GESIS - DBK - ZA5300

ZA5300: Pre-election Cross Section (GLES 2009)

Downloads and Data Access

On this page you find the complete metadata and an overview of all available data sets and documents for the requested study. The download of all files is possible from our central search page under the following link:
ZA5300 Downloads and Data Access

List of Files

List of Files


  • (Dataset) 7 KBytes
  • ZA5300_de_RecodeMissings_v5-0-2.sps (Dataset) 22 KBytes
  • ZA5300_de_v5-0-2.dta (Dataset Stata) 3 MBytes
  • ZA5300_de_v5-0-2.sav (Dataset SPSS) 6 MBytes
  • (Dataset) 6 KBytes
  • ZA5300_en_RecodeMissings_v5-0-2.sps (Dataset) 22 KBytes
  • ZA5300_en_v5-0-2.dta (Dataset Stata) 3 MBytes
  • ZA5300_en_v5-0-2.sav (Dataset SPSS) 6 MBytes


  • ZA5300_fb.pdf (Questionnaire) 466 KBytes
  • ZA5300_Listenheft.pdf (Questionnaire) 708 KBytes
  • ZA5300_Listenheft_Soziodemographie.pdf (Questionnaire) 264 KBytes
  • ZA5300_q.pdf (Questionnaire) 441 KBytes

Other Documents

  • (Other Document) 2 MBytes
  • ZA5300_Aenderungen_v4-0-0_zu_v5-0-0.pdf (Other Document) 31 KBytes
  • ZA5300_sb.pdf (Study Description) 163 KBytes
  • ZA5300_Variablenlabels_v5-0-0.pdf (Other Document) 46 KBytes
Availability Availability A - Data and documents are released for academic research and teaching.
Download of Data and Documents Download of Data and Documents All downloads from this catalogue are free of charge. Data-sets available under access categories B and C must be ordered via the shopping cart with a few exceptions. Charges apply! Please respect our Terms of use.

Bibliographic Citation

Citation Citation GLES (2019): Pre-election Cross Section (GLES 2009). GESIS Data Archive, Cologne. ZA5300 Data file Version 5.0.2,
Study No.ZA5300
TitlePre-election Cross Section (GLES 2009)
Other Titles
  • German Longitudinal Election Study (Project Title)
Current Version5.0.2, 2019-2-26,
Date of Collection10.08.2009 - 26.09.2009
Principal Investigator/ Authoring Entity, Institution
  • Rattinger, Hans - Universität Mannheim
  • Roßteutscher, Sigrid - Universität Frankfurt
  • Schmitt-Beck, Rüdiger - Universität Mannheim
  • Weßels, Bernhard - Wissenschaftszentrum Berlin für Sozialforschung
Contributor, Institution, Role
  • Roßteutscher, Sigrid - Universität Frankfurt - ProjectLeader
  • Scherer, Philipp - Universität Frankfurt - ProjectManager
  • Bieber, Ina - Universität Frankfurt - ProjectMember
  • Bytzek, Evelyn - Universität Frankfurt - ProjectMember
  • Förster, André - GESIS - Leibniz-Institut für Sozialwissenschaften - DataCurator
  • Kratz, Sophia - GESIS - Leibniz-Institut für Sozialwissenschaften - DataCurator
  • Kulick, Manuela - GESIS - Leibniz-Institut für Sozialwissenschaften - DataCurator
  • Roßmann, Joss - GESIS - Leibniz-Institut für Sozialwissenschaften - DataCurator


AbstractEvaluation of parties and politicians. Attitude to political questions.
Categories Categories
  • Political Attitudes and Behavior
Topics Topics
  • 5.6 Social conditions and indicators
  • 5.8 Social behaviour and attitudes
  • 9.2 Mass media
  • 11.5 Mass political behaviour, attitudes/opinion
  • 11.6 Government, political systems and organisation
  • 11.7 Elections


Geographic Coverage
  • Germany (DE)
UniverseThe population comprises all persons with German citizenship resident in the Federal Republic of Germany, who had a minimum age of 16 years and lived in private households at the time the survey was being conducted.
Analysis Unit Analysis Unit
  • Individual
Sampling Procedure Sampling Procedure
  • Probability: Stratified: Disproportional
Mode of Collection Mode of Collection
  • Face-to-face interview: Computer-assisted (CAPI/CAMI)
Time Method Time Method
  • Cross-section
Kind of Data Kind of Data
  • Numeric
  • Text
Data CollectorBIK MARPLAN, Offenbach
Date of Collection
  • 10.08.2009 - 26.09.2009

