The Gay Men’s Sex Survey uses a self-completion questionnaire to collect a limited amount of information from a substantial number of men. The sample is self-selected and opportunistic. We do not pay men to take part nor undertake prize draws or promotions of any type. Instead we stress the benefit to the community of an individual’s participation.
Since 1997, the survey has been undertaken eleven times, with funding from Terrence Higgins Trust as part of the CHAPS programme funded by the Department of Health. During this time it has expanded across England and included men resident in Wales (since 2000); Scotland (since 2001); Northern Ireland (since 2002) and Republic of Ireland (since 2003). The main final reports include men from the whole of the UK, but each UK country and the Republic of Ireland are reported separately in data by country.
Since the Gay Men's Sex Survey first occurred across England in 1997 the number of men recruited every summer has risen from 5,000 to a peak of almost 20,000 in 2002, when all three methods of recruitment were used across the United Kingdom. Since we returned to only two methods of recruitment in 2003 the number of men taking part has varied from 13,000 to 19,000 with this variation largely a function of the efficiency of our online advertising spend.
Number of participants in the Gay Men's Sex Survey 1993-2007
Of all the people recruited to the survey each year, between 15% and 25% are subsequently excluded from any analysis. Those excluded include respondents that have answered less than 20% of the questions; those that are female (or intersex); those under 14 years of age; and those living outside the UK. We also exclude men that have no evidence of sex with men in the last year and either no gay, bisexual or other similar sexual identity or no intention to have sex with a man in the future.
Hence the final samples are men, aged 14 or older, with homosexual experience in the last year or a gay, bisexual or similar sexual identity or an intention to have sex with men in the future and who live in the UK (or Republic of Ireland for the all Ireland reports).
Final sample sizes in the Gay Men's Sex Survey 1993-2007
In the nine years we have used the booklet method the proportion of recruits excluded has varied between 4% (in 2005) and 14% (in 1999). It varies because the precise methods used by distributors, and the actual agencies doing distribution of the booklet changes from year-to-year.
In the seven years we have undertaken the survey online the proportion of recruits excluded has varied between 15% (in 2003) and 31% (in 2001). The majority of men recruited online but excluded are not UK-resident - 13% to 23% of the online sample have been excluded for non-UK residence in the last seven years compared to 0.5% to 3% of booklet-recruited men.