Snowball is not the only method for sampling refugees and similar populations
Keith Sabin, WHO
23 May 2011
We commend Sulaiman-Hill and Thompson for their use of a mixed methods approach to research the resettlement experiences of Afghan and Kurdish refugees in New Zealand and Western Australia. However, we disagree with the authors’ presentation of snowball sampling as the “only feasible way of locating” hidden populations that do not have sampling frames. While we agree that “obtaining statistically representative samples of such socially invisible groups is known to be problematic,” we would like to point out that there are probability-based sampling methods that can be used to generate valid population-based estimates among hidden populations. Time-location sampling (TLS) and respondent driven sampling (RDS) have been used worldwide since 1994 for research among "hidden" populations, such as sex workers, men who have sex with men, injection drug users, artists, high-risk youth, migrants, and other populations that form social networks [1-4].
The success of snowball sampling in the Afghan and Kurdish refugee populations suggests that RDS is a feasible probability-based sampling method in this population. RDS is a chain-referral sampling method whereby peers recruit other peers from within their social networks. Features, such as enforcement of recruitment quotas through the use of uniquely coded coupons and analytical adjustments based on differential recruitment (homophily) and social network sizes, allow RDS to overcome many of the biases found in other chain-referral methods [5-6].
The feasibility of RDS in migrant and refugee populations is due to the key role of social networks in the migration and settlement process [7]. While RDS has not specifically been used to recruit Middle Eastern refugees in western settings, it has been used to recruit diverse migrant populations, such as Polish immigrants in Ireland [8]; rural-to-urban migrants in China [9]; and Southeast Asian [10] and Latino immigrants in the United States [11,12]. Additionally, it is being proposed as a primary sampling method to access migrant workers in South and Southeast Asia [13]. While these populations may not be culturally similar to Afghan and Kurdish refugees in New Zealand and Western Australia, they do share similar social circumstances that present comparable methodological challenges to researchers and would be suitable for sampling via RDS.
References 1. Malekinejad M, Johnston LG, Kendall C, Kerr LR, Rifkin MR, Rutherford GW: Using respondent-driven sampling methodology for HIV biological and behavioral surveillance in international settings: a systematic review. [Review] AIDS & Behavior 2008, 12S(4 Suppl): S105-S130. 2. Heckathorn DD, Jeffri J: Finding the beat: Using respondent-driven sampling to study jazz musicians. Poetics 2001, 28:307-329. 3. Johnston LG, Thurman TR, Mock N, Nano L, Carcani V: Respondent-driven sampling: A new method for studying street children with findings from Albania. Vulnerable Children and Youth Studies 2010, 5(1): 1-11. 4. Kissinger P et al.: HIV/STI Risk behaviors among Latino migrant workers in New Orleans post-Hurricane Katrina disaster. Sexually Transmitted Diseases 2008, 35(11): 924-929. 5. Heckathorn DD. Respondent-driven sampling: A new approach to the study of hidden populations. Social Problems 1997, 44(2): 174-199. 6. Heckathorn D. Respondent driven sampling II: deriving valid population estimates from Chain-referral samples of hidden populations. Social Problems 2002; 49(1):11-34. 7. MacDonald JS, MacDonald LD: Chain Migration, Ethnic Neighborhood Formation, and Social Networks. Milbank Memorial Fund Quarterly 1964, 42(1): 82-97. 8. Muhlau P: Respondent driven sampling of migrant populations. [http://www.methods.manchester.ac.uk/events/2010-09-10/muhlau.pdf]. 9. He N, Wong FY, Huang ZJ, Ding Y, Fu C, Smith BD, et al: HIV risks among two types of male migrants in Shanghai, China: money boys vs. general male migrants. AIDS 2007, 21(Suppl 8): S73-9. 10. Shah AM, Guo L, Magee M, Cheung W, Simon M, LaBreche A, et al: Comparing selected measures of health outcomes and health-seeking behaviors in Chinese, Cambodian, and Vietnamese communities of Chicago: results from local health surveys. Journal of Urban Health 2010, 87(5): 813-826. 11. Strathdee SA, Lozada R, Ojeda VD, Pollini RA, Brouwer KC, Vera A, et al: Differential effects of migration and deportation on HIV infection among male and female injection drug users in Tijuana, Mexico. PLoS ONE 2008, 3(7): e2690. 12. Magis-Rodriguez C, Lemp G, Hernandez MT, Sanchez MA, Estrada F, Bravo-Garcia E: Going North: Mexican migrants and their vulnerability to HIV. Journal of the Acquired Immune Deficiency Syndrome 2009, 51(Suppl 1): S21-5. 13. Joint United Nations Initiative on Mobility and HIV/AIDS in South East Asia (JUNIMA) [http://junima.org].
