=========================================================================== COMPARATIVE STUDY OF ELECTORAL SYSTEMS (CSES) - MODULE 6 (2021-2026) CODEBOOK PART 5: STUDY DESIGN AND WEIGHTS SECOND ADVANCE RELEASE - DECEMBER 16, 2025 CSES Secretariat www.cses.org =========================================================================== HOW TO CITE THE STUDY: The Comparative Study of Electoral Systems (www.cses.org). CSES MODULE 6 SECOND ADVANCE RELEASE [dataset and documentation]. December 16, 2025 version. doi:10.7804/cses.module6.2025-12-16. These materials are based on work supported by the American National Science Foundation (www.nsf.gov) under grant numbers SES-1760058 and SES-2214278, the GESIS - Leibniz Institute for the Social Sciences, the University of Michigan, in-kind support of participating election studies, the many organizations that sponsor planning meetings and conferences, and the numerous organizations that fund national election studies by CSES Collaborators. Any opinions, findings and conclusions, or recommendations expressed in these materials are those of the author(s) and do not necessarily reflect the views of the funding organizations. =========================================================================== =========================================================================== TABLE OF CONTENTS =========================================================================== ))) CSES ADVANCE RELEASE - WHAT USERS NEED TO KNOW ))) CSES DESIGN, DOCUMENTATION, & WEIGHTS - ADVICE TO USERS ))) OVERVIEW OF "CODEBOOK PART 5: STUDY DESIGNS AND WEIGHTS" ))) HOW TO NAVIGATE THE CSES MODULE 6 CODEBOOK ))) LIST OF TABLES IN CODEBOOK PART 5 ))) CSES MODULE 6: SUMMARY OF CSES MODULE 6 DATA >>> IDENTIFICATION, WEIGHT, AND STUDY ADMINISTRATION DATA >>> MODULE 6 QUESTIONNAIRE VERSION >>> POLITY WEIGHTS IN CSES >>> TYPE OF POLITY WEIGHTS BY ELECTION STUDY >>> DERIVATIVE WEIGHTS IN CSES ))) OVERVIEW OF STUDY DESIGN & WEIGHTS BY POLITY & ELECTION YEAR >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - AUSTRALIA (2022) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - AUSTRIA (2024) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - BRAZIL (2022) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - DENMARK (2022) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - FRANCE (2022) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - MONTENEGRO (2023) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - NEW ZEALAND (2023) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - NORTH MACEDONIA (2024) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - POLAND (2023) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - PORTUGAL (2022) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - PORTUGAL (2024) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - SLOVAKIA (2023) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - SLOVENIA (2022) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - SWEDEN (2022) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - SWITZERLAND (2023) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - TAIWAN (2024) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - TURKIYE (2023) >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - UNITED STATES (2024) =========================================================================== ))) CSES ADVANCE RELEASE - WHAT USERS NEED TO KNOW =========================================================================== This dataset and all accompanying documentation is the "Second Advance Release" of CSES Module 6 (2021-2026). An Advance Release is a preliminary version of this CSES product. It thus lacks some of the checking, cleaning, processing, documentation, data, and variables that we anticipate in the Full Release of this product. The product is still in development, and data will be added gradually. Many election studies and variables that will eventually be present in the CSES Module 6 Full Release are unavailable in this file. Advance Releases are provided as a service to the CSES user community for those analysts who find it valuable to work with preliminary versions of the dataset. Users should anticipate future changes and improvements to the naming, data, and documentation of variables and election studies that appear in an advanced release file. If users wish to re-use their programming syntax/code on a future release of this product, we recommend that the code be written to accommodate these potential changes. Users of the Advance Release may also wish to monitor the errata for CSES Module 6 on the CSES website to check for known errors that may impact their analyses. To view errata for CSES Module 6, go to Data Download on the CSES website, navigate to the CSES Module 6 download page, and click on the Errata link in the white box to the right of the page. We hope that until the Full Release of CSES Module 6 is available, users will find this and future CSES Module 6 Advance Releases helpful in their work. =========================================================================== ))) CSES DESIGN, DOCUMENTATION, & WEIGHTS - ADVICE TO USERS =========================================================================== We recommend that researchers become familiar with the CSES design, units of analysis, documentation, and dataset weights before beginning their investigations. For instance, while the set of respondents appearing within each election study represents their respective nations, the group of nations that appear within CSES is not a random sample of countries worldwide. Furthermore, while many election studies include 1,000 or so respondents, other election studies may consist of over 10,000 respondents. Some nations will have studies of more than one election in a CSES module, and occasionally there will be two independent studies of a single election. Last, some election studies include oversamples of specific sub-populations or would otherwise benefit from use of the included weight variables. We hope you find our website and documentation useful as you proceed with your work, and welcome any questions or suggestions you have. =========================================================================== ))) OVERVIEW OF "CODEBOOK PART 5: STUDY DESIGNS AND WEIGHTS" =========================================================================== Part 5 of the CSES Codebook provides users with general information about the design of each election study included in MODULE 6, as well as the original polity level weights for each study. The purpose of these overviews is to provide users with a quick summary of the study and the weights, including information about the sample size, the fieldwork period, sample selection procedures, the mode(s) of interview, and the language(s) the survey was administered in. For more detailed information about each study's design and its weights, including information on response rates, refusal conversion, questionnaire translation, respondent incentives etc., users are advised to consult the Design Reports for each polity included in the study, which are available on the CSES MODULE 6 Study Page: (see https://cses.org/data-download/cses-module-6-2021-2026/). =========================================================================== ))) HOW TO NAVIGATE THE CSES MODULE 6 CODEBOOK =========================================================================== In the CSES MODULE 6 dataset, all variables begin with the letter "F" (F being the sixth letter of the English alphabet and thus signifying MODULE 6). The CSES Codebook is especially extensive, and users are advised that the best way to navigate it is electronically. It is a .txt format, which allows it to be accessed via a variety of programs. The CSES Codebook can be navigated quickly in the electronic files, with the following commands allowing for quick searching: ))) = Section Header. >>> = Sub-section Header 1. <<>> = Sub-section Header 2. +++ = Tables. CSES QUESTION CLASSIFICATION = For survey-level variables only, CSES Question Classification details whether the variable is part of the CSES Core component, which are questions asked repeatedly in CSES Modules, whether a variable is part of the CSES Module Theme component, which are questions specific to the Module Theme under exploration and might not be included in CSES repeatedly, or whether a variable is a Derivative Variable, which is explained below. VARIABLE NOTES = Notes for particular variables. ELECTION STUDY NOTES = Notes for a particular election study. DERIVATIVE VARIABLE = Highlights a variable derived from another variable or variables within the CSES. POTENTIAL POLITY LEVEL BRIDGING IDENTIFIER = Highlights a variable that may be used for data bridging at polity