Internet and Social Network Recruitment: Two Case Studies Kathy A. Johnson1, MS, RN; Jane Peace2, PhD, RN 1 University of Wisconsin – Madison, Madison, WI, US; 2Duke University, Durham, NC, US Abstract The recruitment of study participants is a significant research challenge. The Internet, with its ability to reach large numbers of people in networks connected by email, Facebook and other social networking mechanisms, appears to offer new avenues for recruitment. This paper reports recruitment experiences from two research projects that engaged the Internet and social networks in different ways for study recruitment. Drawing from the non-Internet recruitment literature, we speculate that the relationship with the source of the research and the purpose of the engaged social network should be a consideration in Internet or social network recruitment strategies. Background Recruitment for research involving human subjects is challenging. Efforts to increase study participation, particularly in minority and vulnerable populations, has led to research seeking to understand reasons why people choose to participate. These empirically identified reasons loosely cluster into four overlapping areas: 1) the perceived burden and benefit of participation1-3, 2) the personal health state of the potential participant4-5, 3) the type of research4,6, and 4) the relationship with the source of the research, that is, the level of trust and commitment the potential participant has in the research or in the recruiter 4,7,8. The Internet presents itself as a convenient and cost effective way for researchers to access a large sampling frame of potential participants. Internet recruitment strategies include but are not limited to mass emails within institutions, banners on search engines, Craigslist advertisements, postings on health-related discussion boards, and researchers posing as members engaging in discussions in chat rooms. However, because the type of Internet recruitment strategy, the type of research, and the study populations are diverse, making generalizations about the effectiveness of the Internet as a recruitment tool is difficult. For instance, Im and colleagues, seeking ethnicspecific cancer support groups, report difficulties identifying potential user-groups and gatekeeping by webmasters9. Seeking healthy young adults to complete a survey about smoking, Koo and Skinner report low recruitment using both mass emails and invitation messages on Web discussion boards10. In contrast, Bull and colleagues recruited (and retained) young adults for an online HIV-prevention intervention using banner advertisement11. And, seeking to recruit Hispanic men-seeking-men for HIV research, Fernandez and colleagues report good recruitment using a targeted interactive approach within chat rooms12. Because the Internet facilitates clustering of people around some variable or attribute, chain referral (snowball) sampling on the Web appears a reasonable way to access a targeted group. A snowball sampling strategy begins by asking purposefully identified participants to identify others as possible participants13. There is an underlying assumption that the purposefully identified participants are connected to others through a social network based on an identified attribute or variable. A social network is a set of people with some pattern of contacts or interactions between them14. For example, “PatientsLikeMe” is a health-related Internet-based social network that connects people who share a health condition for purposes of sharing information and experiences, and knowledge building among group members. Combining the Internet and social network snowball sampling would appear to be a reasonable and efficient method of recruitment, but the role of social networks in study participation remains unclear. We present two case studies that demonstrate two different uses of the Internet to tap social networks for research recruitment. The first study involved the intentional use of the researcher’s personal social network via email, building on the trust of existing relationships to recruit a “hard to reach” population: those who had experienced the death of a family member from a geographic distance. In this case, social networks were tapped to help reach a convenience sample. The second study was designed to recruit from a local university-based volunteer research registry (purposeful sampling), but serendipitously engaged an existing social network. Both of the case studies made the assumption that “some kind of connection”-- a relationship with the source of the research-- would encourage participation and thus increase recruitment. The purpose of this paper is to report and contrast these two case studies that used different methods of Internet and social network recruitment.
