THE 2005 BES SURVEY
DATA
UPDATED:
DECEMBER
11, 2005
2005
BES Survey Data Memo (requires Adobe Acrobat)
Note
on sampling and weighting procedures
(requires Adobe Acrobat)
2005
BES Survey Data file-SPSS zip
2005
BES Survey Data file-Stata zip
Pre-Election questionnaire (requires Adobe Acrobat)
Post-Election questionnaire (requires Adobe Acrobat)
Mailback
questionnaire - England (requires Adobe Acrobat)
Mailback questionnaire
- Scotland (requires Adobe Acrobat)
Mailback questionnaire
- Wales (requires Adobe Acrobat)
THE
2005 BES SURVEY DATA
RELEASE:
DECEMBER 11, 2005
This updated
release of the 2005 BES survey data contains a preliminary version of the
validated
vote variable (VOTEVAL)
and updated Heath-Goldthorpe five-category social class variables for
respondent (RGOLDNEW)
and partner (SGOLDNEW). It also contains the self-completion
questionnaire data
and various 'Natcen' filters and weights (including a new set of weights for the
pre-campaign,
post-election, self-completion panel respondents (N =2343)). These data were
gathered by
in-person CAPI interviews and a self-completion questionnaire administered to
all
respondents
participating in the post-election face-to-face survey. Fieldwork was conducted
by
the National
Centre for Social Research (NATCEN) under the supervision of Research Director,
Katarina Thomson.
1. Data
Files: The data files are in SPSS and STATA formats. One file is a
compressed SPSS
(.sav) file called SPSS0214.sav, and the other is a compressed STATA (.dta) file called
STAT0214.dta.
Download one or both of these files. Users requiring the
data in other
formats can contact Harold Clarke (hclarke@utdallas.edu) for assistance.
Alternatively,
they can use a program such as DBMS COPY or STAT TRANSFER to translate
one of the present
files into the format needed.
2. Sample
sizes: The pre-campaign and post-election cross-sectional surveys use
samples of
people 18 years of
age or older drawn from the British postcode address file. There are
substantial
booster samples for Scotland and Wales. The unweighted N for the entire
post-election
sample = 4161. The
post-election sample contains a pre-campaign-post election panel
with a sample size
= 2959, as well as a post-election only "topup", N = 1202. The N for
the pre-campaign
survey is 3589.
For the self-completion questionnaire, the N = 3226. The N for the
pre-campaign,
post-election, self-completion panel is 2343.
3. Filter:
The data files contain a filter variable called WAVE2 which will enable users to
select the pre-campaign
cross-section sample, the post-election cross-section sample, the
pre-campaign-post-election panel,
the pre-campaign-post-election-self-completion panel, and the self-completion
cross-section
sample.
4. Weights:
(i) Pre-campaign
cross-section: PREWTBR = weight for Great Britain; PREWTEN = weight for
England; PREWTSC =
weight for Scotland; PREWTWA = weight for Wales.
(ii) Post-election
cross-section: POSTWTBR = weight for Great Britain; POSTWTEN = weight
for England;
POSTWTSC = weight for Scotland; POSTWTWA = weight for Wales.
(iii) Pre-Post
election panel: PANWTBR = weight for Great Britain; PANWTEN = weight for
England; PANWTSC =
weight for Scotland; PANWTWA = weight for Wales.
(iv) Post-election
self-completion cross-section: CSCWTBR = weight for Great Britain;
CSCWTEN = weight
for England; CSCWTSC = weight for Scotland; CSCWTWA = weight for
Wales.
(v) Pre-campaign,
post-election, self-completion panel: PSCWTBR = weight for Great Britain;
PSCWTEN = weight
for England; PSCWTSC = weight for Scotland; PSCWTWA = weight for
Wales.
5. Variable
Names: Except for miscellaneous technical variables, variable names
reference
question numbers
in the pre-campaign, post-election and self-completion surveys. Pre-campaign
variable names
start with the letter "A," post-election variable names start with the
letter "B," and
self-completion
variables start with the letter "C". For example, the variable name
AQ1 refers to
a variable based
on responses to question 1 in the pre-campaign
survey, BQ1 refers to a variable
based on responses
to question 1 in the post-election
survey, and CQ1 refers to a variable based
on responses to
question 1 in the self-completion
survey.
6. Summary
Variables: These variables summarize information using basic
variables in the
data set. In the
present release, summary variables for the pre-campaign survey include: (i)
AMISTOT -- all
responses for most important issue facing the country; (ii) APARTYID,
ACLOSEID, ACLOSEW
and ASPID -- variables for responses to the traditional BES party
identification
battery; (iii) APTYSUP and APTYSUPW -- variables for an alternative ("party
supporter")
measure of partisanship. Comparable most important issue (BQ2MISZ) and party
identification
variables (BPARTYID, BCLOSEID, BCLOSEW, and BSPID) have been
constructed for
the post-election data. There are also several variables starting with the
letter "T"
that combine
demographic information from the pre-campaign and post-election surveys.
7.
Questionnaires: The
pre-campaign, post-election and self-completion questionnaires can be
downloaded
from the BES website.
8.
Comments, Questions, Problems? Please
contact a member of the BES team:
David
Sanders: sanders@essex.ac.uk
Paul
Whiteley: whiteley@essex.ac.uk
Harold
Clarke: hclarke@utdallas.edu
Marianne
Stewart: mstewart@utdallas.edu
Kristi
Winters: kwinte@essex.ac.uk
The
BES office telephone number at the University of Essex is: 01206873346
The British Election Study 