level. POTENTIAL REGIONAL LEVEL BRIDGING IDENTIFIER = Highlights a variable that may be used for data bridging at regional level. POTENTIAL TIME BRIDGING IDENTIFIER = Highlights a variable that may be used for data bridging by time. POTENTIAL PARTY/COALITION LEVEL BRIDGING IDENTIFIER = Highlights a variable that may be used for data bridging at party/coalition level. POTENTIAL CSES PRODUCT BRIDGING IDENTIFIER = Highlights a variable that may be used for data bridging with other CSES products. For further details on the CSES MODULE 6 documentation, users are advised to consult Part 1 of the CSES Codebook. =========================================================================== ))) LIST OF TABLES IN CODEBOOK PART 5 =========================================================================== Below, we list the Tables located in Codebook Part 5. Tables can be accessed in the electronic version of the CSES Codebook by searching for "+++". - OVERVIEW OF ELECTION STUDIES INCLUDED IN MODULE 6 WITH NUMBER OF OBSERVATIONS, MODE OF DATA COLLECTION, AND FIELDWORK DATES - TYPE OF POLITY WEIGHTS BY INDIVIDUAL ELECTION STUDIES =========================================================================== ))) CSES MODULE 6: SUMMARY OF CSES MODULE 6 DATA =========================================================================== The Second Advance Release of CSES MODULE 6 contains data from the following 18 election studies in 17 polities. They are listed below in alphabetic order with an overview of some method parameters of each election study. | +++ TABLE: OVERVIEW OF ELECTION STUDIES INCLUDED IN MODULE 6 WITH | NUMBER OF OBSERVATIONS, MODE OF DATA COLLECTION, AND | FIELDWORK DATES | | POLITY (ELEC YEAR) No of Mode of Dates of Fieldwork | Observations Interview (Start-End date) | --------------------------------------------------------------------- | AUSTRALIA (2022) 3,269 MX May 23, 2022-Jun 05, 2022 | AUSTRIA (2024) 1,569 INT Sep 30, 2024-Oct 21, 2024 | BRAZIL (2022) 2,001 F2F Nov 19, 2022-Dec 04, 2022 | DENMARK (2022) 1,549 MX Nov 02, 2022-Feb 01, 2023 | FRANCE (2022) 1,575 INT Apr 28, 2022-May 26, 2022 | MONTENEGRO (2023) 1,200 F2F Oct 14, 2023-Oct 30, 2023 | NEW ZEALAND (2023) 910 MX Oct 18, 2023-Mar 09, 2024 | NORTH MACEDONIA (2024) 1,055 F2F Oct 25, 2024-Dec 04, 2024 | POLAND (2023) 1,500 F2F Nov 07, 2023-Dec 16, 2023 | PORTUGAL (2022) 1,010 F2F Feb 11, 2022-Mar 07, 2022 | PORTUGAL (2024) 1,001 F2F Apr 27, 2024-May 07, 2024 | SLOVAKIA (2023) 1,021 F2F Jan 31, 2024-Feb 26, 2024 | SLOVENIA (2022) 855 MX Sep 26, 2022-Dec 03, 2022 | SWEDEN (2022) 2,845 MX Sep 13, 2022-Jan 04, 2023 | SWITZERLAND (2023) 5,033 MX Oct 23, 2023-Jan 11, 2024 | TAIWAN (2024) 1,206 F2F Jan 15, 2024-May 31, 2024 | TURKIYE (2023) 1,508 F2F Aug 25, 2023-Oct 04, 2023 | UNITED STATES (2024) 4,764 MX Nov 07, 2024-Feb 19, 2025 | --------------------------------------------------------------------- | TOTAL 33,871 | | Key: F2F=Face-to-face. | TP=Telephone. | INT=Internet/Online. | MX=Mixed. | | Users are advised to consult the VARIABLE NOTES for Variables F1019_ | in Codebook Part 2 concerning fieldwork date classifications. Users are advised that the above information relates to this Advance Release. The product is still in development, and data will be added gradually. Many election studies and variables that will eventually be present in the CSES Module 6 Full Release are unavailable in this version of the file. --------------------------------------------------------------------------- >>> IDENTIFICATION, WEIGHT, AND STUDY ADMINISTRATION DATA --------------------------------------------------------------------------- CSES MODULE 6 includes several identification, weight, and study administration variables, allowing users the opportunity to account for many methodological differences in the fielding of CSES across different studies and polities. A list of relevant IDENTIFICATION, WEIGHT, AND STUDY ADMINISTRATION DATA are below this explanation and can be navigated in the CSES Codebook by searching for IDENTIFICATION, WEIGHT, AND STUDY ADMINISTRATION DATA in Part 2 of the CSES MODULE 6 Codebook, or by searching the dataset for variables beginning "F1XXX". - F1008 ID COMPONENT - WHETHER POLITY ADMINISTERED CSES MODULE 6 MULTIPLE TIMES - F1012_1 STUDY TIMING - F1012_2 STUDY TIMING WITH RESPECT TO COVID-19 PANDEMIC - F1013 STUDY CONTEXT - F1015_1 MODE OF INTERVIEW - STUDY - FIRST - F1015_2 MODE OF INTERVIEW - STUDY - SECOND - F1015_3 MODE OF INTERVIEW - STUDY - THIRD - F1016_1 MODE OF INTERVIEW - RESPONDENT - FIRST - F1016_2 MODE OF INTERVIEW - RESPONDENT - SECOND - F1016_3 MODE OF INTERVIEW - RESPONDENT - THIRD - F1017 SELF-SELECTION INTO MODE OF INTERVIEW - F1018_1 DURATION OF FIELDWORK - F1018_2 DAYS FIELDWORK STARTED POST ELECTION - F1019_M DATE QUESTIONNAIRE ADMINISTERED - MONTH - F1019_D DATE QUESTIONNAIRE ADMINISTERED - DAY - F1019_Y DATE QUESTIONNAIRE ADMINISTERED - YEAR - F1020_1 DAYS INTERVIEW CONDUCTED POST FIRST ROUND OF ELECTION - F1020_2 DAYS INTERVIEW CONDUCTED POST SECOND ROUND OF ELECTION - F1021 DURATION OF INTERVIEW - F1022_1 INTERVIEWER ID WITHIN ELECTION STUDY - F1022_2 INTERVIEWER GENDER - F1023 LANGUAGE OF QUESTIONNAIRE ADMINISTRATION - F1024 QUESTIONNAIRE VERSION - F1100 ID COMPONENT - SAMPLE COMPONENT - F1101_1 ORIGINAL WEIGHT: SAMPLE - F1101_2 ORIGINAL WEIGHT: DEMOGRAPHIC - F1101_3 ORIGINAL WEIGHT: POLITICAL - F1102_1 FACTOR: MEAN OF SAMPLE WEIGHT - F1102_2 FACTOR: MEAN OF DEMOGRAPHIC WEIGHT - F1102_3 FACTOR: MEAN OF POLITICAL WEIGHT - F1103_1 POLITY WEIGHT: SAMPLE - F1103_2 POLITY WEIGHT: DEMOGRAPHIC - F1103_3 POLITY WEIGHT: POLITICAL - F1104 FACTOR: SAMPLE SIZE ADJUSTMENT - F1105_1 DATASET WEIGHT: SAMPLE - F1105_2 DATASET WEIGHT: DEMOGRAPHIC - F1105_3 DATASET WEIGHT: POLITICAL --------------------------------------------------------------------------- >>> MODULE 6 QUESTIONNAIRE VERSION --------------------------------------------------------------------------- Studies included in the CSES MODULE 6 fielded one of two versions of the CSES questionnaire: the pilot questionnaire or the finalized questionnaire. Variable F1024 "QUESTIONNAIRE VERSION" in the Dataset classifies for Users the studies that fielded the pilot version of the questionnaire and the finalized questionnaire. All studies in the current release fielded the finalized questionnaire. The differences between the pilot and finalized questionnaire will be documented in the VARIABLE NOTES in CSES Codebook Part 2, once pretest studies are integrated into the CSES MODULE 6 dataset. --------------------------------------------------------------------------- >>> POLITY WEIGHTS IN CSES --------------------------------------------------------------------------- CSES provides users with up to three original weights from each national election study (see variables F1101_), namely: - SAMPLE WEIGHT (variable F1101_1): intended to correct for unequal selection probabilities resulting from booster samples procedures for selection within the household, non-response, or other sample design features - DEMOGRAPHIC WEIGHT (variable F1101_2): intended to adjust sample distributions of socio-demographic characteristics to more closely resemble the characteristics of the population - POLITICAL WEIGHT (variable F1101_3): intended to reconcile discrepancies in the reported electoral behavior of respondents vis-a-vis official electoral counts. Users are advised to read carefully about the different weights in CSES to ascertain whether their analyses should be subjected to weighting, and if so, which kind. The CSES project does not provide advice as to which weights are appropriate to use in particular circumstances. This is best left to analysts to decide based on their detailed knowledge of the research question under investigation. We advise analysts to consult each Polity's Design Report on the CSES MODULE 6 Study Page (see https://cses.org/data-download/cses-module-6-2021-2026/). For information on derivative weights calculated for the Cross-National Dataset, users should consult Parts 1 and 2 (variables F1101-F1105) of the CSES Codebook. --------------------------------------------------------------------------- >>> TYPE OF POLITY WEIGHTS BY ELECTION STUDY --------------------------------------------------------------------------- | +++ | TABLE: TYPE OF POLITY WEIGHTS | BY INDIVIDUAL ELECTION STUDIES | | Sample Demographic Political | POLITY (ELEC YEAR) Weight Weight Weight | ----------------------------------------------------------- | AUSTRALIA (2022) X - - | AUSTRIA (2024) - X X | BRAZIL (2022) - X - | DENMARK (2022) - X X | MONTENEGRO (2023) - X - | NEW ZEALAND (2023) - X X | NORTH MACEDONIA (2024) - X X | POLAND (2023) - X - | PORTUGAL (2022) - X X | PORTUGAL (2024) - X - | SLOVAKIA (2023) X X - | SLOVENIA (2022) X X - | SWEDEN (2022) - X - | SWITZERLAND (2023) X - X | TAIWAN (2024) - X - | TURKIYE (2023) - X - | UNITED STATES (2024) X X - | ----------------------------------------------------------- | KEY: X = available; - = not available. | | Weights are unavailable for FRANCE (2022). --------------------------------------------------------------------------- >>> DERIVATIVE WEIGHTS IN CSES --------------------------------------------------------------------------- Additionally to the original polity weights listed in the previous section by national election study, CSES provides users with several derivative weights, constructed from the original weights. They are: - FACTOR WEIGHTS (variable F1102) These variables report the mean weight of each type, within each polity. The resulting factors are then used to create the derivative Polity Weights (F1103 explained below). - POLITY WEIGHTS (variable F1103) These variables report standardized versions (with a mean of 1 within the polity) of the original weights provided with the component election studies, described in F1101. They are the ratios of each weighting factor to the mean weight (F1102) of each type, calculated within each polity. - SAMPLE SIZE ADJUSTMENT WEIGHT (variable F1104) This variable reports the ratio of the average sample size to each election study sample. The resulting factor is then used to create the derivative Dataset Weights, F1105. - DATASET WEIGHTS (variable F1105) These variables are intended for micro-level analyses involving the entire CSES sample. Using the sample size adjustment (F1104), the centered weights (F1103) are corrected such that each election study component contributes equally to the analysis, regardless of the original sample size. Details of the calculation of the above derivative weights, including the precise STATA code used to create the weights, can be found in the variable notes for variables F1102, F1103, F1104, and F1105 in Part 2 of the CSES MODULE 6 Codebook. Analysts are advised to read the weight documentation carefully to ensure their analyses are weighted appropriately (if applicable). The CSES project does not provide advice as to which weights are appropriate to use in particular circumstances. This is best left to analysts to decide based on their detailed knowledge of the research question under investigation. We advise analysts to consult variable notes for F1102-F1105 in Codebook Part 2 for more specific information on how the derivative weights were calculated for the Cross-National Dataset. =========================================================================== ))) OVERVIEW OF STUDY DESIGN & WEIGHTS BY POLITY & ELECTION YEAR =========================================================================== --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - AUSTRALIA (2022) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2022 Federal election in Australia occurred on Saturday, May 21, 2022. Fieldwork for the Australian study began on May 23, 2022, two days after election day. Data collection was completed after 14 days, ending on June 5, 2022. The study contains information from Australian adults aged 18 years or over who were enrolled on the electoral register and eligible to vote. The only excluded group is prison inmates serving sentences below three years who would be eligible to vote. This represents 0.11% of the electorate. The study, undertaken by Australian National University (ANU), was conducted as part of wave 61 of the Social Research Center's Life in Australia panel. Members of the panel are Australian residents aged 18 years or more. A total of 4,338 active panel members were invited to participate in the study, with 3,556 completing the interview. Of these 3,556 interviews, 287 respondents were removed from the data because they were not citizens and therefore, ineligible to vote, reducing the total sample size to 3,269 observations. The survey was administered in English. The study used a mixed-mode approach, including both Internet and telephone (landline and cell) surveys. Respondents were initially contacted and recruited via their landline or mobile phone and offered to complete the questionnaire online. Still, they could self-select into the alternative mode and complete the study via telephone if they were uncomfortable completing surveys over the Internet or if Internet access was problematic. Of the 3,269 respondents comprising the CSES sample, 3,158 observations completed the survey online, while the remaining 111 respondents were interviewed via telephone. A dual-frame random digit dialing (RDD) sample design was employed to undertake recruitment of the Life in Australia panel in 2016, with a 30:70 split between the landline RDD sample frame and a cell phone RDD sample frame. An alternating next/last birthday method was used for the landline phone sample to randomly select respondents from households with two or more people. For the cell phone sample, the individual who answered the call was the selected respondent. Only one member per household was invited to join the panel. Between May 2018 and September 2021, the panel was refreshed multiple times with new panelists. All respondents receive a small incentive for joining the panel and another incentive for each survey they complete. Further details on the panel are provided in the study's Design Report. <<>> POLITY WEIGHTS The 2022 Australian Study provides a COMBINED SAMPLE and DEMOGRAPHIC WEIGHT. This weight adjusts for the unequal probability of selection. It is calculated based on a model-based approach that uses a model to "project" the responding sample to the population. The panel collects more than 20 characteristics about respondents from which the final set of covariates selected for the model was: Adults in the household, age group by highest education, gender, language spoken at home, remoteness, and state. This set was chosen since it yielded a satisfactory between bias and efficiency, and contained a wide range of model covariates. To adjust for wave biases, the weights were further adjusted through the use of propensity scores. Whereas all non-citizens (287 from the overall 3,556 respondents) were dropped from the dataset for CSES Module 6, the weight calculation included these non-citizens. Collaborators note that their statistician, who devised the weights, advised them to retain the weight calculated based on ineligible respondents, as a weight excluding them would make negligible differences to analysis. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - AUSTRIA (2024) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2024 Austrian Federal lower house election was held on Sunday, September 29, 2024. Fieldwork for the study began on September 30, 2024, lasting 22 days, with completion on October 21, 2024. The questionnaire was administered in German as an online survey. The total number of observations from the Austrian Study is 1,569 interviews. The study was undertaken by the Department of Government from the University of Vienna and funded by the same department as well as the Austrian Federal Ministry of Education, Science and Research. The dataset was archived by the Austrian Social Science Data Archive (AUSSDA). The study was conducted as part of wave 7 of the Digitize! Online Panel. The sample is meant to be representative of all eligible voters in Austria as of election day (Austrian citizens aged 16 or older). Respondents younger than 16 on election day, as well as respondents currently not residing in Austria were not eligible. Austrians without Internet access were excluded from the sampling frame (estimated to be 5% of the population). The Digitize! project employs a probability-based sampling method, drawing addresses from residents aged 16 and above from Austria's Central Register of Residents (Zentrales Melderegister, ZMR), with sampling stratified by region (NUTS3). Recruitment for the panel is done on an ongoing basis. In every wave, panelists as well as previously uncontacted individuals are invited to participate in the survey. On September 30, 2024, 3,723 active panelists of the Digitize! Online Panel were invited via email to the online survey. All panelists, who had not yet participated, were sent a reminder on October 3. A final reminder was issued on October 7, 2024, targeting only panelists aged 65 and older, as well as those with lower levels of formal education. The average number of contact attempts is therefore 2 to 3. Overall, 1,569 respondents completed the interview. After the completion of the online survey, all panelists were given the option to receive a 5 EUR SPAR supermarket voucher. To encourage greater participation among younger individuals, a random subsample of panelists under the age of 25 was selected to receive a higher incentive payment. These 232 individuals were informed in their email invitation that they would receive a 10 EUR SPAR voucher upon fully completing the survey. <<>> POLITY WEIGHTS The 2024 Austria Study provides a DEMOGRAPHIC WEIGHT and a COMBINED DEMOGRAPHIC and POLITICAL WEIGHT for post-stratification adjustments according to known population distributions. The demographic weight (F1101_2) adjusts the sample to match known demographic characteristics in the voting age population (age, gender, education, marital status, region, and employment type). The political weight (F1101_3) is a combined weight that adjusts for the reported voting behavior in the 2024 Austrian general election on top of the demographic weight. