This study is part of a larger project called Contraceptive Use, Pregnancy Intention and Decisions (CUPID) of young Australian women. CUPID aims to shed light on the reasons for using, or not needing, contraceptive methods among young women, despite the apparent widespread availability of contraception. This study aims to identify how to encourage the participation of young women in a web-based survey, about contraception and pregnancy, through face-to-face discussions with young women. Ethical approvals were granted by committees at the University of Queensland, University of Newcastle, and Family Planning NSW. Additional information on CUPID is available from the project website.

Young women aged 18-23 years and living in urban and rural NSW, Australia, were invited to participate in focus group discussions. Advertising for the discussions was via posters and information sheets displayed at the University of Newcastle, Family Planning NSW clinics (Ashfield, Fairfield, Penrith, Newcastle, Dubbo), educational settings (library), and one workplace setting with a trainee program including young women aged 18-23 years. Recruitment was achieved using direct contact (face-to-face, telephone, or email) and potential participants inviting their friends (ie, snowballing sampling). No financial incentives were given for participation in the study, but refreshments were provided during the discussions. Women who offered to participate but were aged outside the specified age range were excluded.

Ten semi-structured focus group discussions (lasting between 30 and 60 minutes) were conducted with participants from August to September 2011 by one member of the research team (DH) and an assistant (CR). The focus groups were held in urban areas at clinical locations in Sydney (Ashfield, Penrith) and Newcastle. In rural areas, the focus groups were held at clinical (Dubbo), educational (Tamworth), and workplace (Muswellbrook) locations local to the participants. Interviews were conducted with single participants where it was not possible to form a focus group of two or more participants aged 18-23 years, or the participant specifically requested to have a one-on-one discussion. The interview followed the focus group question schedule.

Before commencing the discussion, participants were given an information sheet describing the aims of the study, a consent form, and short survey to gather information on age, study or work patterns, and Internet use. All discussions were audio recorded after written consent was provided by the participants. A question schedule was used to stimulate discussion and included open-ended questions to encourage participants to give a broad range of responses. The interviewer (equally DH or CR) prompted the participants where appropriate to elicit more detailed information. Participants were asked to avoid revealing their personal experience of contraception and pregnancy but to focus on their reaction to survey questions on these topics. Other topics included access to the Internet, preference for using computers, smartphones or other devices, and responding to sensitive questions. The non-interviewing researcher (equally DH or CR) took notes throughout and summarized the discussion to the participants at the end of each discussion and invited corrections to the recording.

All discussions were de-identified, transcribed verbatim, and manually analyzed using thematic analysis. Our overarching aim was to identify important recommendations for research about contraception and unplanned pregnancy. We were interested in examining broad themes with which to frame recommendations, rather than detailed content analysis. We therefore employed a grounded theory approach [22] to examine the comments provided in the interview about participation challenges and survey challenges. Printed interview transcripts were read by DH. Statements that, directly or indirectly, referred to participation or survey challenges were identified. Excerpts were sorted into initial groupings by DH. These excerpts revealed five themes relating to participation challenges and seven themes relating to survey challenges. The research team reviewed these themes and agreed on the coding framework. All interviews were then coded by DH according to the identified themes using NVivo 9 (QSR International Pty Ltd., 2010). JL reviewed the transcripts and coding in detail and any minor discrepancies were resolved, and interpretations developed, with in-depth discussion to reach a consensus.

A recurring theme in the discussions was the ongoing cost of using the Internet, either on a smartphone, mobile phone or home computer: “But yes, being on the Internet and it’s going to cost you money ... because the Internet is pricey ... it’s just once you’re on there it’s like tick, tick, tick, money gone.” The constant awareness of the increasing Internet charges discouraged participation in a web-based survey. However, free access to social media through most mobile network service providers resulted in most participants using social media for almost all personal communications: “At the moment because I’m with [mobile network provider] ... you get free Facebook access, so that’s the only reason why I go on Facebook, because it’s free.” The majority of participants suggested the use of social media as the “only way” to engage their attention and raise awareness of the web-based survey.

An incentive to participate was desired by almost all participants: “I think it’d probably have to be like $20 [gift card] ... I do think that people do sometimes get a bit greedy, and it’s pretty sad that people just won’t do it to help people out.” Privacy and confidentiality were key concerns about participating in web-based surveys. Participants were concerned about providing their contact details especially if there would be future unexpected contact: “... and they want to interview me or, I don’t know, talk to me over the phone, then I wouldn’t necessarily give it.” Further, participants were worried about who wanted their information: “It’s not a freaky guy behind the computer just trying to get information about you”. Participants also wondered how their responses would be used: “Just you’re worried about where your information is going to go, and who’s going to get it.”

Participants wanted the survey to be as short as possible (15 minute maximum) otherwise they were likely to leave the web-based survey and move to other websites. Asking sensitive questions was appropriate as long as “I prefer not to answer” was given as a response option:

In the absence of this response option, most participants admitted they would give a false response in the survey.