Errata & Versions

VersionDate, Name, DOI
5.0.2 (current version)2019-2-26 Release5-0-2
5.0.12017-6-7 Release5-0-1
5.0.02011-10-25 Pre-Release5-0-0
4.0.02010-6-9 Pre-Release1.3
3.0.02010-4-1 Pre-Release1.2
2.0.02010-3-8 Pre-Release1.1
1.0.02009-11-25 Pre-Release1.0
Errata in current version
Version changes
Changes between version 5.0.2 and it's previous version
DateSubjectDescriptionCorrection Description
2019-2-26Extensive revision of the variable labels and value labels in the English data set2019-2-26
Changes between version 5.0.0 and it's previous version
DateSubjectDescriptionCorrection Description
2011-10-25Overall, nine new weights were added to the dataset in Version 5.0.0.: East-West weight, transformation weight, a combined transformation and East-West weight and six further different sociodemographic and regional weights. They allow for East-West and all-German Analyses (with and without taking the transformation weight into account). A detailed description of all available weights and their computation can be found in the study description.--
2011-10-25The new variable ostwest ("East/West (Berlin citizens attributed to East Germany)") was generated. It indicates, if the respondent lives in East or West Germany (Berlin attributed completely to East Germany). This variable is also the basis for computing the East/West weight. In addition, the variable quelle was renamed into ostwest2 and includes the exact attribution of Berlin citizens to East and West Germany. This means that within Berlin it is distinguished between the former affiliation to the German Democratic Republic (GDR/DDR) and the Federal Republic of Germany (FRG/BRD).--
2011-10-25In a few exceptional cases, respondents received the questions regarding the most important or second important issue and/or the solution expertise, although the respondent had not given a valid answer to the questions regarding the most important, the second most important or the third most important problem in Germany (Q32, Q33, Q34) or had refused the answer. These wrongly filtered respondents were assigned to the code "trifft nicht zu" ("not applicable").--
2011-10-25Eight respondents received the question regarding their occupation (D221 "Berufliche Tätigkeit", "Occupation"), which was a mistake due to wrong filtering. These wrongly filtered respondents were assigned to the code "trifft nicht zu" ("not applicable").--
2011-10-25There is a filter mistake concerning the question D286A ("Im HH gesprochene Sprache", "language spoken at home"). According to the quesionnaire, the question should have been asked if the respondent or his/her spouse was born abroad. The filter was implemented wrongly so that only respondents who were born abroad and whose spouse was born abroad (or who have no spouse) were asked this question. Additionally, respondents were asked, who did not state their country of birth and who do not have a spouse or who did not make a statement concerning their spouse. --
2011-10-25The respondents´ answers to the questions D280 ("Geburtsland","Country of birth") D282 ("Geburtsland Partner/in", "Partner´s country of birth"), D284 ("Geburtsland Mutter", "Mother´s country of birth") and D285 ("Geburtsland Vater", "Father´s country of birth") were recoded according to the unified coding scheme "countries". In addition, all open statements were coded and summed up with the predefined codes of the coding scheme into one variable.--
2011-10-25The indication of the sample point (point), in which the interview was realized, was anonymized for data security reasons and was added to the dataset as the variable vpoint (virtual sample point).--
2011-10-25The variable labels of the questions Q89, Q92, Q95 and Q98 had been misleading and therefore were relabeled into "Nutzungshaeufigkeit TV-Nachrichten Sendername" ("Frequency of use of TV news programmes: television channel name").--
2011-10-25The variables “kand1” and “kand5” partly contained incorrect candidate names due to the fact that not all candidates for the election were certain on the date of the study. All cases, in which the respondents were asked about the wrong candidates (that was the case in presumably two election districts), were marked in the variables problem2 and problem3. Moreover, the variables wkkunion, wkkspd, wkkgruen, wkkfdp und wkklinke were generated, which contain the names of all candidates ordered by party and which base on official documents of the federal elections administrator.--
2011-10-25In 21 cases, year of birth (d201b) and age (q1a) did not correspond. These cases are marked in the variable “problem1”.--
2011-10-25Question D273 ("Religious denomination"): the answers "Judaism", "Islam" and "Other" were summarized to "Other religious denomination" due to data security considerations. Further answers were recoded respective to the coding scheme "Religious denomination", which can be found on the GLES homepage at GESIS. --
2011-10-25The variables D274 "Church attendance", D275 "Synagogue attendance" and D276 "Mosque attendance" were summarized to the variable dkirchg "church/mosque attendance" due to data security considerations.--
2011-10-25The variable label of the variable D278 ("German citizenship") had been misleading, because the question did not ask, whether the respondent holds a German citizenship, but whether the respondent holds a German citizenship since birth. The variable label was renamed into "German citizenship since birth".--