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - BRAZIL (2022) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2022 Brazilian general elections were held on Sunday, October 2, 2022 (first round), with a second round of Presidential elections held on Sunday, October 30, 2022. The Brazilian Election Study was conducted between November 19 and December 4, 2022. Fieldwork lasted 16 days, starting 48 days after the first round. Quaest collected data in face-to-face interviews using an electronic questionnaire. Interviewing was conducted in Portuguese, with all Brazilian citizens aged 16 years old and older and registered to vote eligible to be interviewed. The total number of observations from the Brazil Study is 2,001. The survey employed a multi-stage stratified random sample design. In the first stage, municipalities, which constitute the primary sampling units (PSUs), were selected through systematic PPS (Probability Proportional to Size), based on the voting-age population. Prior to selection, municipalities were stratified by region and state. In total, 119 PSUs were selected. In the second stage, census tracts within municipalities were selected again by PPS. In the third and final stage, a fixed number of respondents were selected from within each census tract by quota sampling based on gender, age, education, and family income. Quotas were set in accordance with the latest numbers from the TSE 2018 (Superior Electoral Court), the Brazilian Institute of Geography and Statistics' (IBGE) national household surveys from 2022 and 2021, as well as the latest census from 2010. The households to be contacted within census tracts were chosen by random walk. If the individual within the selected household matched the quotas and the interview was successfully conducted, the interviewer, walking clockwise, was instructed to skip three households and approach the fourth one. If the respondent was not available or did not fit the quota, the interviewers would move to the next household. <<>> POLITY WEIGHTS The 2022 Brazil Study provides a DEMOGRAPHIC WEIGHT. The weight corrects for imbalances between the sample and the population regarding gender, age, education, household income, and region - based on official records of the Superior Electoral Court 2022 and the IBGE-PNAD 2022 (Brazilian Institute for Geography and Statistics). Calculations of the weight are based on raking/iterative proportional fitting, using the package "anesrake" in the statistical software R. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - DENMARK (2022) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2022 Danish general election took place on Tuesday, November 1, 2022, except for the Faroe Islands, where respondents voted early on Monday, October 31, 2022, due to a regional holiday. Fieldwork for the 2022 Danish post-election study was conducted by Epinion A/S between November 2, 2022 - one day after the election - and February 1, 2023 (lasting 92 days). The questionnaire was administered in Danish as an online survey with the option to answer a paper-based mail-back questionnaire. The sample is meant to represent eligible voters in Denmark - Danish citizens aged 18 and older who are registered to vote automatically if they permanently live in Denmark. The total number of observations from the Denmark Study is 1,549. The study is based on a simple random sample drawn from the Danish person register ("CPR-register"), wherein all Danish citizens and long-term inhabitants of Denmark must be registered. Hence, individuals were the primary sampling unit and had the same probability of being selected. Respondents were initially contacted personally via digital mail ("Digital Post") with the survey being verified by a personal access coding system - or by mail for those exempt from digital postal services. With few exceptions, digital mail is mandatory for Danish citizens. Respondents initially not answering the web survey were mailed a paper questionnaire as an alternative to the web survey. About 1% of the sample made use of this opportunity, with the rest answering the questionnaire online. To boost response rates, reminders and prompts have been employed through various channels (digital post, letters, SMS, postcards). Further, collaborators conducted a lottery among all survey participants in which respondents could win gift cards worth 500 DKK. <<>> POLITY WEIGHTS The 2022 Denmark Study provides a DEMOGRAPHIC WEIGHT and a COMBINED DEMOGRAPHIC and POLITICAL WEIGHT. The demographic weight corrects distributions in the sample to known socio-demographic population characteristics (joint distribution of age groups, region, gender, plus education). The political weight was created using raking and builds upon the demographic weight, additionally adjusting the sample according to vote choice in the 2022 Danish general election. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - FRANCE (2022) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2022 French Presidential elections were held across two rounds. The first round was held on Sunday, April 10, 2022, with the second round run-off held on Sunday, April 24, 2022. Interviewing for the French election study began on April 28, 2022 (18 days after the first round of the Presidential elections) and continued until May 26, 2022 (a total of 36 days). The survey was administered in French, via the Internet. The total number of observations for the study is 1,575. The sample is representative of the voting-eligible population, i.e., French citizens who were 18 years or older at the time of the election, and living in mainland France (excluding Corsica). Respondents for the French 2022 study were sampled from the ongoing ELIPSS panel ("Etude Longitudinale par Internet Pour les Sciences Sociales"). ELIPSS is an internet panel that was established in 2012, with two refresher samples conducted in 2016 and 2020. Panelists are selected randomly and respond online every month to surveys designed by researchers. Further, they are provided with a touchscreen tablet connected to the internet to participate in ELIPSS surveys. Respondents recruited in 2012 and 2016 were sampled from the 2011 and 2014 censuses, respectively. For the 2012 panel wave, participants were recruited via various modes, including postal, phone and face-to-face. The 2016 refresher sample was recruited entirely face-to-face. The second refresher sample conducted in 2020 originates from the 2018 Fideli database (statistical database on housing and individuals based on fiscal administration data). Originally planned as face-to-face, this refreshment phase had to be adapted to a multi-mode approach due to the COVID-19 pandemic. By January 2021, the panel included 2,300 active participants. Variable F1100 in the CSES Module 6 datasets distinguishes between the three panel samples respondents originate from (2012, 2016, 2020). The 2022 French Election Study (FES 2022) was part of the wider ELIPSS panel program, comprising a total of five rounds of questioning conducted from November 2021 to July 2022. The fourth round included the CSES Module 6 survey questions. However, researchers are advised that demographics as envisaged by CSES were not collected as part of this fourth wave. Instead, the respective demographic and socio-economic information was gathered through the regular annual ELIPSS demographic survey. <<>> POLITY WEIGHTS No weights provided. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - MONTENEGRO (2023) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2023 Montenegrin election study covers the Presidential elections, held on Sundays, March 19 and April 2, 2023, and legislative elections held on Sunday, June 11, 2023. The post-election cross-sectional study was fielded between October 14 and October 30, 2023, with fieldwork starting 125 days after the legislative elections. The fieldwork was conducted by De Facto Consultancy based in Podgorica, with surveys administered face-to-face in Montenegrin. The sample is meant to be representative of eligible Montenegrin citizens aged 18 years or older registered to vote. Institutionalized persons were excluded from the sampling frame (about 0.5 percent of the voting-eligible population). The total number of observations from the Montenegro Study is 1,200. The study was conceptualized as a stratified multi-stage random sample with three stages in total. In the first stage, all polling stations were mapped to three geographic regions, North, Center and South. These regions reflect differences in population size, the distribution of ethnic groups, geography, economic strength, and history within Montenegro. Polling stations inside the regions formed the primary sampling units (PSUs) as well as the second level of stratification, as they were divided into three groups according to their size. This proxies for the settlement size and the urban-rural divide within the country, with bigger polling places being more urban, and smaller polling places being more rural. Upon assigning all polling places across the resulting nine cells (three per region, and additionally three within each of the regions), PSUs were selected through the Probability Proportional to Size (PPS) method. The respondent selection within PSUs in the third and final stage followed a predefined procedure. Specifically, interviewers were instructed to select households by random walk, selecting every third house or household. Within each household, respondents were selected from a pool of all voting-eligible household members according to the last birthday method. The final sample includes 15 interviews from each PSU. <<>> POLITY WEIGHTS The 2023 Montenegro Study provides a DEMOGRAPHIC WEIGHT based on national census statistics from 2011, the latest available year. The demographic weight is composed of two separate weights, one adjusting for the four main national groups (Montenegrins, Serbs, Albanians, and Bosniaks-Muslims), the other adjusting for gender and age groups (18-34, 35-54, 55+). The weight included in F1101_2 is a multiplication of the two individual weights. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - NEW ZEALAND (2023) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2023 New Zealand Parliamentary election was held on Saturday, October 14, 2023. Fieldwork for the post-election study, part of the New Zealand Election Study (NZES), was conducted between October 18, 2023, and March 9, 2024 (121 days in total). The data collection was conducted by COMPASS and funded by the Victoria University of Wellington and the New Zealand Electoral Commission. The data is archived by the Australian Data Archive. The survey was administered in English, Chinese, and Maori, and the study employed a mixed-mode approach, with respondents having the choice to participate via the Internet or via self-completion mail-back. Hence, respondents could self-select into interview modes. The total number of observations from the New Zealand Study is 910 interviews: 454 via self- completion mail-back and 456 via the Internet. The study is meant to be representative of the New Zealand Electoral Roll (about 94% of the eligible population). A person needed to be registered and above 18 years old to be interviewed. The survey employed a stratified random sampling procedure, in which postal addresses were sampled from the computerized electoral rolls. Young voters and voters of Maori descent were oversampled to adjust for lower response rates. For all completed surveys, survey verification checked whether the reported age and gender correctly matched the electoral roll. <<>> POLITY WEIGHTS The 2023 New Zealand Study provides a DEMOGRAPHIC and a COMBINED DEMOGRAPHIC and POLITICAL WEIGHT. Weights were created with iterative proportional fitting. The demographic weight controls for gender, age, Maori descent, education and region. Therefore, the weight makes the survey representative of the electoral roll from which it was sampled. The combined demographic and political weight controls for the same factors from above, but includes vote choice. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - NORTH MACEDONIA (2024) --------------------------------------------------------------------------- <<>> STUDY DESIGN The North Macedonian parliamentary election, the country's main election, was held on Wednesday, May 8, 2024, together with the second round of the Presidential elections. The first round of the Presidential elections was held two weeks earlier on Wednesday, April 24, 2024. Fieldwork, carried out by Indago, started on October 25, 2024, 170 days (or five months, 17 days) after the parliamentary elections, lasting until December 4, 2024 (41 days). Interviews were conducted face-to-face in Macedonian and Albanian. The sample is meant to be representative of the adult residential population of North Macedonia, i.e., citizens aged 18 or older at the time of sampling in September 2024. Collaborators advise that this includes the entire vote- eligible population, as voter registration in the country is passive - citizens automatically gain the right to vote upon turning 18 (the only exception applies to individuals who have been legally deprived of voting rights). Further, the original sample included citizens who became eligible to vote between the elections and the start of the interviewing process. This concerned one respondent born in July 2006, who turned 18 only after the April/May 2024 elections. This respondent was removed from the dataset and is hence not included in CSES. Since sampling focused exclusively on private dwellings, institutionalized individuals (nursing homes, hospitals, correctional facilities, etc.) were excluded from the sampling frame. While imprisoned individuals make up 0.14% of the voting-eligible population, estimates on the share of other excluded groups are unavailable. The total number of observations from the North Macedonia study, as included in CSES, is 1,055 interviews. The sample was conceptualized as a three-stage stratified random sample, of which the State Statistical Office of North Macedonia conducted the first two stages. The sampling frame from which households were sampled is based on two sources, namely the 2021 census of the North Macedonian population, households, and dwellings, as well as an internal document by the State Statistical Office, detailing the distribution of dwellings across enumeration areas. In the first stage, 188 enumeration areas were selected randomly, ensuring proportional representation of the population across NUTS-3 regions and the urban-rural divide. Second, from within each of the 188 selected enumeration areas, 10 addresses (dwelling units) were sampled randomly (amounting to 1,880 addresses in total). In the third and last stage, interviewers were instructed to select individuals from within households based on the next-birthday method, thereby ensuring a random selection of individuals from within the household. <<>> POLITY WEIGHTS The 2024 North Macedonia Study provides a DEMOGRAPHIC WEIGHT and a POLITICAL WEIGHT. The demographic weight serves as a post-stratification weight. It is based on iterative fitting, aligning weighted marginal distributions in the sample with known margins in the target population for gender, age, region, place of living, and ethnicity. The political weight is designed to align voting behavior in the sample (as reported in variable F3011_LH_PL) with official election results for the 2024 parliamentary election. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - POLAND (2023) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2023 Polish parliamentary election was held on Sunday, October 15, 2023. The post-election survey was fielded between November 7 and December 16, 2023 (a total of 25 days). The survey was conducted by DANAE face-to-face and in Polish. The total number of observations from the Poland Study is 1,500 interviews. The sample is meant to be representative of the Polish voting-age population. The survey is based on a multi-stage stratified random sample design, with quotas employed in the last stage for respondent selection. In the first stage, the sample was stratified by voivodeships to ensure adequate regional representation, and by locality size based on four categories: villages, cities up to 50,000 inhabitants, cities with a population from 50,000- 200,000, and cities with over 200,000 inhabitants. Second, to select starting addresses within each stratum, systematic random sampling from an address-based sampling frame was applied without replacement. From each sampled starting point, field interviewers followed a random route procedure to select households. Lastly, respondents within households were selected using predefined quotas on gender, age, and education. <<>> POLITY WEIGHTS The 2023 Poland Study provides a DEMOGRAPHIC WEIGHT. The weight adjusts the sample to population estimates for gender, age, education, voivodeship, and locality size. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - PORTUGAL (2022) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2022 Portuguese legislative election took place on Sunday, January 30, 2022. The post-election study was administered between February 11 and March 7, 2022 (25 days in total), with interviewer-administered in-person interviewing. Fieldwork was conducted by GfK Portugal, with questionnaires administered in Portuguese. The total number of observations from the Portugal 2022 Study is 1,010 interviews. The sample is meant to be representative of individuals aged 18 years or older residing in continental Portugal (mainland) and having the right to vote in the 2022 Parliamentary elections. The two autonomous regions of Madeira and the Azores were excluded from the sample frame (5 percent of the eligible population). The same holds for institutionalized persons (less than 1 percent of the eligible population). The study is based on a multi-stage stratified random sample. The sample was stratified according to locality size ("habitat") and EU NUTS II regions; a hierarchical geocode standard subdividing the territories of EU member states. In the first step, the random selection of localities within each of the NUTS II regions was carried out following the General Index of INE's Population Census, based on the sample matrix of NUTS II regions and habitat. In each designated locality, sample points were then again sampled randomly. Afterward, households were selected through random route. From within each household, the eligible adult who was the next to have a birthday was selected for the interview in the last step. <<>> POLITY WEIGHTS The 2022 Portugal Study provides a DEMOGRAPHIC WEIGHT and a COMBINED DEMOGRAPHIC and POLITICAL WEIGHT. The demographic weight (F1101_2) adjusts the sample to known population characteristics regarding gender, age, and educational attainment. Population estimates were taken from the 2011 Census. The political weight (F1101_3) is a combined weight that adjusts the sample to match the 2022 legislative election results on top of the demographic weight. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - PORTUGAL (2024) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2024 Portuguese legislative election took place on Sunday, March 10, 2024. The post-election study was administered between April 27 and May 7, 2024 (10 days in total), with interviewer-administered in-person interviewing. Fieldwork was conducted by GfK Portugal, with questionnaires administered in Portuguese. The total number of observations from the Portugal 2024 Study is 1,001 interviews. The sample is meant to be representative of individuals aged 18 years or older residing in continental Portugal (mainland) and having the right to vote in the 2024 Parliamentary elections. The two autonomous regions Madeira and the Azores were excluded from the sample frame (5 percent of the eligible population). The same holds for institutionalized persons (less than 1 percent of the eligible population). The study is based on a multi-stage stratified random sample. The random selection of localities within each NUTS II region was carried out according to the Population Census, taking into account the sample matrix of NUTS and locality size ("Habitat"). In each locality, a random selection of sampling points was carried out. Households were selected by random route and the individual selected within each household was the next to have a birthday, among all household members who were 18 years old and older. <<>> POLITY WEIGHTS The 2024 Portugal Study provides a DEMOGRAPHIC WEIGHT. The demographic weight (F1101_2) adjusts the sample to known population characteristics regarding gender, age, educational attainment, region, and habitat, with population estimates taken from the 2021 Census. The weight also includes religious attendance and union membership, with estimates taken from the European Social Survey Round 10. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - SLOVAKIA (2023) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2023 Slovakian Parliamentary elections were held on Saturday, September 30, 2023. Fieldwork began on January 31, 2024, and hence started 124 days after the election. It was completed on February 26, 2024. Data were collected via face-to-face interviewing. The survey was administered in Slovak. The total number of observations from the Slovakia Study is 1,021 interviews. FOCUS Centrum pre socialnu a marketingovu analyzu conducted the survey field work and the Slovak Research and Development Agency funded the data collection. The sample is meant to be representative of Slovak citizens aged 18 years or older on election day. The selection procedure can be described in three steps. In the first step, the number of primary sampling units (PSUs) were determined. Each PSU represents a certain territory within a municipality/city and the residents living there. 120 PSUs were included in the sample. The number of PSUs in individual regions was determined proportionally to the number of residents aged 18 and above living in individual regions of Slovakia. In the second step, specific households were randomly selected in each PSU. Since there is no complete publicly accessible list of all households available that can be used for random selection, the interviewer had to select individual households in a specified PSU according to the established rules, namely by the method of the so-called random walk within the given PSU. In each PSU, interviewers were randomly assigned a starting/exit point from which they began the random walk. From the selection point, the interviewer proceeded according to the precise rules of the random walk and searched for every third household. Interviewer recorded their progress on a special control sheet so that it was possible to check the correctness of the household selection. In the third step, interviewers attempted to interview household members randomly selected by the software from all household members aged 18 and above. This required listing all adult household members at the beginning of the interview. <<>> POLITY WEIGHTS The 2023 Slovakia Study provides a SAMPLE WEIGHT and a COMBINED SAMPLE and DEMOGRAPHIC WEIGHT. The sample weight compensates for the different probabilities of selecting respondents in terms of the total number of adult household members. This is done to correct for an over-representation of people from households with a lower number of adult household members. The weight was constructed by assigning a weight of "1" to a respondent from a single-member household, a weight of "2" to a respondent from a two-member household, etc. Finally, the individual weights calculated for individual respondents are standardized so that the average value of the design weights is "1", which ensures that the total number of respondents in the set with the weight turned on was equal to the number of respondents in the unweighted set. The combined sample and demographic weight adjusts the data to match known demographic characteristics, namely gender (two categories), age (six categories), region (eight categories), education (three categories), and community size (eight categories). It was calculated using an iterative proportional fitting algorithm and carried out with the design weight turned on, so the result of this weighting is post-stratification weights, which also take into account the design weights. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - SLOVENIA (2022) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2022 Slovenian parliamentary election took place on Sunday, April 24, 2022. The post-election survey, including CSES Module 6, was administered between September 26 and December 3, 2022 (lasting 61 days in total). Fieldwork was carried out by the Public Opinion and Mass Communication Research Center of the University of Ljubljana, Faculty of Social Sciences. Data collection was self-administered using a mixed-mode design combining online and paper (mail-back) questionnaires, both available from the start of fieldwork. The survey was conducted exclusively in Slovenian, with translation and pre-testing completed by the research team. The total number of completed interviews is 855. The study is designed to be representative of the population of the Republic of Slovenia aged 18 years and older. The sampling frame was the Central Population Register of Slovenia, from which a systematic random probability sample of 3,000 individuals was drawn. The sample was stratified by 12 statistical regions and 5 types of settlements, ensuring proportional coverage of the Slovenian population. Each person in the population had an equal probability of selection. To encourage participation, respondents received one invitation letter, followed by two reminders and a 5 EUR gift card incentive after completion. Both online and paper modes were available throughout the fieldwork period, and respondents could choose their preferred mode. Approximately 90.4% of interviews were completed online, and 9.6% via mail. The inclusion of the paper questionnaire was intended to minimize coverage error by allowing participation of individuals without Internet access (estimated at about 6% of the population). <<>> POLITY WEIGHTS The 2022 Slovenia Study provides a COMBINED SAMPLE and DEMOGRAPHIC WEIGHT. The weight was created due to partial discrepancies in the sample structure and the mode of data collection (self-administered online survey). The weight corrects for potential deviations in the results arising from differential response rates across subpopulations. A post-stratification adjustment was applied iteratively to align the sample distribution with population benchmarks derived from official statistical sources (the sampling frame). The weight is based on a combination of variables: Gender (two categories) x Age (four categories), and Region (twelve categories) x Settlement type (five categories). The resulting weight is included in variables F1101_1 and F1101_2. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - SWEDEN (2022) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2022 Swedish lower house election took place on Sunday, September 11, 2022. The study was fielded between September 13, 2022, and January 4, 2023 (a total of 114 days) by Statistics Sweden (SCB), employing a mixed-mode approach: Respondents were initially contacted via postal mail and received a postcard notification about the upcoming survey. A week later, respondents were contacted again and sent the paper mail-back survey as well as a QR-code taking them to the online interview. Hence, respondents could self-select into interview modes. The total number of observations from the Sweden Study is 2,845 interviews: 2,187 via self-completion mail-back and 658 via the Internet. The sample was conceptualized as a simple random sample and is meant to be representative of Swedish citizens living in Sweden who were entitled to vote and aged 18 years or older. Eligible Swedish citizens living abroad were excluded from the sample frame (about 2% of the voting-eligible population), as were citizens with protected identity (<1%). Individuals were sampled directly from population registers provided by Statistics Sweden and were hence the primary sampling unit. To encourage participation, sampled respondents were offered a lottery ticket worth 30 SEK upon completion of the survey. <<>> POLITY WEIGHTS The 2022 Sweden Study provides a DEMOGRAPHIC WEIGHT. Weights were calculated using an iterative proportional fitting algorithm (raking). The demographic weight (F1101_2) adjusts the data to match known demographic characteristics, namely gender (two categories), age groups (four categories), and education (seven categories). As data was missing for education for 16 respondents in the survey, the deposited weight was coded missing for these observations as well. For CSES, the 16 respective cases were set to the average value of F1101_2, i.e., 1. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN & WEIGHTS - SWITZERLAND (2023) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2023 Swiss Federal elections were held on Sunday, October 22, 2023. Fieldwork for the Swiss election study began on October 23, 2023, and was completed on January 11, 2024 (81 days in total). The fieldwork was administered by FORS - the Swiss Center for Expertise in the Social Sciences. The total number of observations from the Switzerland Study is 5,033 interviews. The study employed a mixed-mode approach with interviews conducted via the Internet and mail-back self-completion. All selected respondents were first invited to complete the survey via the Internet. With the second reminder (out of three), a paper version of the questionnaire was sent to those who had not yet participated in the survey. Hence, all respondents who completed the survey after the reception of the second reminder self-selected into the mode of interview. The survey was administered in French, German, and Italian. The sample was designed to be representative of all Swiss citizens residing in Switzerland who were at least 18 years old on Election Day (October 22, 2023) and thus eligible to vote. Only Swiss citizens living abroad were not included in the sample (about 4% of potential voters). The sample was drawn by the Federal Statistical Office (FSO) from their sampling frame of Swiss citizens who were at least 18 years old and residing in Switzerland at the time of the elections, with individuals as primary sampling units. With the use of stratification by canton (electoral district), the sample size was adjusted to have a minimum of at least 80 respondents per canton in each of the 26 cantons (which implied an oversampling of small cantons). The oversampling aims to obtain a larger sample of individuals in two cantons with important regional parties, Geneva and Ticino, to facilitate analyses at the cantonal level. To maximize the response rates, a cash incentive of a 10 CHF bill was included in the invitation letter. <<>> POLITY WEIGHTS The 2023 Switzerland Study provides a SAMPLE WEIGHT and a COMBINED SAMPLE and POLITICAL WEIGHT. The Sample Weight (F1101_1) corrects for the oversampling of small cantons as well as the cantons of Geneva, Ticino and Zurich. This weight was calculated as the quotient of the proportion of eligible voters in the population of each canton and the proportion of respondents in the sample that falls within the canton. The Political Weight (F1101_3) corrects biases that result from oversampling, turnout, and party choice in the data. The weight is calculated by multiplying three weights, namely the Design, Turnout, and Party Choice Weights. The Design Weight corresponds to F1101_1. The Turnout Weight corrects for respondents' overreporting of participation in the federal elections (according to the official election turnout), whereas the Party Choice Weights corrects for party choice biases in the survey (according to the official election results). --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - TAIWAN (2024) --------------------------------------------------------------------------- <<>> STUDY DESIGN General elections in Taiwan were held on January 13, 2024. The post-election survey was fielded between January 15 and May 31, 2024 (138 days in total). The data was collected by the Election Study Center at National Chengchi University. The survey was conducted face-to-face and administered in Chinese. The total number of observations from the Taiwan Study is 1,206 interviews. The sample is meant to be representative of Taiwanese citizens with household registration in Taiwan and a minimum age of 20 on election day, i.e., voting age. Convicted criminals deprived of civil rights were excluded from the sampling frame, as they are ineligible to vote. The survey employed a multi-stage stratified random sample design. To ensure that sampling proportions align with actual population distributions, the sample was stratified by Taiwan's major geographical regions in the first stage (such as northern, central, southern, and eastern regions), followed by a selection of electoral districts by the probabilities proportional to size principle (PPS). In the second stage, villages or neighborhoods were selected randomly from within the electoral districts in each region. The third stage involved address-based sampling of households based on a geographical information system (GIS). The official village, street, and household address database maintained by the Department of Household Registration, Ministry of the Interior, R.O.C., served as the basis for address sampling, ensuring that all addresses had a chance of being selected. Finally, respondents were selected within households based on pre-defined random selection rules, namely the last-two-digit address method. That method first lists the number of voting-eligible respondents in the household. Then, using the last two digits of the address number, interviewers referenced a table developed by scholars based on official government statistics to identify the designated respondent - for example, the youngest eligible person in the household, the oldest, or another household member as determined by the sampling table. To boost response rates, respondents received a convenience store gift voucher worth 250 TWD after finishing the interview. Researchers are advised that most of the original survey questions included additional missing value categories not envisaged by CSES, such as "It's hard to say", "No opinion", or "It depends". As CSES does not envisage these codes, respondents selecting one of these answer options were coded to missing for the respective CSES item. <<>> POLITY WEIGHTS The 2024 Taiwan Study provides a DEMOGRAPHIC WEIGHT, whose purpose is to adjust the sample to the general population of eligible voters in the 2024 Presidential and legislative elections. Weights were created by raking based on gender, age (five groups), education (five groups), and area (six regions). The population parameters were adopted from the Statistical Yearbook of Interior published by the Ministry of the Interior, Republic of China, in 2023. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - TURKIYE (2023) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2023 Turkish general elections were held on Sunday, May 14, 2023 (first round) with a second round of Presidential elections held on Sunday, May 28, 2023. The 2023 Turkish election study is a post-election cross-sectional study, with fieldwork starting 103 days after the first round of elections. Fieldwork was conducted between August 25, 2023 and October 4, 2023, by Frekans Arastirma and the questionnaire was administered face-to-face in Turkish. The total number of observations from the Turkiye Study is 1,508. The sample is designed to be representative of the Turkish voting age population (i.e., citizens aged 18 years or older). Institutionalized persons in prisons or hospitals and military personnel were excluded from the sampling frame (estimated to be less than 3% of the voting-eligible population). Further, the sample excludes the NUTS-II region TR63 covering areas in the provinces of Hatay, Kahramanmaras, and Osmaniye (about 4% of the total eligible population). Collaborators note the region was excluded due to the earthquakes on February 6, 2023, which caused major devastation in the area, impeding proper sampling and interviewing. The sample was realized as a stratified multi-stage random sample. NUTS-2 regions, as given by the Turkish Statistical Institute (TUIK), formed the basis for stratification. In the first step, the target sample of 1,500 envisaged interviews was stratified according to each region's share of the urban and rural population as published by the Supreme Electoral Council following the Address-Based Population Registration System (ADNKS) records. After determining the desired number of interviews, collaborators requested household addresses from the Turkish Statistical Institute with a block size of 150. Ten voting-eligible respondents were targeted from each block, and no substitution was used. The probability-proportionate-to-size (PPS) principle was used in distributing the address blocks to the NUTS-2 regions. Respondents within visited household addresses were selected according to a lottery method and based on the reported number of eligible individuals in each household. If, for any reason, the selected individual could not be interviewed on the first visit, the same household was visited up to three times until the interview was conducted successfully. <<>> POLITY WEIGHTS The 2023 Turkiye Study includes a DEMOGRAPHIC WEIGHT. The weight adjusts the sample to the following known population distributions of gender, age, and educational attainment: Males and Females, eleven age groups (ranging from 18-21 to 65+) and six categories for education (ranging from "no formal education" to "college graduate"). The resulting 132 cells (2 x 11 x 6) were filled with the respective respondents from within the sample. Afterward, the resulting distributions were compared to the latest available 2022 population statistics from the Turkish Statistical Institute. Demographic weights were then created based on identified differences across population and sample distributions. --------------------------------------------------------------------------- >>> OVERVIEW OF STUDY DESIGN AND WEIGHTS - UNITED STATES (2024) --------------------------------------------------------------------------- <<>> STUDY DESIGN The 2024 United States general election was held on Tuesday, November 5, 2024. The 2024 study - part of the American National Election Study (ANES) 2024 Time Series - used a mixed-mode design that included in-person, online, video, phone, and paper-and-pencil interviews. Further, the ANES 2024 Time Series study has a pre- and post-election panel design in which respondents were interviewed twice: once before the election and once after. Spouses or partners of respondents in the fresh samples were invited to complete a brief companion survey during the pre-election period. The study design comprises four independently drawn components: a fresh in-person sample, a fresh online sample, a re-interview panel of 2016 and 2020 respondents, and a smaller re-interview group from the 2024 General Social Survey. The CSES sample includes respondents from the first three components (fresh in-person sample, N = 925; fresh online sample, N = 1,769; 2016-2020 re-interviews, N = 2,070) who participated in the 2024 pre- and post-election interviews. The total number of observations from the United States Study is 4,764 interviews. Fieldwork took place between August 3 and November 5, 2024 (pre-election); and November 7, 2024, and February 19, 2025 (post-election). Interviews were conducted in English and Spanish. Median interview length was about 93 minutes for in-person, 63 for web, and 72 for panel respondents; post- election interviews were slightly longer. The target population consisted of U.S. citizens aged 18 or older living in private households across the 50 states and Washington, D.C., excluding those in institutional or group quarters. The in-person sample covered the 48 contiguous states and D.C., while the online sample also included Alaska and Hawaii. The fresh in-person sample used a multi-stage stratified cluster design based on 60 primary sampling units (PSUs) drawn across the contiguous states and Washington, D.C. The PSUs were metropolitan statistical areas (MSAs) or non-MSA counties, or combinations of counties, to achieve a minimum population of 50,000 adult citizens. PSUs were stratified by quintiles of county-level Republican vote share and terciles of turnout in 2020 and by region, with Los Angeles, Cook, Harris, and Maricopa counties selected with certainty. Within each PSU, smaller areas (Census Block Groups) were drawn and stratified by home ownership, education (college graduates), and modeled Trump vote probability in 2020 using voter file data. The online sample was drawn in two phases from the U.S. Postal Service's address file, stratified by county-level 2020 Republican vote share (quintiles), turnout (terciles), and census regions (four), and later refined by predicted 2020 and 2022 address average turnout and Trump vote probabilities. Respondents in the in-person sample received a 5 USD cash advance and between 25 USD and 50 USD for participation, later increased to 100 USD or 150 USD. Web respondents received 10 USD upfront and 40 USD upon completion, with later increases to 100 USD. The same incentives were used for panel and spouse surveys. The overall post-election re-interview rate was 89.9% of pre-election respondents (88.1% in the fresh in-person component, 85.3% in the fresh online component, and 95.3% for the 2016-2020 panel group, respectively). The pre-election re-interview rate for the 2016-2020 panel cases was 82.0%. As post-election re-interviews were conducted with 95.3% of panelists, the cumulative 2024 re-interview rate was thus 78.2%. <<>> POLITY WEIGHTS The 2024 United States Study includes a COMBINED SAMPLE and DEMOGRAPHIC WEIGHT. The weight adjusts for unequal probability of selection, non- response, age by gender, marital status by gender, race/ethnicity by educational attainment, race/ethnicity by Census region, nation of birth, home tenure by metropolitan status, population density, household income, and early voter status. The weight deposited with CSES (variable V240108b in the ANES 2024 Time Series data) is targeted to the full post-election dataset, namely all four modes of data collection. Users interested in conducting analyses with a subset of these sample components only are referred to the ANES 2024 Time Series Study, which provides individual weights for selected components. /// END OF FILE