CBS News Public Pulse Data and Instant Polls of Undecided Voters for 2004 Presidential Debates 1 and 3 and Vice-Presidential Debate (ICPSR 4177)

Published: Aug 18, 2005

Principal Investigator(s):
Kathy Frankovic, CBS News


Version V1

The data were gathered through two different processes. The data in Parts 1, 3, and 5 were gathered by asking respondents questions concerning the 2004 United States Presidential Election and the debates between the presidential and vice-presidential candidates. Questions from Parts 1, 3, and 5 concerned opinions of presidential candidates Senator John F. Kerry (Democrat) and President George W. Bush (Republican) and vice-presidential candidates Senator John Edwards (Democrat) and Vice-President Dick Cheney (Republican) before and after their respective debates, opinions of each presidential and vice-presidential candidate's qualities and attributes, which candidate was more likely to win his debate, and which candidate won his debate. Further debate questions addressed the likelihood respondents were going to watch each debate, the level of influence and importance of the debates, and whether one candidate unfairly attacked his opponent in the debate. Also, respondents were asked for their opinion of George W. Bush's handling of the presidency, whether the country was going the right direction, the degree to which their opinion of the vice-president affected their vote for president, their opinion of the condition of the national economy, whether the candidates shared the moral values of many United States citizens, the likelihood the respondent was going to vote in the 2004 presidential election, and what mattered more in the presidential election: national security or the national economy. Pulse data (Parts 2, 4, and 6) were collected in order to obtain response time reactions to the candidates and their statements. They were obtained by panelists being instructed to move their cursor to the left and right to indicate how much they like or dislike the messages being delivered by each candidate throughout the debate. Background information for Parts 1, 3, and 5 includes age, education, employment status, head of household status, home ownership status, household income and whether the household was dual-income, housing type, labor union membership status, likelihood of voting in the 2004 election, marital status, military service status, number of children in the household, political party affiliation, political ideology, race, religious affiliation, sex, voter registration status, and whether the respondent voted in the 2000 United States presidential election.

Frankovic, Kathy. CBS News Public Pulse Data and Instant Polls of Undecided Voters for 2004 Presidential Debates 1 and 3 and Vice-Presidential Debate  . Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2005-08-18.

Export Citation:

  • RIS (generic format for RefWorks, EndNote, etc.)
  • EndNote

2004-09 -- 2004-10

2004-09-30 -- 2004-10-13

(1) In Parts 2, 4, and 6 the format of the TIME variable was changed from DATE (hh:mm:ss:ss) (F11.2) to Numeric (F11.2) for the sake of conversion between statistical packages. Users may want to convert the TIME variable back to its original date format. (2) The data in Parts 2, 4, and 6 were measured every second in order to get aggregate data concerning the degree to which the message was liked or disliked. Within eight minutes after each debate was over, data were collected on respondents' views of each candidate's performance during the debate.

A panel sample was used.

Registered voters who were undecided at the time of debate and would be very likely or somewhat likely to watch the debate.


survey data, and reaction data

web-based survey

web-based response instrument

Presidential Debate 1: 79 percent, Vice-Presidential Debate: 68 percent, Presidential Debate 3: 66 percent




  • Data in this collection are available only to users at ICPSR member institutions.

ICPSR logo

This study is provided by ICPSR. ICPSR provides leadership and training in data access, curation, and methods of analysis for a diverse and expanding social science research community.