Recruiting online: Lessons from a longitudinal survey of contraception and pregnancy intentions of young Australian women

Recruiting participants for health surveys has become increasingly difficult, particularly where young people are concerned. Traditional recruitment practices involve methods such as postal, telephone and face-to-face invitations, which may not be compatible with the high usage and reliance on more modern forms of technology amongst younger people.

In recent years researchers have made use of social media as a means of recruiting young participants into health surveys, but it is not yet clear whether this is a cost-effective approach, or whether it can provide samples that demographically reflect the general population, particularly for longitudinal research.

In the current study, the researchers assessed the effectiveness of online recruitment methods in recruiting women aged 18-23 for the Contraceptive Use, Pregnancy Intention, and Decisions (CUPID) Study. A variety of methods were used, including Facebook advertisements and posts, Twitter, and online forums, as well as face-to-face events, distribution of promotional material, and media releases.

Over the one-year recruitment period, a total of 3,795 women were recruited to take part in the online survey, almost double the original target of 2,000. Women were recruited at an average cost of $11 per participant, substantially lower than that for the pilot version of the study, which, using mailed invitations, attracted only 54 participants at a cost of around $100 each. The sample was found to be broadly representative of 18-23 year old women in Australia in terms of demographics, with the exception of a higher proportion of women who had completed year 12 education.

While the use of multiple approaches makes it difficult to determine the success of individual strategies, Facebook appears to be a particularly effective means of recruitment, with a large daily increase in respondents observed following changes to the placement of advertisements from the sidebar to the central newsfeed.

facebook-793049_1280The findings from the current study suggest that it is possible to recruit a demographically representative sample of young women using online methods, and that this can be done at a reasonable cost.

This paper generated significant interest, with an invited commentary commissioned by the American Journal of Epidemiology. In this commentary Dr Jenifer Allsworth from the University of Missouri – Kansa City praised the study design, suggesting that the study was well conducted and made an important contribution to the literature.

Contact person: Melissa.Harris@newcastle.edu.au

Citation: Harris ML, Loxton D, Wigginton B, Lucke J. Recruiting online: Lessons from a longitudinal survey of contraception and pregnancy intentions of young Australian women. American Journal of Epidemiology, 2015. doi: 10.1093/aje/kwv006

Postal recruitment for young women is not cost effective

What you need to know:

Researchers have often used mailed invitations to ask people to participate in surveys. It is important to have enough people participate, and that they represent the population of interest. This study showed that postal recruitment methods are not cost effective in achieving representative samples of young people. When we chose young women from the Medicare Australia database and mailed them an invitation to participate in a survey only 5.4% of eligible participants completed the entire survey. At a cost of around AU$100 per participant this was not cost-effective. Innovative and flexible approaches with a focus on social media may hold the key.

What this research is about:

Population-based studies usually rely on randomly sampling methods to achieve samples that are representative of the larger population. In Australia, this has previously been achieved by using Medicare Australia’s health insurance database. This database has the most up-to-date age, sex and contact details of Australian citizens and permanent residence. The Australian Longitudinal Study on Women’s Health was able to successfully recruit a sample of 18-23 year old women in 1996 at a cost of AU$30 per participant using this approach. Since this time, response rates to population studies have declined. In this study we report on the cost-effectiveness of recruiting a representative sample of women aged 18-23 years living in NSW using the Medicare database for an online survey about contraception and unintended pregnancy.

What did the researchers do: 2

Using a stratified sampling frame (based on age and area of residence), a total of 900 young women were invited to participate in the pilot study. Assuming an 18% response rate, it was estimated that 150 participants would be achieved. Invitations were distributed by Medicare Australia on behalf of the research team in two batches using a modified Dillman protocol.

What did the research find:

Only 6.2% of eligible participants consented and initiated the survey. Of these, 5.8% completed at least one survey item and 5.4% completed the entire survey. The cost per participant was around AU$88 for survey initiation and AU$100 for survey completion.

How can you use this research:

Postal recruitment methods are not cost effective for achieving representative samples for young people. In a changing technological landscape, innovation and flexibility in recruitment strategies are required. Internet-based technologies (including social media) may hold the key but further research is required regarding the ability to achieve representative samples using this approach.

Contact person: Dr Melissa Harris: Melissa.Harris@newcastle.edu.au

Citation: Harris, ML, Herbert D, Loxton D, Dobson AJ, Wigginton B, Lucke J. Recruiting young women for health surveys: traditional random sampling methods are not cost-effective. Aust N Z J Public Health. 2014, 38 (5); p 495.