ANES logo TIME SERIES CUMULATIVE DATA FILE
line3

ERRATA

Description: Miscellaneous corrections
Posted:2006-Oct-17

line3

1. Cases in VCF0127 (Household Union Membership) that are from 2004 and coded 5 should have been coded 2 in VCF0127. In the codebook documentation for VCF0127, the 2004 source variable was also omitted: V043290.

2. The codebook documentation "General Note" for VCF0152 should read as follows:
     GENERAL NOTE:
     Built from VCF0152, respondent categorization of type of
     Christianity, church attendance, and Christian born-again
     status.

3. Thermometer variables with the codebook documentation cross-refereence "see VCF0201 note" for code DK are incorrect and should refer to the annotation accompanying the introductory text preceding VCF0201.

4. Variable VCF0118 has 2004 data incorrectly coded. The following recode will correct in SPSS:
         compute VCF0118 = V043260b.
         if V043260b = 4 VCF0118 = 2.
         if V043260b = 6 VCF0118 = 2.
         if V043260b = 5 VCF0118 = 3.
         if V043260b = 7 VCF0118 = 4.
         if V043260b = 8 VCF0118 = 5.
         execute.

5. For 1990 cases, VCF0702 should be recoded as follows (SPSS):
         Do if (VCF0004 = 1990).
         Recode VCF0702 (1=2) (2=1).
         End if.

6. The CDF codebook appendix "1972-LATER PARTY-CANDIDATE MASTER CODE" correctly documents code 42 for Robert (Bob) Dole except in 1976, when Dole mentions are instead coded 37. Code 37 mentions from 1976 in relevant CDF likes-dislikes variables should be recoded to code 42 only if consistency is desired (note: code 42 does not otherwise apply in 1976).

7. Codebook documentation for income variable VCF0114 does include income ranges from 2004. They are as follows:
   1. 0-16,999
   2. 17,000-34,999
   3. 35,000-69,999
   4. 70,000-119,000
   5. 120,000+

8. For VCF0106a, the code value label in the release statement files for code 0 should be: "0. DK; NA; no Pre IW; short-form new Cross Section or Spanish language (1992)".
In addition, VCF0106a code 0 should be assigned to missing data in release statement files.

9. 7 cases from 2004 that are coded 0 in VCF0106 should have been coded 9; 170 cases coded 7 should have been coded 3.
   SPSS:
         recode VCF0106 (0=9)) (7=3).

10. Some cases from the 2000 were coded to regional south rather than political south in the CDF variable, VCF0113. The cases are in DE,KY,MD,OK,DC,WV.
   SPSS:
         if VCF0004 = 2000 and VCF0901a = 10 VCF0113 = 2.
         if VCF0004 = 2000 and VCF0901a = 21 VCF0113 = 2.
         if VCF0004 = 2000 and VCF0901a = 24 VCF0113 = 2.
         if VCF0004 = 2000 and VCF0901a = 40 VCF0113 = 2.
         if VCF0004 = 2000 and VCF0901a = 11 VCF0113 = 2.
         if VCF0004 = 2000 and VCF0901a = 54 VCF0113 = 2.

divider