Differences between revisions 1 and 3 (spanning 2 versions)
⇤ ← Revision 1 as of 2024-11-16 21:17:58 →
Size: 778
Editor: DominicRicottone
Comment: Initial commit
← Revision 3 as of 2024-11-16 21:54:28 → ⇥
Size: 810
Editor: DominicRicottone
Comment: Styling
Deletions are marked like this. Additions are marked like this.
Line 3: Line 3:
ANES time series studies are administered around each presidential election. They are all two-wave studies, corresponding to a ''pre-election'' interview and a ''post-election'' interview. '''ANES time series studies''' are administered around each presidential election. They are all two-wave studies, corresponding to a ''pre-election'' interview and a ''post-election'' interview.
Line 37: Line 37:
----



== Data ==

ANES Time Series

ANES time series studies are administered around each presidential election. They are all two-wave studies, corresponding to a pre-election interview and a post-election interview.

Contents

  1. ANES Time Series
    1. Weights
    2. Data

Weights

Declare the complex survey design like: 

proc surveyfreq;
  stratum STRATUM;
  cluster CLUSTER;
  weight WEIGHT;
  table VOTE PARTY VOTE*PARTY ;
run;

Or: 

svyset CLUSTER [pweight=WGT], strata(STRATUM) vce(linearize) singleunit(centered)

Until the 2006 pilot, clusters and strata were published as a single variable. This could be broken apart like: 

compute STRATUM = trunc(v040103/10).
compute CLUSTER = v040103 ‐ (10*STRATUM)

Data

CategoryRicottone

AmericanNationalElectionStudies/TimeSeries (last edited 2024-11-16 21:54:28 by DominicRicottone)