1. How we did it
The questionnaire was sent out in the post to 3,000 randomly selected households from throughout the District. These households were selected completely at random from the Land and Property Gazeteer database, which is the most comprehensive and up to date list of addresses.
Respondents receiving one in the post were given the option to complete it online if they wished.
A similar questionnaire was sent out in June 2013, June 2014 and June 2016 using the same method, so results can be compared over time. Where accurate comparisons are available the results for previous surveys are included.
We received 607 responses altogether (this compares to 772 in 2016). This was 59 online responses (a decrease of 23 online responses from last year) and 548 paper based.
The results were weighted (according to best practice) for Ward and age. This is a technique used to get accurate levels of representation of certain groups in the District. Data for under represented age groups and Wards was weighted to make the sample a better representation of the actual population. For instance, if the population is known to be 50:50 male to female but the sample has 48:52 then weighting is used to adjust the sample data to represent the true population. In this case it would be done by giving each male in the sample a weight of 50/48 and each female a weight of 50/52.
NOTE: None of the comments given by respondents are weighted like the quantitative (numerical) statistics are. They are reflecting the comments of the actual age groups and Ward profiles that initially responded to the questionnaire. Where the ‘% of all 607 respondents’ is shown in the categorisation of comments this is not representative of the actual age groups or Wards within East Devon, purely the % of respondents to this survey.