# The 2016 ANES Time Series Study with pre-election interview A subset of 2,188 participants of the 2016 American National Election Time Series Study, which was to track the enduring social trend and record the political moment of 2016 (DeBell, 2018). This study consisted of two surveys with same population. The pre-election interview was during the weeks before the 2016 general election, including 4,271 respondents in total. The post-election interview is the re-interview during the weeks after the election, including 3,649 respondents (662 respondents did not complete post-interviews). data ## Format A data frame with 2188 rows and 10 variables. * `age` Respondent's age in years. * `edu.year` Respondent's education year, which is mapped from `education` level: `MS`=8, `HSdrop`=11, `HS`=12, `Coll`=14, `CCdeg`=15, `BAdeg`=17, `MAdeg`=19. * `education` Respondent's education level. * `income.num` Respondent's family income in thousands: an numerical variable. It is median value of the range of each `income` level * `income` Respondent's family income level: '(01) 01. Under $5,000' = 5, '(02) 02. $5,000-$9,999', '(03) 03. $10,000-$12,499', '(04) 04. $12,500-$14,999', '(05) 05. $15,000-$17,499', '(06) 06. $17,500-$19,999', '(07) 07. $20,000-$22,499', '(08) 08. $22,500-$24,999', '(09) 09. $25,000-$27,499', '(10) 10. $27,500-$29,999', '(11) 11. $30,000-$34,999', '(12) 12. $35,000-$39,999', '(13) 13. $40,000-$44,999', '(14) 14. $45,000-$49,999', '(15) 15. $50,000-$54,999', '(16) 16. $55,000-$59,999', '(17) 17. $60,000-$64,999', '(18) 18. $65,000-$69,999', '(19) 19. $70,000-$74,999', '(20) 20. $75,000-$79,999', '(21) 21. $80,000-$89,999', '(22) 22. $90,000-$99,999', '(23) 23. $100,000-$109,999', '(24) 24. $110,000-$124,999', '(25) 25. $125,000-$149,999', '(26) 26. $150,000-$174,999', '(27) 27. $175,000-$249,999', '(28) 28. $250,000 or more' = 250. * `PID` Party identification: a numeric variable with value from 1 to 7 representing strong Democrat, `strDem` \< weak Democrat, `weakDem` \< independent Democrat, `indDem` \< independent independent `indind` \< independent Republican, `indRep` \< weak Republican, `weakRep` \< strong Republican, `strRep`. * `selfLR` The respondent' self-placement about own left-right in 7 ordinal levels (from extremely liberal to extremely conservative). `extLib`: extremely liberal, `Lib`: liberal, `sliLib`: slightly liberal, `Mod`: moderate, `sliCon`: slightly conservative, `Con`: conservative, `extCon`: extremely conservative. `extLib` \< `Lib` \< `sliLib` \< `Mod` \< `sliCon` \< `Con` \< `extCon`. * `TrumpLR` The respondent's opinion about Donald Trump's left-right placement (same scale as selfLR). * `ClinLR` The respondent's opinion about Hilary Clinton's left-right placement (same scale as selfLR). * `PreVote` The respondent's voting preference between Donald Trump and Hilary Clinton two months preceeding the November election (Pre-election interview). It is a factor with levels `HillaryClinton` and `DonaldTrump`. * `PreVote.num` Recode the PreVote to numeric values, 'HillaryClinton'=0, 'DonaldTrump'=1. * `WeightforPreVote` Pre-election weight of a respondent. ```r data(ANES2016) ``` ## Details The Pre-election preference is recorded as "PreVote" and the "PreVote.num" is the numeric of it. Observations with missing values, or "No thought" responses have been removed. Respondents expressing a voting preference other than Clinton or Trump have been removed. ## Examples ```r head(ANES2016) ``` ## References DeBell, Matthew, Jon A. Krosnick, Katie Gera, David S. Yeager, and Michael P. McDonald. The turnout gap in surveys: Explanations and solutions. **Sociological Methods & Research**, 2018. tools:::Rd_expr_doi("10.1177/0049124118769085") Enamorado, T., Fifield, B., & Imai, K. (2018). **User’s guide and codebook for the ANES 2016 time series voter validation supplemental data**. Technical Report, American National Election Studies.

ANES2016 dataset