Abstract

Introduction

Researchers are constantly challenged to connect with participants in new and innovative ways. While traditional in-person recruitment has been successful in behavioral medicine, finding ways to connect with people outside of one’s immediate network can improve the generalizability of results to individuals who may not present in clinic. In particular, the COVID-19 pandemic drove researchers who may have previously relied on in-person recruitment to the internet to connect with potential participants.

There are many benefits to online recruitment for health-related research including cost-effectiveness (Ryan, 2013), the potential for a broader reach to individuals who may have less contact with their healthcare system (Benedict et al., 2019), and as a platform to reach historically hard-to-reach populations (Wozney et al., 2019). Social media recruitment like Facebook may reach more diverse samples—including age, race, and educational status—than traditional recruitment methods in large hospitals and academic centers. Despite these benefits, leveraging social media for recruitment in health research has lagged in the medical setting. A scoping review of 30 studies which used social media recruitment for medical research studies showed that almost half of included studies (n = 12) found social media to be more effective than other recruitment methods (Topolovec-Vranic and Natarajan, 2016). However, researchers may be hesitant to use social media for recruitment due to lack of knowledge of the efficiency of this recruitment strategy.

Online recruitment methodology has been used in various contexts. For example, recruitment utilizing Facebook has been used to reach samples regarding smoking cessation (Frandsen et al., 2014; Heffner et al., 2013; Sadasivam et al., 2013), condom use (Bull et al., 2012), diabetes (Balfe et al., 2012; Barnard et al., 2015), and cancer screenings (Kapp et al., 2013; Miyagi et al., 2014). While Facebook recruitment has often been used to target adolescents (Amon et al., 2014), more than half of adults over age 65 who use the internet are active Facebook users (Auxier and Anderson, 2021; Duggan et al., 2015). Little research utilizing targeted advertising on Facebook targets the older adult populations, representing a potential missed opportunities for researchers focused on adults and older adults (Thornton et al., 2016). More research is needed that targets older adults using Facebook to assess whether this is a viable recruitment strategy.

In the context of lung cancer screening, utilizing the Facebook platform to reach and target current and former smokers has proven efficacious. Lung cancer is one of the most stigmatized cancers and individuals may be less likely to present to care face-to-face in a timely manner (Carter-Harris et al., 2014). Facebook can provide a new way to reach individuals who are current or former smokers and has successfully reached this population in previous work (Carter-Harris et al., 2016; Strong and Renaud, 2020). However, research that utilizes targeted advertising using Facebook still represents just a sliver of the overall research in lung cancer screening and in the broader umbrella of health research.

Drawing attention to the strengths of social media recruitment for behavioral health research can help researchers leverage such platforms appropriately and efficiently. This paper describes recruitment for a study of current and former smokers focused on lung cancer screening. The aim of the current paper is to describe steps taken to recruit participants and to provide metrics about recruitment, including advertisement reach and participant demographics.

Methods

Results

Facebook advertisement reach

The ad was active for 34 days, with a brief pause to fix a problem with the study link. 46,153 unique Facebook users saw the ad on their web-page and 29,187 unique Facebook users clicked on the recruitment ad. Of those individuals, 2,111 clicked on the actual survey link which brought them to the REDCap introductory page. The total cost of the advertisement campaign was $600, with each click on the survey link costing approximately $0.28. A total of 433 entries were collected through REDCap. Out of those entries, 151 were excluded for various reasons (Figure 1). These reasons include failing the data integrity check, previous participation in the study, incomplete screening information including missing email, below age eligibility, not agreeing to participate, and never having smoked cigarettes. Two hundred eighty-two participants were randomized to either the control or intervention group. In total, 191 participants completed study questionnaires. The cost of the advertisement per completing participant was $3.14.

Open in a separate window

Figure 1.

Flow diagram.

Participant demographics

Demographic information is described in Table 1. Participants primarily identified as White (n = 168, 93.3%) and few identified as Black or African American (n = 6, 3.3%), American Indian or Alaska Native (n = 5, 2.8%), or Asian/Pacific Islander (n = 1, 0.6%). Participant age ranged from 55 to 77 years old (M = 62.9, SD = 5.1). Participants educational status ranged, with most individuals completing high school or equivalent diploma (n = 39, 21.7%), completing partial college (n = 73, 40.6%), standard college (n = 42, 23.3%), or a graduate degree (n = 24, 13.3%). Most participants were retired (n = 71, 39.7%) or employed full-time (n = 34, 19.0%) or disabled (n = 34, 19.0%). The majority of individuals reported being married/living with a partner (n = 80, 45.4%) or divorced/separated (n = 58, 33.0%). Over half of participants reported their annual income status between $10,000 and $40,000 (n = 95, 52.8%), and one-quarter reported their annual income status to be between $40,000 and $75,000 (n = 45, 25.0%).

Table 1.

Demographic characteristics of participants.

