Overall, 2,094 questionnaires June and November 2001. Some contained the replies of more than one person in the household. All in all responses were received from 2,313 individuals. This includes the responses of 64 children aged 6 - 8 who did not answer the main questions, but who did project work which we have coded as answers to the comments section of the questionnaire form. The sample for the analysis of the main questions therefore comprises 2,249 people. This is much larger than the typical sample of 1,000 which is commonly used by pollsters.


The number of questionnaires returned represents a response rate of 5.1% (see note after Table 2). We are advised that cold-calling market research normally attracts a response rate of less than 2%. Furthermore, we have performed an analysis (see Appendix A) on the responses which reinforces the hypothesis that the sample was big enough to draw statistically valid conclusions. So we are satisfied with the level of response and the coverage, which came from a wide area around the Park; see Figure 7 below (for full response list see Table 1 and Table 2).

Figure 7 : Responses to questionnaires - by postcode


Naturally, we could not obtain answers from every single household in the community, and those responding to the questionnaire are, of course, self-selecting. We have been concerned, therefore, to test whether the replies received are representative of the park-using public. We did this in four stages.


First, we conducted a survey in the Park (see Table 19). Members of the public were randomly selected and asked to complete the consultation questionnaire. This produced 245 questionnaire responses.


Second, we carried out Park survey responses by plotting the cumulative results, which tested whether enough responses had been received to enable valid conclusions to be drawn. This demonstrated the required level of consistency, so that it was not necessary to carry out further consultation of the Park users (see Appendix A).


Third, we carried out a similar cumulative data analysis on all of the questionnaires returned, which again demonstrated that samples were of sufficient size (see Appendix A).


Fourth, we matched responses on key questions from the overall sample with the responses to those questions from the Park sample. These demonstrated a remarkable level of consistency. The results for the Top Site, managed parkland question may be seen below in Figure 8.

Figure 8 : Top Site YES responses to "managed parkland"
&endash; all respondents and park respondents compared


For example, 70% of all respondents said they would like the Top Site to be used for ecology, 16% said they would like a commercial leisure centre there, and 42% a building. The results for the park users were 72%, 17% and 37% respectively. Similarly close correlations can be seen from the responses to questions regarding the Main Park and the Sports Centre.


These results tend to show that the responses from the general public are representative of the park users as a whole. While, therefore, it would be possible to carry out further sampling, we are satisfied that the results we have achieved are a valid tool for determining the initial views of the community and are a platform for further consultation.

Top of page
Report Contents
Go to Chapter 3 -Distribution (previous page)
Go to Chapter 5 - Methodology (next page)

Crystal Palace Campaign March 2002 - Consultation Starts Here
Copyright: Day, Kolvin, Sacks 2002
Last updated: 26 March 2002