The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., 6720A Rockledge Drive, Suite 100, Bethesda, MD, 20817, USA.
Naval Health Research Center, 140 Sylvester Road, San Diego, CA, 92106, USA.
BMC Med Res Methodol. 2018 Oct 24;18(1):114. doi: 10.1186/s12874-018-0575-x.
While enrolling dyads in research studies is not uncommon, there is limited literature on the utility of different recruitment strategies and the resulting selection biases. This paper examined two recruitment strategies used to enroll military couples in a longitudinal study, assessing the impact of both strategies on the representativeness of the final study sample.
Descriptive and bivariate analyses were conducted to 1) identify characteristics associated with spouse referral, 2) compare response rates based on recruitment strategy and assess whether recruitment strategy modified correlates of response propensity among spouses, and 3) assess whether referred spouse characteristics differed from non-referred spouses in the final sample. The study sample consisted of married US service members with 2-5 years of military service as of October 2011 and their spouses.
Service members who referred their spouses to participate in the Millennium Cohort Family Study were more likely to be male, have children, serve in the Army, and have combat deployment experience than those who did not refer their spouse. Nearly two-thirds (n = 5331, 64.9%) of referred spouses participated in the Family Study, compared with less than one-third (n = 3458, 29.5%) of directly contacted spouses. Spouse characteristics also differed significantly between recruitment groups.
Overall results suggest that minimal bias was introduced by using a referral recruitment methodology. Service members appeared to be more likely to refer their spouses if they perceived the research topic as relevant to their spouse, such that male service members with combat deployment experience were more likely to refer female spouses caring for multiple children. Referred spouses were significantly more likely to respond to the Millennium Cohort Family Study survey than those who were directly contacted; however, the overall success rate of using a referral strategy was less than recruiting spouses through direct contact. Differences between referred spouses and spouses contacted directly mirrored service member referring characteristics.
尽管招募对偶个体参与研究并不罕见，但关于不同招募策略的效用以及由此产生的选择偏差的文献有限。本文研究了两种招募策略，用于招募军事夫妇参与一项纵向研究，评估这两种策略对最终研究样本代表性的影响。
进行描述性和双变量分析，以 1）确定与配偶推荐相关的特征，2）比较基于招募策略的响应率，并评估招募策略是否改变配偶响应倾向的相关性，以及 3）评估最终样本中被推荐配偶的特征与未被推荐配偶的特征是否存在差异。该研究样本包括截至 2011 年 10 月，有 2-5 年军龄的已婚美国现役军人及其配偶。
与没有推荐配偶的军人相比，推荐配偶参加千年队列家庭研究的军人更有可能是男性、有子女、在陆军服役且有作战部署经验。近三分之二（n=5331，64.9%）被推荐的配偶参加了家庭研究，而不到三分之一（n=3458，29.5%）直接联系的配偶参加了家庭研究。配偶特征在招募组之间也存在显著差异。
总体结果表明，使用推荐招募方法引入的偏差最小。如果军人认为研究主题与配偶相关，他们更有可能推荐配偶，因此有作战部署经验的男性军人更有可能推荐照顾多个孩子的女性配偶。与直接联系的配偶相比，被推荐的配偶更有可能对千年队列家庭研究调查做出回应；然而，使用推荐策略的总体成功率低于通过直接联系招募配偶。被推荐的配偶和直接联系的配偶之间的差异反映了军人推荐的特征。
