changes in russian nationalist public opinion 2013–14
<www.surveysystem.com/sscalc.htm#one> (last accessed 5 April
2015).
- Resulting in an estimated margin of sampling error of about 2.8
per cent with a 95 per cent confidence level, assuming responses
to a question are fairly equally split across response options. See
<www.surveysystem.com/sscalc.htm#one> (last accessed 5 April
2015).
- The 2005 survey, conducted by the Moscow- based Levada Centre
and commissioned by Mikhail A. Alexseev, involved 680 respond-
ents and had an estimated margin of sampling error just under
4 per cent at a 95 per cent confidence level. For a description
of survey methodology, data and results, see http://www.rohan.
sdsu.edu/~alexseev/migration_and_ethnic_conflict/data.html (last
accessed 5 April 2015).
- Excluded from the sample in 2013 and 2014 were zones deemed
to be in armed conflict (Chechnya and Ingushetia) and remote,
sparsely populated areas of northern Siberia and the Russian Far
East (Nenets, Khanty- Mansi, Yamal- Nenets, Kamchatka, Chukotka
and Sakhalin). Altogether these omitted regions account for less
than 4 per cent of the total population of the Russian Federation.
Further, in 2005, respondents were not sampled in the insurgency-
prone republics of Dagestan and North Ossetia. The non- response
rate in these surveys was between roughly three-quarters and four-
fifths – rather typical for industrially developed, predominantly
urban societies.
- The figures on presidential vote choice are calculated from ran-
domly selected sub- samples of respondents: 553 in 2013 and 549 in
2014. This is because for the other half of the sample the question
was differently worded, as part of a study on various factors that
may influence presidential voting (Hale 2014).
- To ensure that substantive results remain equitable over time, the
missing data (including responses of ‘hard to say’ and refusal to
answer) are excluded here and thereafter, unless otherwise stated.
In general, this makes little difference for the figures reported. For
ethnic pride, the missing data involved 3.5 per cent of respondents
in 2005, 2 per cent in 2013 and 2.4 per cent in 2014.
- However, in some contexts, the term russkii can also be interpreted
more broadly, reflecting a certain ambiguity about ethnic categories
inherent in the language that is also explored by the NEORUSS
survey and discussed further below. There is strong reason to believe
that the Kremlin actively exploits this ambiguity (Shevel 2011).