2011-10-25Based on the age of the respondents at the time point of the interview, the variable ragelt18 ("Respondents, who are younger than 18 years" (Source: q1a)) was generated. It indicates respondents in the dataset, who are younger than 18 years.--
2011-10-25Based on the month of birth (D201a) and the year of birth (D201b/in English questionnaire titled as Q201b)) the age of each respondent (in years) at the date of 09-30-2009 was computed (D201rc).--
2011-10-25Based on the categories of the variable D270b ("Net household income, with categories") the variable D270 ("Net household income, categorized") was generated. This variable summarizes the information of D270a and D270b.--
2011-10-25Several variables with open statements were coded. Questions regarding "other" parties were mostly concerned. If possible, these codings were not captured in separate variables, but instead summed up with the other (closed) codings of the respective question into one variable. --
2011-10-25Concerning some questions, the answers of the respondents were captured in several steps. For instance, mentions of the parties that are represented in the German Bundestag and mentions of "other" parties were mostly captured in separate variables. In order to facilitate working with the dataset, answers to these questions were summed up into two variables. Usually, the first variable ("Version A") contains the predefined answers plus an "other"-category and the second variable ("Version B") contains the further mentions.--
2011-10-25All umlauts (ä, ö, ü) and all "ß" in the variable labels and value labels were substituted by "ae", "oe", "ue" and "ss" in order to guarantee international usability of the dataset.--
2011-10-25The missing values were recoded in some cases due to harmonisation of the GLES studies. The answer "don´t know" is coded by the value (98), (998) or (9998), Refusals or missing answers ("No answer") are coded by the value (99), (999) or (9999). In case a question did not apply to a respondent due to filtering ("Not applicable"), this is coded by the value (100), (1000) or (10000).--
2011-10-25In order to ensure usability of the GLES datasets in many different programmes, the maximum length of variable labels has been limited to 80 characters and the maximum length of value labels has been limited to 120 characters. For some variables, the maximum length of the variable labels and the maximum length of the value labels were exceeded. In these cases, the labels were shortened so that the contentual meaning remained persistent.--
2011-10-25Due to harmonisation of the GLES datasets, questions with regard to the political parties (e.g. voting intention, voting decision, party Identification and form of party identification et cetera) were coded by a unified coding scheme. This scheme is available on the GLES homepage at GESIS.--
Changes between version 4.0.0 and it's previous version
DateSubjectDescriptionCorrection Description
2010-6-9The vocational education was extended by one code in the post-election cross section. To prevent matching problems, the labels of the pre-election cross section were adjusted to the labels in the post-election cross section (variable: d208).--
2010-6-9Correction of value labels (variables: q24a bis q24g, q43b, d241)--
2010-6-9Renaming of the weights (uniformity of pre- and post-election cross section) (variables: IPFWEIGHT, gewicht_ow--
2010-6-9Adding weights (variables: IPFWEIGHT_VWOST, IPFWEIGHT_VWWEST)--
2010-6-9Construction of a new variable which indicates, if the respondent has given the correct answer to the first/second vote knowledge question (variables: q6rf)--
Changes between version 3.0.0 and it's previous version
DateSubjectDescriptionCorrection Description
2010-4-1Version 3.0.0 (Pre1.2) was published only including the cumulation of cross sections--
Changes between version 2.0.0 and it's previous version
DateSubjectDescriptionCorrection Description
2010-3-8Correction of erroneous codings (variables: q32, q33, q34)--
2010-3-8Generation of id-number, flag variable for problematic and implausible values, comment variable (variables: Lfdn, ProbFlag, ProbKomm)--
2010-3-8Addition of a variable, which indicates the age on the election day (variable: D201_c)--
2010-3-8Correction of erroneous codings (variables: q32, q33, q34)--
2010-3-8Marginal correction of variable labels and value labels (no substantial changes)--

Further Remarks

NotesThe documented study was conducted in the context of the German Longitudinal Election Study (GLES), 2009. The data of this study are also available as cumulation (ZA5302) with the Post-election Cross Section.
Number of Units: 2173
Number of Variables: 638
Analysis System(s): SPSS, Stata


Relevant full texts
from SSOAR (automatically assigned)


Research Data Centre
  •  German Longitudinal Election Study (GLES)
    The German Longitudinal Election Study (GLES) is a DFG-funded project which made its debut just prior to the 2009 federal election. GLES is the largest and most ambitious election study held so far in Germany. Although the initial mandate is to examine and analyse the electorate for three consecutive elections, the aspired goal is to integrate the project within GESIS as an institutionalized election study after the federal election of 2017, and hence to make it a permanent study.