Participants wanted to receive a combined feedback and reminder to complete the follow-up surveys because this information indicated their prior participation had contributed to some results: “… if you get some results back so you know... this is what your information gave, so then the next time they know it’s actually going to do something ... they’re more willing to do the survey in the next six months.” Participants from rural areas identified the benefit of using the feedback to advocate for change in their own community:

Many participants suggested certain personality types who were proactive, organized, and appreciated the benefits of research would always be more likely to participate in a web-based survey than other types of young women.

This study examined two key challenges to researchers in engaging young Australian women in studies of contraception and pregnancy. These areas are (1) ensuring adequate participation and (2) appropriate survey design to ensure the web-based survey is completed. Challenges relating to achieving adequate participation in a survey included the adequacy of participants’ access to the Internet, the use of social media, and avoidance of additional financial costs to access the web-based survey for young women living with socioeconomic disadvantages. Financial incentives were identified as the most effective method of ensuring recruitment of participants for short web-based surveys. Survey design challenges included the need for the survey to be short, to include the option to not answer questions, and for participants to receive feedback about the results. These challenges indicated that web-based surveys should be tailored to the needs of young women, especially those living in lower socioeconomic circumstances.

The web-based recruitment of research participants have been found to overcome the reluctance of young people to attend face-to-face discussions [23] and can reach hidden populations, eg, people with disabilities [10]. The use of the most popular social networking site, Facebook, has been identified as effective for web-based recruitment [14]. Fenner et al (2012) found that from 7940 Australians who clicked on the Facebook advertisement of their project, 3.5% went on to participate in the study. Among the young female participants, 28% were from a rural area, which was higher than the 21% of the target population from a rural area [14]. A US study of adult substance use also recruited through targeted Facebook advertising and successfully recruited 10% (n=1548) of the 14,808 people who clicked on the Facebook advertising, met the inclusion criteria and completed the survey [17]. However, web-based recruitment cannot occur when there is a digital divide beyond economics, eg, people living in rural Australia with unreliable or non-existent Internet availability. Notwithstanding these limitations, the promotion of research participation to young people as a means to have a voice, be heard, and contribute to their community may be the key driver to participate [24]. The findings of this study have shown that young women, especially those in rural areas, may be willing to participate in web-based research on issues important to them that will help them in their local community. Further, the clear message from young women was the need for a social media presence to grab their attention and ensure their participation.

One option that has been used to counteract the demographic challenge presented by the unequal numbers of young women in rural and urban areas of Australia is the use of a stratified random sampling frame to ensure young women living in rural areas are well represented and the findings are not urban-centric. Such a sampling frame is an established method, eg, the ALSWH has a successful history of recruiting and maintaining three generational cohorts of Australian women using mailed invitations and surveys [6,25]. However, the preference of young women to use social media or email instead of postal mail limits the effectiveness of using a similar methodology in a new generation of young women [14,15]. Participants in the ALSWH and in the Fenner et al (2012) study were broadly representative of the Australian female population with recognized over-representation of higher educated women whether the participants responded to postal invitations [6] or Facebook advertising [14]. A limitation of web-based recruitment, eg, social media, and subsequent volunteer participation is the potential loss of sample representativeness and generalizability of the findings. However, a very low response to postal invitations based on a stratified random sampling frame would also result in a non-representative sample of young women. A challenge for researchers is to adapt to generational changes within their target population and choose the least limiting web-based methodology while ensuring sufficient statistical power from their sample.

This study has identified key design features of a web-based survey to increase the likelihood of young women’s ongoing commitment to follow-up web-based surveys. Young women need to be reassured that their anonymity and confidentiality are maintained. They need to feel supported and secure that the researchers conducting the survey are legitimate and trustworthy by having recognizable logos and contact details. Young women agreed that university and government logos were seen as more legitimate than other novel logos viewed on websites, and these logos would encourage their participation. Feedback was imperative to ensure their continued commitment to the research. A study from the University of Wisconsin, USA, showed the suggestion of sending future email reminders to finish a partially completed survey was sufficient to encourage participants to immediately finalize the survey [26]. Further, the findings reported from this current study showed young women had a limited tolerance of reminders and preferred a short survey to be completed as quickly as possible.

This study successfully engaged a group of young women from rural and urban NSW in the development of a web-based survey of contraception and unplanned pregnancy. The findings of these discussions have implications for the research design and implementation of web-based surveys about sexual activity, contraception, and pregnancy, which must include potentially sensitive questions [27]. It is imperative to offer young women an appropriate range of response options to avoid forced responses that create false or blank reporting [27]. Reliable data are needed in order to make recommendations for policy and clinical practice interventions to reduce rates of unintended pregnancy.

Information about contraception use and pregnancy intention in young Australian women is essential to inform policy but data are lacking, especially for those living in rural areas. Based on our findings, future research efforts are needed to understand the full extent of the role social media and incentives play in the decision of young women to participate in web-based research.