2024 Time Series: Updates & Errata
Current Dataset Release: May 19, 2026 version
The current release (May 19, 2026) is a re-release of the 2024 Time Series study dataset with some important updates:
- Corrected data due to misapplication of forward/reverse randomization in the following variables: V241352, V241353x, V241620, V242213, V242233, V242234x, V242243, V242244, V242245x, V242246, V242247, V242248x
- Miscellaneous data corrections in the following variables: V241106x, V241445x, V242051
- Missing data recode in variable V242020
- Responses incorrectly coded -5 (breakoff) for case id 141189 and have been recoded to -1 (inapplicable) in the post-election data
- Corrected value labels have been added to the following variable: V241004, V241287x, V241444x, V241445x, V242002, V242003, V242364x, and
V242509a to V242509f - Unrestricted versions of variables V242530 to V242535 (IDENT) have been merged into the main dataset after having previously been available as a standalone data file
- Data from the variables originating from the WEBSITES and INETNEWS items in the instrument have been temporarily removed to investigate a technical error in the data
- Previously unreleased variables from the MIP submodule (coded responses for mention 1, mention 2, mention 3, and the most important problem) are now included in the dataset: V242165y1 to V242165y5, V242167y1 to V242167y5, V242169y1 to V242169y5, V242171y1 to V242171y5
Previous Dataset Release: August 8, 2025 version
April 20, 2026
We are currently investigating a potential issue with variable V242246, V242447 and V242448x; we will provide an update as soon as we have more information.
February 5, 2026
Nineteen cases were incorrectly coded in the religion summary variable V241445x. The (SPSS) syntax below can be used to recode those values. This variable will be updated in a forthcoming release of the 2024 Time Series Study.
do if V240001 = 140815.
compute V241445x = 4.
else if V240001 = 141130.
compute V241445x = 4.
else if V240001 = 142011.
compute V241445x = 4.
else if V240001 = 142036.
compute V241445x = 4.
else if V240001 = 142461.
compute V241445x = 4.
else if V240001 = 202542.
compute V241445x = 4.
else if V240001 = 203201.
compute V241445x = 4.
else if V240001 = 212289.
compute V241445x = 4.
else if V240001 = 218657.
compute V241445x = 4.
else if V240001 = 218970.
compute V241445x = 4.
else if V240001 = 222436.
compute V241445x = 4.
else if V240001 = 223601.
compute V241445x = 4.
else if V240001 = 235139.
compute V241445x = 4.
else if V240001 = 235273.
compute V241445x = 4.
else if V240001 = 237359.
compute V241445x = 4.
else if V240001 = 327462.
compute V241445x = 4.
else if V240001 = 344010.
compute V241445x = 4.
else if V240001 = 395901.
compute V241445x = 4.
else if V240001 = 141810.
compute V241445x = 4.
end if.
execute.
We are currently looking into a potential issue with V241445x; further details will be available once we conclude our investigation.
October 16, 2025
There was an error in the value labels for the nine category religion summary variable, V241445x. The value labels for this variable should be as follows:
-2. Missing: DK, RF, did not specify
-1. Inapplicable
1. Mainline Protestant
2. Evangelical Protestant
3. Black Protestant
4. Undifferentiated Protestant
5. Roman Catholic
6. Other Christian
7. Jewish
8. Other religion
9. Not religious
We would like to thank Professor Marty Cohen of James Madison University for bringing this to our attention.
We are currently looking into a possible issue with some of the variables from media source (MEDSRC). Pending the outcome of the investigation, revisions to variables V241605a to V241605t and V241606a V241606f may be forthcoming.
Variable V242051 contains a single case coded ‘5’. This was part of data correction for CaseID 141189 but the value recovered from the raw data was not recoded to the standard code of ‘2’ (for “No” responses) in ANES data. Users should recode this value to ‘2’ to match the coding scheme for V242051.
The interview length variables (V243055 for pre-election interview length in minutes, and V243064 for post-election length in minutes) were constructed by adding together individual session times for each survey, as respondents sometimes would complete a survey across multiple sessions. Analysts using these variables will notice some outliers where the survey lengths are either implausibly short or implausibly long and thus not believable. Collecting timing data is challenging and sometimes imperfect – these outliers are likely due to a combination of technical errors and other unusual situations that are unique to specific cases, and are not currently possible to resolve. Outliers for these variables have been left in place in the dataset, rather than deleting them, so that users can decide for themselves whether to exclude outliers or handle them otherwise in their analyses.
March 10, 2025
[2026-04-16: The most recent dataset has corrected this erratum, and thus the csv file download link below has been removed]
An error in the application of the randomization for two variables that were included in the preliminary release of the pre-election survey data (version: 20250219) has been discovered. Revised versions of both V241219 (SPLITTICKET) and V241407 (ISRPAL) are now available on the ANES website in a standalone CSV file. Also included in the new file is a revised version of the branching summary variable V241409x (V241407 + V241408). These can be merged into the preliminary pre-election data file using the ID variable V240001.
An additional error in the construction of variable V241385x (GAYMARR2, GAYMARR2STR) was identified. A corrected version of this variable is also included in the standalone CSV file.
We would like to thank Dr Laura Stoker of Stanford University and Dr Allen Downey of Olin College for bringing these issues to our attention.