Two case studies using the Internet and social networks for recruitment Study 1: From a distance: The experience of geographically distant family members at a relative’s end of life phase The purpose of this study was to understand the experiences of family members who live at a distance from a relative who is dying. Clinicians identify family dynamics (difficulty designating a decision maker, reaching consensus among family members, or navigating family tensions) as one of multiple barriers to end of life conversations. In the future, a computer-mediated communication tool could facilitate these discussions. To effectively design for these technologies, we must first understand the experience of distant family members during the dying phase of a relative. An online open-ended questionnaire was designed to assess this. Accessing this “distant” population posed a challenge, as family members and their roles (except for Power of Attorney) are not officially documented in the clinical record. Therefore, one of the study purposes was to determine the feasibility of recruitment through a sampling procedure that capitalized on the social networks of the researcher and the first generation respondents (FGRs). Unlike traditional snowball or chain referral sampling plans, the FGRs were not identified by the shared attribute (the death of a geographically distant relative), but instead were identified by their relationship with the researcher, and were used to access a large geographic and social range of people for the purpose of finding people with the desired characteristic. This recruitment strategy was employed to minimize any coercion for the FGR, and to increase the social distance between the researcher and potential respondent, providing increased confidentiality or anonymity for respondents who would be sharing potentially emotional responses. No incentives were offered to either the FGR or the SGR (second generation respondents). An invitational email message was sent to 80 FGRs identified from the researcher’s email contact list (professional and personal contacts, no family members), based on the perceived likelihood that they would forward the message. The invitational email messages were extended to 59 women and 21 men in 4 countries. (Figure 1). The FGRs were asked to forward the email message (with included hyperlink to the questionnaire) to potential SGRs, personal contacts who they felt might meet the study inclusion criteria: 18 years old or older; experienced the death of a legal or blood relative within the last 6 months; live more than 50 miles away from the relative; had reason to talk to a physician or other health care provider during that relative’s dying phase. No attempt was made to track either the disposition of the original invitational email or the number of times the email message was forwarded to potential SGRs. Figure 1. Study 1 Planned Recruitment Design The results were disappointing. Personal email responses were received from 4 FGRs apologizing that they did not know anyone who met the inclusion criteria. Five (second generation) respondents started the survey, but only three completed it. Of these three, demographic information was completed for two: one from Wisconsin (the originating state of the research) and the other from Colorado. No new data was received after 4 days of the initial email invitation. No reminder emails were sent to the FGRs. Though the literature is sparse about this type of recruitment strategy, this result was not as successful as Rock’s attempt to access another “hard to reach” population. Seeking pet owners who had been treating cats or dogs for diabetes, Rock found that chain referral sampling via email was more efficient and effective at yielding a small but diverse group of respondents than veterinarian referral, letters posted in pet-friendly places and word of mouth15. One difference between Rock’s study and this study is that FGRs in Rock’s study could participate in the research. In Study 1, the attempt to distance the researcher from the respondent by recruiting SGRs but not FGRs may have discouraged participation. Other possible reasons for the low response rate include the small sampling pool and inclusion criteria that were overly limiting (occurrence within the last 6 months was chosen to access only recent
experiences). Additionally, an online survey may have seemed too impersonal for this potentially emotional topic, or people who did experience this may not have wanted to re-live the experience by writing about it. Study 2: Unanticipated Internet social network recruiting: Experiences of Adoptive Parents of Children with FXS The purpose of this study was to learn about the experiences and perceptions of adoptive parents of children with Fragile X Syndrome (FXS), the most common genetically mediated cause of mental retardation. Caused by a genetic mutation on the X chromosome, FXS affects males more often and more severely than females. Due to increased penetrance in subsequent generations, the FXS trait is usually present in families for several generations before an individual is sufficiently affected to warrant evaluation and diagnosis. For this reason, children with FXS may be placed for adoption without anticipation of the diagnosis. This telephone interview study explored adoptive parents’ experiences obtaining their children’s FXS diagnosis and sharing information about the diagnosis with birth families. Understanding the experiences of this population may inform the design of informatics applications to support managing and sharing health information for such families. The FXS community is actively involved in research, and research registries have been developed to facilitate and manage recruitment of this population. These research registries confidentially manage limited information about potential research participants who volunteer to belong to the registry. For this study, a university-based FXS research registry was accessed. The registry included nine adoptive families of children with FXS. None of the nine families were currently participating