Authors: Jane R Montealegre, Lisa G Johnston, Keith Sabin
Affiliations: Jane R Montealegre, The University of Texas School of Public Health, USA Lisa G Johnston, University of California, San Francisco, Global Health Science, USA Keith Sabin, World Health Organization, Hanoi, Vietnam
Snowball is not the only method for sampling refugees and similar populations
23 May 2011
We commend Sulaiman-Hill and Thompson for their use of a mixed methods approach to research the resettlement experiences of Afghan and Kurdish refugees in New Zealand and Western Australia. However, we disagree with the authors’ presentation of snowball sampling as the “only feasible way of locating” hidden populations that do not have sampling frames. While we agree that “obtaining statistically representative samples of such socially invisible groups is known to be problematic,” we would like to point out that there are probability-based sampling methods that can be used to generate valid population-based estimates among hidden populations. Time-location sampling (TLS) and respondent driven sampling (RDS) have been used worldwide since 1994 for research among "hidden" populations, such as sex workers, men who have sex with men, injection drug users, artists, high-risk youth, migrants, and other populations that form social networks [1-4].
The success of snowball sampling in the Afghan and Kurdish refugee populations suggests that RDS is a feasible probability-based sampling method in this population. RDS is a chain-referral sampling method whereby peers recruit other peers from within their social networks. Features, such as enforcement of recruitment quotas through the use of uniquely coded coupons and analytical adjustments based on differential recruitment (homophily) and social network sizes, allow RDS to overcome many of the biases found in other chain-referral methods [5-6].
The feasibility of RDS in migrant and refugee populations is due to the key role of social networks in the migration and settlement process [7]. While RDS has not specifically been used to recruit Middle Eastern refugees in western settings, it has been used to recruit diverse migrant populations, such as Polish immigrants in Ireland [8]; rural-to-urban migrants in China [9]; and Southeast Asian [10] and Latino immigrants in the United States [11,12]. Additionally, it is being proposed as a primary sampling method to access migrant workers in South and Southeast Asia [13]. While these populations may not be culturally similar to Afghan and Kurdish refugees in New Zealand and Western Australia, they do share similar social circumstances that present comparable methodological challenges to researchers and would be suitable for sampling via RDS.
References
1. Malekinejad M, Johnston LG, Kendall C, Kerr LR, Rifkin MR, Rutherford GW: Using respondent-driven sampling methodology for HIV biological and behavioral surveillance in international settings: a systematic review. [Review] AIDS & Behavior 2008, 12S(4 Suppl): S105-S130.
2. Heckathorn DD, Jeffri J: Finding the beat: Using respondent-driven sampling to study jazz musicians. Poetics 2001, 28:307-329.
3. Johnston LG, Thurman TR, Mock N, Nano L, Carcani V: Respondent-driven sampling: A new method for studying street children with findings from Albania. Vulnerable Children and Youth Studies 2010, 5(1): 1-11.
4. Kissinger P et al.: HIV/STI Risk behaviors among Latino migrant workers in New Orleans post-Hurricane Katrina disaster. Sexually Transmitted Diseases 2008, 35(11): 924-929.
5. Heckathorn DD. Respondent-driven sampling: A new approach to the study of hidden populations. Social Problems 1997, 44(2): 174-199.
6. Heckathorn D. Respondent driven sampling II: deriving valid population estimates from Chain-referral samples of hidden populations. Social Problems 2002; 49(1):11-34.
7. MacDonald JS, MacDonald LD: Chain Migration, Ethnic Neighborhood Formation, and Social Networks. Milbank Memorial Fund Quarterly 1964, 42(1): 82-97.
8. Muhlau P: Respondent driven sampling of migrant populations. [http://www.methods.manchester.ac.uk/events/2010-09-10/muhlau.pdf].
9. He N, Wong FY, Huang ZJ, Ding Y, Fu C, Smith BD, et al: HIV risks among two types of male migrants in Shanghai, China: money boys vs. general male migrants. AIDS 2007, 21(Suppl 8): S73-9.
10. Shah AM, Guo L, Magee M, Cheung W, Simon M, LaBreche A, et al: Comparing selected measures of health outcomes and health-seeking behaviors in Chinese, Cambodian, and Vietnamese communities of Chicago: results from local health surveys. Journal of Urban Health 2010, 87(5): 813-826.
11. Strathdee SA, Lozada R, Ojeda VD, Pollini RA, Brouwer KC, Vera A, et al: Differential effects of migration and deportation on HIV infection among male and female injection drug users in Tijuana, Mexico. PLoS ONE 2008, 3(7): e2690.
12. Magis-Rodriguez C, Lemp G, Hernandez MT, Sanchez MA, Estrada F, Bravo-Garcia E: Going North: Mexican migrants and their vulnerability to HIV. Journal of the Acquired Immune Deficiency Syndrome 2009, 51(Suppl 1): S21-5.
13. Joint United Nations Initiative on Mobility and HIV/AIDS in South East Asia (JUNIMA) [http://junima.org].
Authors: Jane R Montealegre, Lisa G Johnston, Keith Sabin
Affiliations:
Jane R Montealegre, The University of Texas School of Public Health, USA
Lisa G Johnston, University of California, San Francisco, Global Health Science, USA
Keith Sabin, World Health Organization, Hanoi, Vietnam
Competing interests
We have no competing interests