Updates & Errata




Cross-Study ID File

1956-1960 Panel Study

About the Dataset

  • Type of Study: ANES panel study
  • Sample universe: eligible voters in U.S. households
  • Sample composition: panel-only/panel-and-Time-Series cases
  • Sample selection: area probability
  • Number of waves: 5 total
  • Completions: per component year
  • Modes used: face-to-face
  • Instrument format: paper questionnaire
  • Weights: V600569
  • Auxiliary files: none

Study Content Highlights

Panel respondents were asked the same questions documented for the ANES 1956 Time Series Study, the ANES 1958 Time Series Study, and the ANES 1960 Time Series Study.

Study Design Highlights

This file provides a panel of 1,132 respondents who were interviewed in 1956, 1958, and 1960 with an interview status defined each year as either ‘panel-only’ or ‘panel and Time Series’. The portion of respondents defined each year as ‘panel and Time Series’ were also considered members of the cross-section for the same year’s ANES Time Series study. Among the 1,132 total cases of the 1956-1958-1960 panel, 1,096 are also 1956 Time Series cross-section cases, 895 are also included in the 1958 Time Series cross-section, 779 are also members of the 1960 Time Series cross-section, and 752 cases can be found in all three Time Series cross-sections. Selection variables are included in this release to select by participation each year, as panel-only or as panel-and-Time-Series (V600841-V600843).

Respondents in this file were interviewed in 1960 and at least once previously. In addition to the 1956-1958-1960 panel, there are 275 cases of panel respondents who were interviewed (either as panel-only or as panel-and-Time-Series) in 1958 and 1960 but not in 1956, and there are 107 cases of respondents (either panel-only or panel-and-Time-Series) who were interviewed in 1956 and 1960 but not in 1958. Respondents with sample membership during years 1956-1960 confined to a single Time Series study’s cross-section are not included in this file but are found only in the data of the Time Series study (see highlighted rows below).

1956-1960 participation of respondents is summarized below.

Row Count Panel Study 1956 status 1958 status 1960 status 1956 TS 1958 TS 1960 TS
184 TS only 184
110 110 TS and Panel Panel only** 110
165 165 TS and Panel TS and Panel 165 165
280 TS only 280
4 4 Panel only* Panel only**
6 6 Panel only* TS and Panel 6
7 7 Panel only* Panel only** Panel only
7 7 Panel only* TS and Panel Panel only**
22 22 Panel only* TS and Panel TS and Panel 22 22
50 50 TS and Panel Panel only** 50
47 47 TS and Panel TS and Panel 47 47
225 225 TS and Panel Panel only** Panel only 225
5 5 TS and Panel Panel only** TS and Panel*** 5 5
114 114 TS and Panel TS and Panel Panel only** 114 114
752 752 TS and Panel TS and Panel TS and Panel 752 752 752
569 1956 TS only 569
1514 1762 1450 1181

TS=Time Series
* pre-only or post-only in 1956
** R had moved since preceding interview
*** it is undocumented why these cases were restored to the Time Series in 1960
Note: for the 1956-1960 panel, variables for 1956 are named exactly as they are named for the 1956 Time Series study (V560001-V560298), however names for 1958 and 1960 variables are ‘numbered’ differently than for the 1958 and 1960 Time Series studies; 1958 variables are V580299-V580566 and 1960 variables are V600567-V600840.

For 1956, only those respondents who had provided both pre-election and post-election interviews were included in the 1956 Time Series study release. In this file, there are also 1956 interview data for 44 pre-only and 2 post-only cases which are considered ‘1956 panel-only.’ For the 1958 Time Series study, respondents who had previously been interviewed in 1956 (even as 1956 panel-only cases) were eligible for inclusion in the 1958 Time Series cross-section if they had not moved, while movers acquired ‘1958 panel-only’ status; new cases for the 1958 Time Series study were drawn among current dwellers in households vacated by 1956 respondents. There are 237 interviews from 1958 in this file which are not present in the 1958 Time Series release. In 1960, all 1956 respondents and all 1958 cross-section respondents were eligible for panel re-interviewing (movers followed again), however for the 1960 Time Series study only non-movers among these same groups were eligible for cross-section inclusion. The 1960 Time Series study also included as fresh cases 166 respondents selected from newly constructed sample dwelling units or from 1956 addresses.

Weights Summary

Weight V600569 compensates for 1) differential sampling rates used in 1958 and 2) panel attrition. In 1958, the rate of selection for respondents first interviewed in 1958 was half the basic rate of selection for respondents interviewed at their 1956 addresses in both 1956 and 1958. For the set of respondents lost to the panel through attrition, 100 panel respondents interviewed in all years (1956, 1958, and 1960) were selected for weight value 2 or greater.

Auxiliary Files

No auxiliary file is present for this release.

Associated Derivative Files:

An ANES derivative file is a file which merges together related ANES studies. Within the ANES archive, study data cases may be represented in no more than 1 derivative file, in addition to its presence as a stand-alone study release.

No ANES derivative file release is associated with this study.