close× Call Us
close×
Title Recruiting Subjects for Online Surveys: Facebook versus Mechanical Turk
Post date 10/27/2015
C1 Background and Explanation of Rationale

Online surveys are increasingly popular in political science. While subjects can be recruited
through commercial Internet panels, scholars often rely on one of two lower-cost methods:
Amazon.com’s Mechanical Turk for research in the United States and India, and Facebook advertisements
when drawing subjects in other countries. Yet little work compares these various
methods, perhaps because they tend to be used by scholars in different subfields. In this paper
we outline a research design to compare recruitment via Facebook, Mechanical Turk, and a
Qualtrics Internet panel in India and the United States, the two countries where all three methods
are viable. We discuss existing research and advance hypotheses concerning demographics,
political attitudes and knowledge, participant cooperation, and experimental replication. We
outline a proposal to compare the various online sources to each other and to national probability
samples. This research will constitute the first large-N comparison of survey sampling
via Facebook and Mechanical Turk and the first to compare online convenience samples for
political science research in India and the U.S.

C2 What are the hypotheses to be tested?

Too many to list here; see pre-analysis plan.

C3 How will these hypotheses be tested? *

See pre-analysis plan.

C4 Country United States of America, India
C5 Scale (# of Units) 6000 (in each country, 1000 each recruited through Facebook, Mechanical Turk, and a Qualtrics panel)
C6 Was a power analysis conducted prior to data collection? No
C7 Has this research received Insitutional Review Board (IRB) or ethics committee approval? Yes
C8 IRB Number 3891X
C9 Date of IRB Approval 9/21/15; clarification approval 10/21/15
C10 Will the intervention be implemented by the researcher or a third party? Researchers
C11 Did any of the research team receive remuneration from the implementing agency for taking part in this research? No
C12 If relevant, is there an advance agreement with the implementation group that all results can be published? not provided by authors
C13 JEL Classification(s) D72, C83, C90