Congressional districts (N=108) as well as individuals were sampled. Contextual data about these congressional districts and the election campaigns therein are included in the dataset.

Introduced greatly expanded Congressional context, most of which became "core" items repeated in each subsequent survey. Major emphasis was placed on the respondents' evaluations of their district's candidates along several dimensions; and on how much and what kinds of contact Rs had with candidates. Specific candidate/incumbent names are associated with each congressional district. Voter validation variables are included.

STUDY CONTENT HIGHLIGHTS:
In addition to content on electoral participation, voting behavior, and public opinion, the 1978 ANES contains questions in other areas such as media exposure and media evaluation. New content includes new batteries on candidate recall and on contact with House incumbents and candidates, and the first appearance of Congressional candidate feeling thermometers, the open-ended Congressional candidate 'likes' and 'dislikes' questions, and party thermometers.

STUDY DESIGN HIGHLIGHTS:
In some respects, the 1978 ANES was a watershed study, introducing much content subsequently considered core to the time series. Congressional context was greatly expanded, emphasizing respondent evaluations of candidates in the areas of issues, performance, behavior, and ability. Congressional-level questions which specified district-appropriate names of incumbents and candidates appeared for the first time. The ballot card was also introduced.

In the sample design for the 1978 ANES, 108 congressional districts constituted the primary areas. The sample was designed to yield estimates for congressional districts, however the sample was not intended to produce estimates of each district's constituency. As in all time series studies, the 1978 ANES sample was drawn to represent a national cross-section. Administration of the post-election interview began immediately after the November 7 elections, and continued until January 25, 1979. The average interview length was 73 minutes, and 2,304 completions were yielded from 3,343 eligible in the sample.