in research, and in fact, due to their adoptive status, these families were often excluded from research despite their eagerness to participate. A traditional recruitment strategy was used, in which a letter of invitation was sent via postal mail from the research registry to the nine eligible families, the FGRs. Six of the nine families responded, and five consented to participate. This recruitment took place over a six-week period. One FGR posted information about the study on a Facebook page for parents of children with FXS, initiating unanticipated Internet recruitment for the study. (Figure 2). Three adoptive parents who had learned about the study on Facebook contacted the researcher via telephone and requested information and consent materials. Although all three of these SGRs initially expressed enthusiasm about participating, only one SGR signed and returned a consent form. Recruiting this SGR required a phone call and a reminder email message from the researcher, and the signed consent form was received almost twelve weeks after recruitment to the study began. Therefore, recruiting this subject required a disproportionate amount of researcher time and effort. Figure 2. Study 2 Serendipitous Recruitment Design Recruitment via the research registry was considerably more successful than recruitment via the Facebook group (Table 1). The registry could be considered a type of network, Research Facebook in that members share a specific attribute (parenting a child registry with FXS), as well as a willingness to participate in research. # Eligible 9 Unknown The unanticipated extension of recruitment to SGRs in this # Responded to 6 3 study provided an opportunity to compare recruitment via the invitation registry with recruitment in a social network based on the # Consented 6 1 same attribute. The recruitment rate in the social network was much lower, and may have been still lower than it appears, Recruitment rate 67% 33% because the network may have included eligible persons who # Completed study 5 0 did not respond to the invitation to inquire about the study. Table 1. Study 2 Recruitment Results
Discussion These case studies are only two of a myriad of ways to incorporate the Internet and social networks into recruitment strategies. However, previously identified barriers to Internet recruitment (e.g., difficulty identifying user-groups, gatekeeping by webmasters) do not appear to apply in these studies. In order to understand the added aspect of social networks to our Internet recruitment efforts, we first contrast the case studies (Table 2), and then reflect on the motivations for study participation in reported in the general recruitment literature. Study 1 - From a Distance
Study 2A – Families/FXS Volunteer Registry
Study 2B – after FGR posting on Facebook
Method of Internet recruitment
Email-based
None
(Unanticipated) Web-based
Geographic range of recruitment
Worldwide (diversity)
National (U.S.)
Unknown; potentially worldwide
Whose social network accessed
Researcher’s and FGRs’
Volunteer Research Registry
FGR’s
Type of social networks
Personal and professional friends
Health-related (FXS) group with research interest
Health-related (FXS) group formed for social support
1
1
(FGR)
(Research Registry)
Number of intermediaries between researcher and participant
2 (Research registry and FGR)
Table 2. Case comparison of use of Internet and social networks Although social networks can be characterized in other ways, the social networks tapped in these two cases are described as either personal or health-related. A critical review of these case studies allows us to compare recruitment between types of social networks that each had one intermediary (Study 1 and Study 2A). In these studies recruitment was more successful within social networks that were health-related than in social networks that were personal. In Study 1, the several FGRs (those with a direct relationship to the researcher) demonstrated investment in the relationship with the researcher by replying to the researcher’s email message. That is, the relationship between the FGR and the researcher appeared to require a response of some kind, even though none was expected. Continuing this line of thinking, we speculate that the SGRs may have similarly demonstrated investment in their relationships with FGRs. However, the SGRs did not have a relationship with the researcher, and therefore had no overt reason to be invested in the research. In Study 2A, participation in the volunteer research registry likely demonstrated an ongoing commitment to research, whereas participation in the Facebook group (Study 2B) likely demonstrated commitment to the Facebook group, not to research—a social tie. In the presented case studies, the more direct relationship with the researcher or with research resulted in more participation. Additionally, given the same number of intermediaries, recruitment from a group with membership based on research was more successful. The relationship to the source of the research, an identified factor of research participation in the general recruitment literature, refers to the level of trust and commitment the potential participant has in the research or in the recruiter. Lowton describes trust in the clinic and its staff as one of the factors that motivate cystic fibrosis patients to participate in clinical trials4. And, in the case of childhood cancer tumor tissue banking, participation in research has been described as being part of the childhood cancer community7. Frost and colleagues report on an instance of grassroots research that occurred within PatientsLikeMe16. Their study describes how the news of a possible effect of lithium on Amyotophic Lateral Sclerosis was developed and diffused through an online community. The patientdriven observational study of the use of lithium that Frost and colleagues describe was conducted by and for the PatientsLikeMe group members. In this instance, research recruitment and participation were facilitated by the presence of the Internet and a trusted health-related social network that is committed to research.
