Challenges of internet recruitment: a case study with disappointing results.
Challenges of internet recruitment: a case study with disappointing results.
复制标题
DOI：
10.2196/jmir.7.1.e6
复制
发表时间：
2005-03-19
影响因子：
7.4
通讯作者：
Skinner H
中科院分区：
文献类型：
--
作者：
Koo M;Skinner H
互联网为创新研究提供了巨大的机会，但关于利用互联网招募研究参与者的出版物很少。本文总结了我们的经验，从两个研究中，我们试图招募青少年在互联网上的问卷调查研究，以评估一个戒烟网站。 评估通过互联网招募青少年评估戒烟网站的策略。在研究1（定义社区招募）中，我们向青年健康网站CyberIsle的注册成员发送了邀请电子邮件。共有3801个电子邮件地址被随机分为两组。在第一组中，电子邮件表明，前30名受访者将收到一个20加元的电子礼券，用于在网上书店使用，如果他们去吸烟锌网站，并回答一个简短的调查。对于第二组，电子邮件还表明，如果受访者将他们的朋友介绍给这项研究，他们将获得额外的10加元礼券。提醒电子邮件在发送初始邀请电子邮件后10天发送。在研究2（公开招聘）中，我们在网络讨论板、Usenet论坛和一个专门的招聘网站上发布邀请信息，并尝试了滚雪球式的招聘策略。当潜在的参与者到达研究中心时，他们被自动随机分为高激励组（15加元电子礼券）或低激励组（5加元礼券）。在研究1（定义的社区招募）中，成功发送了2109封电子邮件。只有5名受试者（0.24%）（包括朋友推荐的1名受试者）通过了招募过程并完成了问卷;另有6名受试者访问了研究信息页面，但未完成研究。在研究2（公开招聘）中，看到广告的用户数量是未知的。共有35名用户访问了该网站，其中14名参与者被招募（8名来自15加元礼券组，6名来自5加元礼券组）。另外5人是从一般互联网社区招募的（3人来自讨论板，2人来自研究志愿者网站）。剩下的9名参与者是通过朋友推荐和雪球策略招募的。总体而言，征聘率低得令人不安。在我们的案例中，使用互联网技术，包括电子邮件，电子讨论板，Usenet论坛和网站的招聘并没有被证明是一个有效的方法，征求年轻的主题参与我们的研究。讨论了可能的原因，包括参与者的观点。一个主要的挑战是区分可信的和合法的邮件从垃圾邮件和欺诈性的错误信息在互联网上。从研究人员的角度来看，需要采用更大的样本，验证参与者的属性，并评估和调整与互联网招聘相关的潜在偏见。
The Internet provides tremendous opportunities for innovative research, but few publications on the use of the Internet for recruiting study participants exist. This paper summarizes our experiences from 2 studies in which we attempted to recruit teenagers on the Internet for a questionnaire study to evaluate a smoking-cessation website. To evaluate strategies of recruiting teenagers for the evaluation of a smoking-cessation website through the Internet. In Study 1 (Defined Community Recruitment), we sent invitation emails to registered members of a youth health website, CyberIsle. A total of 3801 email addresses were randomly divided into 2 groups. In the first group, emails indicated that the first 30 respondents would receive a Can $20 electronic gift certificate for use at an online bookstore if they would go to the Smoking Zine website and respond to a short survey. For the second group, the email also indicated that respondents would receive an additional Can $10 gift certificate if they referred their friends to the study. Reminder emails were sent 10 days after the sending of the initial invitation email. In Study 2 (Open Recruitment), we posted invitation messages on Web discussion boards, Usenet forums, and one specialized recruitment website, and attempted a snowball recruiting strategy. When potential participants arrived at the study site, they were automatically randomized into either the higher incentives group (Can $15 electronic gift certificate) or lower incentive group (Can $5 gift certificate). In Study 1 (defined community recruitment), 2109 emails were successfully delivered. Only 5 subjects (0.24%), including 1 referred by a friend, passed the recruitment process and completed the questionnaire; a further 6 individuals visited the information page of the study but did not complete the study. In Study 2 (open recruitment), the number of users seeing the advertisement is unknown. A total of 35 users arrived at the website, of whom 14 participants were recruited (8 from the Can $15 gift certificate group and 6 from the Can $5 gift certificate group). Another 5 were recruited from the general Internet community (3 from discussion boards and 2 from the Research Volunteers website). The remaining 9 participants were recruited through friend referrals with the snowball strategy. Overall, the recruitment rate was disappointingly low. In our case, recruitment using Internet technologies including email, electronic discussion boards, Usenet forums, and websites did not prove to be an effective approach for soliciting young subjects to participate in our research. Possible reasons are discussed, including the participants' perspective. A major challenge is to differentiate trustable and legitimate messages from spam and fraudulent misinformation on the Internet. From the researchers' perspective, approaches are needed to engage larger samples, to verify participants' attributes, and to evaluate and adjust for potential biases associated with Internet recruitment.