You are on page 1of 3

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/291162231

Snowball sampling

Article in The BMJ · December 2013


DOI: 10.1136/bmj.f7511

CITATIONS READS

58 21,071

1 author:

Philip M. Sedgwick
St George's, University of London
437 PUBLICATIONS 9,743 CITATIONS

SEE PROFILE

All content following this page was uploaded by Philip M. Sedgwick on 19 January 2016.

The user has requested enhancement of the downloaded file.


BMJ 2013;347:f7511 doi: 10.1136/bmj.f7511 (Published 20 December 2013) Page 1 of 2

Endgames

ENDGAMES

STATISTICAL QUESTION

Snowball sampling
Philip Sedgwick reader in medical statistics and medical education
Centre for Medical and Healthcare Education, St George’s, University of London, London, UK

Researchers examined the association between plasma HIV-1 same probability of being selected independently of all others.
RNA concentrations and the incidence of HIV in injecting drug The probability that a population member will be chosen is
users. A prospective cohort study design was used. Participants known in advance.
were injecting drug users, with or without HIV, followed up In the above study, snowball sampling initially involved
every six months between 1 May 1996 and 30 June 2007. Cohort identifying an easily accessible small number of injecting drug
members were recruited from an inner city community in users for recruitment to the cohort. These cohort members then
Vancouver, Canada, using the method of snowball sampling. recommended other injecting drug users as potential cohort
In total, 622 injecting drug users with HIV and 1429 injecting members, who in turn suggested others. The method is so called
drug users without HIV were recruited.1 because the number of sample members increases with time,
The researchers reported that in the community of injecting drug analogous to a snowball accumulating snow as it rolls down a
users, a longitudinal measure of plasma HIV-1 RNA hill. Snowball sampling is used to recruit samples when
concentration was positively correlated with the HIV incidence members of the desired population are hard to reach or access
rate, independently of unsafe sexual behaviours and sharing because they feel disempowered, socially excluded, or
used syringes. It was concluded the results could help inform vulnerable.
HIV prevention and treatment interventions. Snowball sampling was probably the only sampling method
Which of the following statements, if any, are true? that could have been used to obtain a reasonably representative
a) Snowball sampling constitutes probability sampling sample of injecting drug users. Unlike random sampling
methods, it did not require the construction of a sampling frame
b) Snowball sampling requires the construction of a sampling
(b is false). However, it would probably not have been possible
frame
to construct a list of injecting drug users in the community
c) Snowball sampling is prone to selection bias because they would have been hard to access. Even if they were
Answers identified, they might have shown a certain amount of distrust
and been reluctant to be recruited if approached at random. The
Statement c is true, whereas a and b are false. use of existing cohort members to recommend other potential
Two types of sampling methods can be used to recruit members probably encouraged recruitment through an element
participants to a study—random sampling (sometimes called of trust in the community. To increase recruitment and ensure
probability sampling) and non-random sampling (sometimes that loss to follow-up was minimised, cohort members were
called non-probability sampling) methods. Snowball sampling offered $C20 (£11.5; €13.7; $18.8) at each study visit.
is a type of non-random (non-probability) sampling method (a The cohort that resulted from snowball sampling was prone to
is false). selection bias (c is true). Selection bias, described in a previous
Random sampling methods involve some form of random question,2 would have occurred if the injecting drug users
selection of the members of the population. Each member of recruited to the cohort were systematically different from those
the population has a known and typically equal probability of who were not recruited—resulting in the cohort not being
being selected for the sample. The most straightforward type representative of the population. Recruitment of injecting drug
of random sampling method is simple random sampling users depended on the recommendations of existing cohort
(sometimes referred to as random sampling). Random sampling members. Therefore, the cohort was unlikely to be representative
requires knowledge of exactly who is in the population, with of the population owing to biases in the injecting drug users
construction of a sampling frame—that is, a list of everyone in who were recommended. For example, injecting drug users who
the population. A sample of a fixed size is selected at random did not have many friends were less likely to be recruited.
from the list, with all members of the population having the Typically, any sample resulting from non-probability sampling

p.sedgwick@sgul.ac.uk

For personal use only: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe
BMJ 2013;347:f7511 doi: 10.1136/bmj.f7511 (Published 20 December 2013) Page 2 of 2

ENDGAMES

methods is considered not to be representative of the population. 1 Wood E, Kerr T, Marshall BDL, Li K, Zhang R, Hogg RS, et al. Longitudinal community
plasma HIV-1 RNA concentrations and incidence of HIV-1 among injecting drug users:
Little is known about the properties of samples resulting from prospective cohort study. BMJ 2009;338:b1649.
snowball sampling. 2 Sedgwick P. Selection bias versus allocation bias. BMJ 2013;346:f3345.

Competing interests: None declared. Cite this as: BMJ 2013;347:f7511


© BMJ Publishing Group Ltd 2013

For personal use only: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe
View publication stats

You might also like