In summary, we speculate that the relationship with the source of the research, identified in non-Internet recruitment literature, is a consideration in Internet and social network recruitment as well, particularly the relationship with either the research or the researcher, and the purpose of the tapped social network. Conclusion Indeed, the small sample sizes of these studies make it impossible to draw conclusions. However, there seems to be an indication that the relationship to the source of the research, an identified factor in non-Internet recruitment, may also be a motivating factor for Internet research participation. The influence of social networks as a trusted source or connection with a researcher appears especially important for recruitment via the Internet, and warrants further investigation. Finally, Study 1 sought cultural and experiential variety by sending invitations worldwide, using geographic distance as a proxy for cultural diversity. Though the Internet has the potential to reach geographically distant populations, we speculate that cultural factors influence motivation to participate and are not yet well understood.
References 1. Cadigan, RJ and Davis, AM. Deciding whether to participate in a biobank: The concerns of healthy volunteers. Principles and Practice in Biobank Governance. J. Kaye and M. Stranger. Surrey, Ashgate: 2009; 117-136. 2. Wasan AD, Taubenberger SP, Robinson WM. Reasons for participation in pain research: Can they indicate a lack of informed consent? Pain Medicine 2009;10(1):111-9. 3. Peel E, Parry O, Douglas M, Lawton J. "It's no skin off my nose": Why people take part in qualitative research. Qual Health Res 2006, Dec;16(10):1335-49. 4. Lowton K. Trials and tribulations: Understanding motivations for clinical research participation amongst adults with cystic fibrosis. Social Science & Medicine 2005;61(8):1854-65. 5. Kossman S, Hsieh Y, Peace J, Valdez R, Severtson L, Burke L, Brennan PF. A theory-based problem-solving approach to recruitment challenges in a large randomized field trial. Appl Nurs Res 2011, Aug;24(3):179-87. 6. Hallowell N, Cooke S, Crawford G, Lucassen A, Parker M, Snowdon C. An investigation of patients' motivations for their participation in genetics-related research. J Med Ethics 2010, Jan;36(1):37-45. 7. Dixon-Woods M, Wilson D, Jackson C, Cavers D, Pritchard-Jones K. Human tissue and ‘the public’: The case of childhood cancer tumour banking. Biosocieties 2008;3(1):57-80. 8. Bailey JM, Bieniasz ME, Kmak D, Brenner DE, Ruffin MT. Recruitment and retention of economically underserved women to a cervical cancer prevention trial. Applied Nursing Research 2004;17(1):55-60. 9. Im EO, Chee W, Tsai HM, Bender M, Lim HJ. Internet communities for recruitment of cancer patients into an internet survey: A discussion paper. Int J Nurs Stud 2007, Sep;44(7):1261-9. 10. Koo M, Skinner H. Challenges of internet recruitment: A case study with disappointing results. J Med Internet Res 2005;7(1):e6. 11. Bull SS, Vallejos D, Levine D, Ortiz C. Improving recruitment and retention for an online randomized controlled trial: Experience from the youthnet study. AIDS Care 2008, Sep;20(8):887-93. 12. Fernandez, MI, Varga LM, Perrino T, Collazo JB, Subiaul F, Rehbein A, et al. The internet as recruitment tool for HIV studies: Viable strategy for reaching at-risk hispanic MSM in Miami? AIDS Care 2004; (8):953-63. 13. Patton MQ. Qualitative research and evaluation methods . Thousand Oaks, Calif. : Sage Publications,; 2002. 14. Freeman LC. The development of social network analysis. Empirical Press Vancouver, BC; 2004. 15. Rock MJ. Harnessing social networks along with consumer-driven electronic communication technologies to identify and engage members of 'hard-to-reach' populations: A methodological case report. BMC Med Res Methodol 2010;10:8. 16. Frost JH, Massagli MP, Wicks P, Heywood J. How the social web supports patient experimentation with a new therapy: The demand for patient-controlled and patient-centered informatics. AMIA Annu Symp Proc 2008:217-21.
