Contact your Parish Council

Maidstone Housing Needs Survey 2005

A note on sampling methodology

1.    The Housing Needs Survey was carried out in partnership with Ashford Borough Council and provides robust data on which to base our future housing strategies and plan services.

2.    The Council commissioned a housing needs survey to examine housing requirements in urban and rural areas for all population groups and to predict demand over the next 10 years.  Our consultants reported in mid-2005.  The survey consisted of three elements:

·         A postal questionnaire to 9,550 households (some 16% of all resident households) in 20 wards/parishes;

·         A housing market survey utilising Land Registry and Halifax databases, together with an internet/telephone survey of estate agents;

·         Secondary data analysis drawing upon HIP and Housing Register data, 2001 Census, household and population projections and other national research.

3.    The questionnaire was in three parts:

·         Part One – sought information about respondents’ existing housing situation;

·         Part Two - sought information about respondents moving intentions;

·         Part Three – collected information on the moving intentions of newly forming and concealed households.

To create awareness of the survey and its importance, it was promoted across the borough through posters and a press release.

4.    Sampling methodology:  Sample size depends on two factors: the degree of accuracy required and the extent to which there is variation in the population concerning key characteristics.  However:

·         Beyond a certain size, there is no benefit in a bigger sample;

·         The size of the population is largely irrelevant for the accuracy of the sample – it is the size of the sample that is important.

5.    The survey was structured to achieve a 95% confidence level across the borough.  The postal sample was selected at random from the Council tax register and stratified into 20 wards/parishes.  In order to achieve statistical validity in the smaller wards/parishes (so that parish-based recommendations are founded in a reasonable degree of statistical confidence), it was necessary to sample near 100% of households in some parishes. 

6.    It can be seen that in order to ensure a reasonable number of responses – such that the confidence interval (CI) was as small as practicable – some wards/villages had near 100% sampling e.g. Hollingbourne/Hucking/Bicknor with 500 questionnaires being issued for 556 households.  This ensured that response rates from the questionnaires administered were similar at around 30% across the borough, and the CI was between ±6 to 8% in even the smallest ward, although due to the high overall sample size it was ±1.5% across the Borough.

7.    Postal survey process:  9,250 questionnaires were posted on 15/2/05, with a return deadline of 8/3/05.  An additional 300 questionnaires were dispatched to the four lowest response areas – Parkwood, Town east/north, Shepway and High Street.

8.    Survey weighting:  The data was checked against the 2001 Census housing tenure data and the HSSA[1] for bias and re-weighted where necessary.  Given the nature of the random sample in the postal survey, tenure type was chosen a providing the main validation of the representativeness of the sample.

9.    The second element of the survey drew on Land Registry and Halifax databases to assess the cost of access level property, and the supply and cost of private rented sector housing.  An Estate Agency survey was carried out to assess entry prices for new households in each sub-area.

10. And thirdly, an analysis was carried out on HIP[2] and Housing Register data on the flow of social stock and need.

11. Age of data:  Wherever possible, the data used was as at 31 December 2004, which was the most recent available.

12. Details of the sub-area sampling and response rates are tabled overleaf.  The sub-areas shown represent housing market areas within the Borough.

Response rates by Ward