Shared equity schemes: evaluation reports

Annex: Buyer survey methodology

The buyer survey was issued to all buyers using HtB, NSSE or OMSE for whom contact details were available. Buyer survey fieldwork incorporated two strands; email survey invites issued to buyers for whom an email address was held, and a 'postal web drive' comprising a letter with a short link to the websurvey issued to all buyers for whom an email address was not recorded. Options to respond via online survey, paper copy and telephone were offered to all buyers.

Table 12 provides an overview of the survey sample structure and response. Of the total of 25,288 surveys issued, 16,622 (66%) were issued via email and 8,666 (34%) by post. This mix of email and postal web drive surveys varied across the three schemes; postal web drive was used for 44% of HtB, 100% of NSSE and 1% of OMSE buyers. A postal reminder letter was also issued to a random sample of 10,000 non-respondents at the mid-point of survey fieldwork, equivalent to 46% of non-respondents at the time.

A total of 4,063 survey responses were received, representing an overall response rate of 16%. The response rate varied across the three schemes (14% for HtB, 13% for NSSE and 20% for OMSE). Response rates also varied dependent on survey method, being higher for those who initially received the survey by email (20% response) and lower for those who initially received the survey by postal web drive (9%). This appears to have contributed to differences in response across the shared equity schemes; response rates were highest for OMSE buyers (with the largest proportion of surveys issued via email) and lowest for NSSE buyers (all of whom received the postal web drive).

* The 'postal web drive' response includes those from buyers included in the initial email sample who received a postal reminder at the mid-point of survey fieldwork.

Applying weights to data, while tending to make the quoted figures more representative of the population of interest, also serves to reduce the statistical reliability of the data. As such the 'effective' base size, which is used in any statistical testing, is smaller than the unweighted base size. This effect has been taken into account in determining whether or not differences described throughout the report are statistically significant.

Table 13 compares the profile of survey respondents with all buyers who received the survey, based on shared equity scheme and urban/rural geography (using the Scottish Government's 6-fold urban/rural geography).^[22] As a result of differences in response rates across the three shared equity schemes, survey responses under-represent HtB buyers (by 7 points) and over-represent OMSE buyers (by 8 points). Areas of more limited over and under-representation are also evident across some geographic groupings. Survey responses were weighted by shared equity scheme and the Scottish Government's 6-fold urban/rural geography to correct for this.

Survey analysis involved producing frequency results for all survey questions. This included cross-tabulation of all questions by the following key buyer groups: shared equity scheme used, first-time buyers, purchase price, initial equity share, year of purchase, urban/rural geography and household type.

Differences across respondent groups are only included in the body of the report if statistically significant at a 95 per cent confidence level.