Two case studies using the Internet and social networks for recruitment Study 1: From a distance: The experience of geographically distant family members at a relative’s end of life phase The purpose of this study was to understand the experiences of family members who live at a distance from a relative who is dying. Clinicians identify family dynamics (difficulty designating a decision maker, reaching consensus among family members, or navigating family tensions) as one of multiple barriers to end of life conversations. In the future, a computer-mediated communication tool could facilitate these discussions. To effectively design for these technologies, we must first understand the experience of distant family members during the dying phase of a relative. An online open-ended questionnaire was designed to assess this. Accessing this “distant” population posed a challenge, as family members and their roles (except for Power of Attorney) are not officially documented in the clinical record. Therefore, one of the study purposes was to determine the feasibility of recruitment through a sampling procedure that capitalized on the social networks of the researcher and the first generation respondents (FGRs). Unlike traditional snowball or chain referral sampling plans, the FGRs were not identified by the shared attribute (the death of a geographically distant relative), but instead were identified by their relationship with the researcher, and were used to access a large geographic and social range of people for the purpose of finding people with the desired characteristic. This recruitment strategy was employed to minimize any coercion for the FGR, and to increase the social distance between the researcher and potential respondent, providing increased confidentiality or anonymity for respondents who would be sharing potentially emotional responses. No incentives were offered to either the FGR or the SGR (second generation respondents). An invitational email message was sent to 80 FGRs identified from the researcher’s email contact list (professional and personal contacts, no family members), based on the perceived likelihood that they would forward the message. The invitational email messages were extended to 59 women and 21 men in 4 countries. (Figure 1). The FGRs were asked to forward the email message (with included hyperlink to the questionnaire) to potential SGRs, personal contacts who they felt might meet the study inclusion criteria: 18 years old or older; experienced the death of a legal or blood relative within the last 6 months; live more than 50 miles away from the relative; had reason to talk to a physician or other health care provider during that relative’s dying phase. No attempt was made to track either the disposition of the original invitational email or the number of times the email message was forwarded to potential SGRs. Figure 1. Study 1 Planned Recruitment Design The results were disappointing. Personal email responses were received from 4 FGRs apologizing that they did not know anyone who met the inclusion criteria. Five (second generation) respondents started the survey, but only three completed it. Of these three, demographic information was completed for two: one from Wisconsin (the originating state of the research) and the other from Colorado. No new data was received after 4 days of the initial email invitation. No reminder emails were sent to the FGRs. Though the literature is sparse about this type of recruitment strategy, this result was not as successful as Rock’s attempt to access another “hard to reach” population. Seeking pet owners who had been treating cats or dogs for diabetes, Rock found that chain referral sampling via email was more efficient and effective at yielding a small but diverse group of respondents than veterinarian referral, letters posted in pet-friendly places and word of mouth15. One difference between Rock’s study and this study is that FGRs in Rock’s study could participate in the research. In Study 1, the attempt to distance the researcher from the respondent by recruiting SGRs but not FGRs may have discouraged participation. Other possible reasons for the low response rate include the small sampling pool and inclusion criteria that were overly limiting (occurrence within the last 6 months was chosen to access only recent
experiences). Additionally, an online survey may have seemed too impersonal for this potentially emotional topic, or people who did experience this may not have wanted to re-live the experience by writing about it. Study 2: Unanticipated Internet social network recruiting: Experiences of Adoptive Parents of Children with FXS The purpose of this study was to learn about the experiences and perceptions of adoptive parents of children with Fragile X Syndrome (FXS), the most common genetically mediated cause of mental retardation. Caused by a genetic mutation on the X chromosome, FXS affects males more often and more severely than females. Due to increased penetrance in subsequent generations, the FXS trait is usually present in families for several generations before an individual is sufficiently affected to warrant evaluation and diagnosis. For this reason, children with FXS may be placed for adoption without anticipation of the diagnosis. This telephone interview study explored adoptive parents’ experiences obtaining their children’s FXS diagnosis and sharing information about the diagnosis with birth families. Understanding the experiences of this population may inform the design of informatics applications to support managing and sharing health information for such families. The FXS community is actively involved in research, and research registries have been developed to facilitate and manage recruitment of this population. These research registries confidentially manage limited information about potential research participants who volunteer to belong to the registry. For this study, a university-based FXS research registry was accessed. The registry included nine adoptive families of children with FXS. None of the nine families were currently participating