Panel Recruitment and Representativeness

Great research builds from great sample recruitment. Opinionpanel takes this proposition so seriously that we've addressed the two greatest conundrums facing the online research industry: how to reduce recruitment bias and how to improve identity verification.

Population inclusion

Since 2004 we have recruited our panels by email invitation sent by UCAS. UCAS is the UK's central body for handling undergraduate applications and is therefore able to invite almost the whole of each starting cohort. This means there is very little systematic bias in the student portion of our panel - almost the entire undergraduate student population has the opportunity to join. This massively reduces the likelihood of sampling error.

Identity assurance

Professional and fraudulent respondents could - and in fact do - join the same panels many times simply by creating multiple email addresses. If these respondents are invited twice to the same survey, are they likely to answer both surveys honestly and diligently? It's doubtful. All our student respondents have verified 'ac.uk' academic email addresses, which means only bona fide students take part and, just as importantly, because every 'ac.uk' address is unique, this verification process almost eliminates the danger of participants with multiple identities - a real problem for some consumer panels. All members of The Graduate Panel are former members of The Student Panel. All members of The Future Panel are recruited by UCAS.

Making representative samples

Once the panel is recruited, how do we ensure that the samples we draw for particular surveys are representative? It's quite straightforward, actually. Student population statistics are some of the most detailed and up to date in the UK. We receive detailed figures from the Higher Education Statistics Agency (HESA), enabling us to replicate the demographic and academic of most sub-groups you could think of. We are able to set quotas for attributes including course, university, year of programme, mode and level of study, age, gender, region, ethnicity, domicile, fee eligibility, parental socio-economic classification, school type (e.g. state or private), and tariff points.

Response rates and incentives

Good data collection is about validity as well as representativeness. Our panellists are highly responsive (we average about a 50% response rate to surveys). We believe this is due to a combination of a good incentive programme, relevant surveys, careful panellist engagement through a dedicated panel manager, quarterly newsletter updates on recent survey results and our 'article of the month' programme. The incentive scheme gives BonusBond high-street shopping vouchers to members of The Future Panel, which are sent by post to panellists after each survey. For students and graduates, incentives are based on accruing points which get converted into Amazon Gift Certificates. All panellists start with a credit balance of 10 points. Once a balance of 25 points is reached, they are emailed a £25 Amazon Gift Certificate. As a loyalty bonus we then credit their (now empty) account with a new 'base' of 10 points. To panellists, 1 point is, therefore, effectively worth a credit of £1.67 in Amazon Gift Certificates.