The final results for the Windsor residents’ survey has shown overwhelming support for its proposals, with 95% of Windsor residents agreeing that it is important to Windsor and 92% petitioning the council to look actively at the ideas.
The survey targeted the centre of Windsor (the Castle Without ward) as this is the area most affected by the proposals, both positively in terms of new rail services and improvements to the riverside environment as well as negatively, e.g. from disruption during construction.
Residents were invited to participate by a post card delivered to 3,000 addresses between 16 March and 2 April 2013. The survey was also promoted via Facebook and local restaurants as well as the RBWM website. The online survey was open from 3 March to 9 April. 325 responses were received in total of which 285 were completed. This included 24 hardcopy returns, where residents could return the postcard indicating their support.
The population of the Castle Without ward is about 5,000, according to the latest census data, and as such the survey is a statistically significant sample of the target audience with a ±10% margin of error at 90% confidence.
The survey was also tested for representativeness by age and by method of going to work. The ages of respondents broke down as follows.

The mode of commute, that is method of travelling to work, of respondents was as below. There was no statistically significant difference in replies according to this answer, as can be seen from the chart below. Car users, for example, gave similar answers to train users.

The survey was also tested for geographical representativeness. Responses were evenly spread over the residential areas of central Windsor and the surrounds. The slider graphic above illustrates both petition responses across the borough and answers to question 1 in central Windsor.
The rate of positive responses was slightly higher for valid respondents compared with that for all respondents (including those who didn’t complete the survey or gave fake post codes etc.) For example, completed valid surveys saying phase 1 was important for Windsor was 95% but only 93% for valid respondents as a whole and 90% for all respondents. At first glance, this may seem to imply that people living outside of Windsor are less supportive of the railway than those living in the town. This is counter-intuitive as any disadvantages (e.g. from building works) would be entirely for Windsor whereas there would be only benefits for those travelling through on additional services. A possible explanation is that people outside Windsor are jealous of the potential improvements and have thus been somewhat ungenerous in their responses. However, a better explanation would appear to be that this is simply sample bias. That is, the returns from outside Windsor are not statistically significant and are thus showing bias which would reduce if a representative sample size were collected. This likely also applies to respondents who didn’t complete the survey or who provided false verification information, fake post codes etc. For this reason, only validated, completed responses in the target area are counted in the results unless otherwise stated.