in research, and in fact, due to their adoptive status, these families were often excluded from research despite their eagerness to participate. A traditional recruitment strategy was used, in which a letter of invitation was sent via postal mail from the research registry to the nine eligible families, the FGRs. Six of the nine families responded, and five consented to participate. This recruitment took place over a six-week period. One FGR posted information about the study on a Facebook page for parents of children with FXS, initiating unanticipated Internet recruitment for the study. (Figure 2). Three adoptive parents who had learned about the study on Facebook contacted the researcher via telephone and requested information and consent materials. Although all three of these SGRs initially expressed enthusiasm about participating, only one SGR signed and returned a consent form. Recruiting this SGR required a phone call and a reminder email message from the researcher, and the signed consent form was received almost twelve weeks after recruitment to the study began. Therefore, recruiting this subject required a disproportionate amount of researcher time and effort. Figure 2. Study 2 Serendipitous Recruitment Design Recruitment via the research registry was considerably more successful than recruitment via the Facebook group (Table 1). The registry could be considered a type of network, Research Facebook in that members share a specific attribute (parenting a child registry with FXS), as well as a willingness to participate in research. # Eligible 9 Unknown The unanticipated extension of recruitment to SGRs in this # Responded to 6 3 study provided an opportunity to compare recruitment via the invitation registry with recruitment in a social network based on the # Consented 6 1 same attribute. The recruitment rate in the social network was much lower, and may have been still lower than it appears, Recruitment rate 67% 33% because the network may have included eligible persons who # Completed study 5 0 did not respond to the invitation to inquire about the study. Table 1. Study 2 Recruitment Results
Discussion These case studies are only two of a myriad of ways to incorporate the Internet and social networks into recruitment strategies. However, previously identified barriers to Internet recruitment (e.g., difficulty identifying user-groups, gatekeeping by webmasters) do not appear to apply in these studies. In order to understand the added aspect of social networks to our Internet recruitment efforts, we first contrast the case studies (Table 2), and then reflect on the motivations for study participation in reported in the general recruitment literature. Study 1 - From a Distance
Study 2A – Families/FXS Volunteer Registry
Study 2B – after FGR posting on Facebook
Method of Internet recruitment
Email-based
None
(Unanticipated) Web-based
Geographic range of recruitment
Worldwide (diversity)
National (U.S.)
Unknown; potentially worldwide
Whose social network accessed
Researcher’s and FGRs’
Volunteer Research Registry
FGR’s
Type of social networks
Personal and professional friends
Health-related (FXS) group with research interest
Health-related (FXS) group formed for social support
1
1
(FGR)
(Research Registry)
Number of intermediaries between researcher and participant
2 (Research registry and FGR)
Table 2. Case comparison of use of Internet and social networks Although social networks can be characterized in other ways, the social networks tapped in these two cases are described as either personal or health-related. A critical review of these case studies allows us to compare recruitment between types of social networks that each had one intermediary (Study 1 and Study 2A). In these studies recruitment was more successful within social networks that were health-related than in social networks that were personal. In Study 1, the several FGRs (those with a direct relationship to the researcher) demonstrated investment in the relationship with the researcher by replying to the researcher’s email message. That is, the relationship between the FGR and the researcher appeared to require a response of some kind, even though none was expected. Continuing this line of thinking, we speculate that the SGRs may have similarly demonstrated investment in their relationships with FGRs. However, the SGRs did not have a relationship with the researcher, and therefore had no overt reason to be invested in the research. In Study 2A, participation in the volunteer research registry likely demonstrated an ongoing commitment to research, whereas participation in the Facebook group (Study 2B) likely demonstrated commitment to the Facebook group, not to research—a social tie. In the presented case studies, the more direct relationship with the researcher or with research resulted in more participation. Additionally, given the same number of intermediaries, recruitment from a group with membership based on research was more successful. The relationship to the source of the research, an identified factor of research participation in the general recruitment literature, refers to the level of trust and commitment the potential participant has in the research or in the recruiter. Lowton describes trust in the clinic and its staff as one of the factors that motivate cystic fibrosis patients to participate in clinical trials4. And, in the case of childhood cancer tumor tissue banking, participation in research has been described as being part of the childhood cancer community7. Frost and colleagues report on an instance of grassroots research that occurred within PatientsLikeMe16. Their study describes how the news of a possible effect of lithium on Amyotophic Lateral Sclerosis was developed and diffused through an online community. The patientdriven observational study of the use of lithium that Frost and colleagues describe was conducted by and for the PatientsLikeMe group members. In this instance, research recruitment and participation were facilitated by the presence of the Internet and a trusted health-related social network that is committed to research.
