Quality issues for the research on forced marriage and the potential effects of raising the minimum age for a sponsor, and of

Quality issues for the research on forced marriage and the potential effects of raising the minimum age for a sponsor, and of leave to enter the UK as a spouse or fiancé(e)

Individuals experiencing forced marriage are a hard-to-reach group and the methodology used in the research was deemed the most effective. However, the reporting undermines the reliability and validity of the findings for the following reasons.

Unsubstantiated reporting of findings

A number of statements cannot be substantiated due to insufficient sample sizes and a lack of research evidence. Conclusions such as, mental illness as a significant risk factor among men, cannot be justified on the basis of a single respondent. The numbers of respondents who cited different points is not always given making the findings misleading.
The type of respondents, for example, individuals, organisations, victims and survivors included in different groups cannot be established. Therefore it is not possible to identify the specific groups for and against an increase in age.
The quantitative data obtained from the mapping survey was insufficient to draw any conclusions regarding the prevalence of forced marriage, due to the small number of participating organisations.

Unclear terminology

The report frequently used unclear and poorly defined terminology, such as `control over sexuality'. These terms are not generally understood by the general public.

Sampling bias

The nature of the research may have created distrust and suspicion amongst research participants and a sampling bias in favour of organisations and individuals who did not have a critical or radical perspective on the issues being explored. Participants from some groups might perceive the study as instrumental in further restricting immigration and in potentially interfering with cultural practices around arranged marriages.
It must be borne in mind that the research was undertaken in only three case study areas, therefore the findings are not representative of the whole country. Additionally, there was a sample bias towards South Asian populations.
It was necessary to use snowballing techniques and agencies to facilitate access to all participants and particularly victims/ survivors of forced marriage. It was not therefore a representative sample. Snowballing introduces bias as it favours those who are active in the community, and includes only the people who want to be interviewed or to whom agencies allow access.
Some respondents' beliefs regarding forced marriage may have introduced bias. The term `forced' can be misleading because it presupposes that physical force must play a part for a marriage to be forced. Furthermore the widely held view among respondents that forced marriage only occurs in South Asian communities may have reduced response rates.

Overall, the findings should be treated with considerable caution and can only be considered as indicative and not representative of all communities or community organisations.

Immigration Research and Statistics. 3^rd April 2008.