Demographic characteristic	N (%)
Gender
Female	148 (82.2)
Male	32 (17.8)
Age
55–64	120 (67.1)
65–74	55 (30.7)
>75	4 (2.2)
Race
American Indian or Alaska Native	5 (2.8)
Asian/Pacific Islander	1 (0.6)
Black or African American	6 (3.3)
White	168 (93.3)
Ethnicity
Hispanic	176 (98.9)
Non-Hispanic	2 (1.1)
Marital Status
Divorced/Separated	58 (33.0)
Married/Living with partner	80 (45.4)
Single	19 (10.8)
Widowed	19 (10.8)
Educational Status
Partial high school	2 (1.1)
High school graduate or GED	39 (21.7)
Partial college	73 (40.6)
Standard college or university	42 (23.3)
Graduate degree or professional training	24 (13.3)
Employment status
Employed (full time)	34 (19.0)
Employed (part time)	16 (8.9)
Disabled	34 (19.0)
Homemaker	10 (5.6)
Retired	71 (39.7)
Unemployed	12 (6.7)
Other	2 (1.1)
Income status
Less than $10,000	14 (7.7)
$10,000–$40,000	95 (52.8)
$40,000–$75,000	45 (25.0)
$75,000–$150,000	21 (11.7)
$150,000 or more	5 (2.8)

Open in a separate window

Percent values represent valid percents.

Discussion

This paper outlines steps taken in recruitment using targeted advertising on Facebook and displays metrics about ad reach with the aim of providing transparency. We successfully targeted and acquired data from individuals using Facebook targeted advertisement as our recruitment platform. Attention and data integrity checkpoints helped ensure the integrity of the final dataset.

The single ad generated 29,187 clicks from unique Facebook users over 34 days and 2,111 individuals clicked on the ad and were led to the introductory page. Ramo et al. (2014) tested 36 unique ads over 7 weeks manipulating the ad image (i.e. logo, cartoon, photo) and placement on Facebook (i.e. newsfeed, sponsored stories, promoted posts) with the aim to quantify the success of each distinct ad. Their total results culminated in 5,895 clicks between all ads and found that ads using a logo were the most successful. The success of the current study in generating ad clicks and responses suggests the growing viability of Facebook to recruit individuals for health-related research. Additionally, our randomized intervention design supports the feasibility of online recruitment for intervention studies in behavioral health which have previously proven difficult amongst researchers (Topolovec-Vranic and Natarajan, 2016).

Results from this study show that online recruitment using targeted advertising on Facebook is a successful platform for recruiting individuals with a smoking history and further provides support for recruitment via social media for health-related research. The advertisement reached its target audience and 20.5% of the clicks on the study link resulted in data entries, with almost two-thirds (65.1%) of entries resulting in randomized participants. Almost two-thirds (64.2%) of randomized participants completed the study surveys with valid data (i.e. did not fail the data integrity check). Findings show that utilizing a checking question to identify “bots” successfully identified a considerable number (n = 61) of responses which may have been fraudulent. Further, utilizing an attention check to ensure participant engagement with study information identified 10 participants whose data could be excluded to improve data integrity.

Overall, it cost $0.28 per click and $3.14 per completed participant. In a systematic review of recruitment using Facebook for health and psychosocial research, the mean cost per completing participant was $17.48 (range: 1.36, $110) (Thornton et al., 2016). Our mean ad cost falls within the lower end of the range of ad cost, suggesting that our ad efficiently reached individuals who were interested in participation. Utilizing targeted advertising on Facebook can be a cost-effective way to recruit individuals, and specifically targeting your sample as was done in this study can help improve efficiency.

This study is not without limitations. The sample was primarily White and does not mirror the racial representation of Facebook—about 74% and 72% of all Black and Hispanic U.S. adults, respectively, report utilizing Facebook across all age categories (Auxier and Anderson, 2021). Additionally, across all age groups about 14.4% of all smokers identify as Black, about 8.0% identify as Asian, about 27.1% identify as American Indian or Alaska Native, and about 8.0% identify as Hispanic (Centers for Disease Control and Prevention, 2022). The current study did not capture the racial representation of tobacco users and calls for increased research on recruiting adult and older adult tobacco users on Facebook.

Implications for practice and research

These findings join the growing body of literature recruiting individuals for health-related research using social media (Carter-Harris et al., 2016; Naslund et al., 2017; Sedrak et al., 2020). Additionally, the successful recruitment of individuals leveraging a social media platform like Facebook provides further evidence that this is a viable method of recruitment, particularly in the adult and older adult population. The inclusion of survey questions to ensure records are, first, completed by legitimate individuals and, second, completed by participants who are accurately attending to the questionnaires, increases the reliability and validity of the data. The transparency about the ad and its successful reach and retention of participants can aid not only researchers interested in leveraging social media to reach potential participants, but also for practitioners who may want to find new ways to access clients for health-related interventions.

Supplementary Material

Funding

Footnotes

References