VAR CF0012 FORM FILTER

VAR CF0012    FORM FILTER
              COLUMNS 78-78
              NUMERIC
              MD EQ 0


        FORM OR INTERVIEW TYPE FILTER
        -------------------------------------------------------------------

        This variable is study specific.

        SEE ALSO VCF0012A, VCF0012B.

        SEE General Note, VCF0012A.

        The 1970 Ns below are of the unweighted cross-section.

        Note that in Pre-Post Studies 1972 and 1988, question content was
        planned across both waves, so that the same Form type would be
        assigned in the Post as had been administered in the Pre (e.g. a
        respondent who had been interviewed with Pre Form A was automatically
        assigned to Post Form A).

        1986: In all instances of Form A/ Form B question designation in 1986,
        questions identified as "Form A" were not asked of Form B respondents
        and vice versa, with two exceptions.  In 1986, both Form A and Form B
        respondents were asked the question in Cumulative Data File var
        VCF0830, however the wording differed by form type.  VCF0012 may be
        used to filter for Form A/Form B in order to obtain data for each
        form independently from VCF0830.

        1988: Unlike other years, in 1988 all questions were present in both
        Form A and Form B, however some questions differed in wording or
        placement according to whether the form administered was A or B.  1988
        questions appearing in the Cumulative Data File which varied in
        wording according to form type are flagged in this codebook.  In such
        instances, filter VCF0012 may be used to filter for Form A/Form B in
        order to obtain data for each independently; otherwise, Form A and
        Form B data are combined within variables.  1988 vars affected are:
        VCF0517-VCF0518, VCF0830, VCF9076, VCF9084, VCF9092.

        1994: In 1994, the study included a 'ballot card' experiment.  Two
        ballot cards were drafted with identical content but different layout.
        The variable for form of interview was used exclusively to assign
        respondents to one design or the other for the vote question.

        1996 and later: Due to the introduction of CAPI administration, this
        variable no longer applies.  See VCF0012A and VCF0012B.

         1970:  700003         1972:  700020         1986:  860046
          1.  Type 1 (N=758)    1.  Form I*  (N=1372)  1.  Form A (N=1086)
          2.  Type 2 (N=749)    2.  Form II* (N=1333)  2.  Form B (N=1090)

         1988:  880040         1990:  900061         1994:  940611
          1.  Form A (N=1006)   1.  Form A (N=1007)    1.  Form A ballot card
          2.  Form B (N=1034)   2.  Form B (N= 993)    2.  Form B ballot card
                                                       9.  NA
         other years:
          0. Not applicable to this study

        * note:  for the 1972 post-election wave, 57 Rs who were Form I in the
        pre and 37 Rs who were Form II in the pre were sent post mail
        questionnaires that contained most, but not all, of the questions
        contained in Form I and Form II post personal questionnaires.  In the
        1972 NES codebook, for variables which were present in the post
        personal questionnaire but not in the post mail questionnaire:   Rs
        who were Form I in the pre but completed mail questionnaires for the
        post were described as  "FORM III"  in INAP codes, while Rs who were
        Form II in the pre but completed mail questionnaires for the post were
        described as  "FORM IV" in INAP codes.  These INAP descriptions have
        been retained in INAP codes in the Cumulative Data File codebook for
        corresponding vars.    SEE VCF0015.


==============================