In summary, we speculate that the relationship with the source of the research, identified in non-Internet recruitment literature, is a consideration in Internet and social network recruitment as well, particularly the relationship with either the research or the researcher, and the purpose of the tapped social network. Conclusion Indeed, the small sample sizes of these studies make it impossible to draw conclusions. However, there seems to be an indication that the relationship to the source of the research, an identified factor in non-Internet recruitment, may also be a motivating factor for Internet research participation. The influence of social networks as a trusted source or connection with a researcher appears especially important for recruitment via the Internet, and warrants further investigation. Finally, Study 1 sought cultural and experiential variety by sending invitations worldwide, using geographic distance as a proxy for cultural diversity. Though the Internet has the potential to reach geographically distant populations, we speculate that cultural factors influence motivation to participate and are not yet well understood.
References 1. Cadigan, RJ and Davis, AM. Deciding whether to participate in a biobank: The concerns of healthy volunteers. Principles and Practice in Biobank Governance. J. Kaye and M. Stranger. Surrey, Ashgate: 2009; 117-136. 2. Wasan AD, Taubenberger SP, Robinson WM. Reasons for participation in pain research: Can they indicate a lack of informed consent? Pain Medicine 2009;10(1):111-9. 3. Peel E, Parry O, Douglas M, Lawton J. "It's no skin off my nose": Why people take part in qualitative research. Qual Health Res 2006, Dec;16(10):1335-49. 4. Lowton K. Trials and tribulations: Understanding motivations for clinical research participation amongst adults with cystic fibrosis. Social Science & Medicine 2005;61(8):1854-65. 5. Kossman S, Hsieh Y, Peace J, Valdez R, Severtson L, Burke L, Brennan PF. A theory-based problem-solving approach to recruitment challenges in a large randomized field trial. Appl Nurs Res 2011, Aug;24(3):179-87. 6. Hallowell N, Cooke S, Crawford G, Lucassen A, Parker M, Snowdon C. An investigation of patients' motivations for their participation in genetics-related research. J Med Ethics 2010, Jan;36(1):37-45. 7. Dixon-Woods M, Wilson D, Jackson C, Cavers D, Pritchard-Jones K. Human tissue and ‘the public’: The case of childhood cancer tumour banking. Biosocieties 2008;3(1):57-80. 8. Bailey JM, Bieniasz ME, Kmak D, Brenner DE, Ruffin MT. Recruitment and retention of economically underserved women to a cervical cancer prevention trial. Applied Nursing Research 2004;17(1):55-60. 9. Im EO, Chee W, Tsai HM, Bender M, Lim HJ. Internet communities for recruitment of cancer patients into an internet survey: A discussion paper. Int J Nurs Stud 2007, Sep;44(7):1261-9. 10. Koo M, Skinner H. Challenges of internet recruitment: A case study with disappointing results. J Med Internet Res 2005;7(1):e6. 11. Bull SS, Vallejos D, Levine D, Ortiz C. Improving recruitment and retention for an online randomized controlled trial: Experience from the youthnet study. AIDS Care 2008, Sep;20(8):887-93. 12. Fernandez, MI, Varga LM, Perrino T, Collazo JB, Subiaul F, Rehbein A, et al. The internet as recruitment tool for HIV studies: Viable strategy for reaching at-risk hispanic MSM in Miami? AIDS Care 2004; (8):953-63. 13. Patton MQ. Qualitative research and evaluation methods . Thousand Oaks, Calif. : Sage Publications,; 2002. 14. Freeman LC. The development of social network analysis. Empirical Press Vancouver, BC; 2004. 15. Rock MJ. Harnessing social networks along with consumer-driven electronic communication technologies to identify and engage members of 'hard-to-reach' populations: A methodological case report. BMC Med Res Methodol 2010;10:8. 16. Frost JH, Massagli MP, Wicks P, Heywood J. How the social web supports patient experimentation with a new therapy: The demand for patient-controlled and patient-centered informatics. AMIA Annu Symp Proc 2008:217-21.
