2002. Scores were compiled according to a Likert-style scale.