Results of the Medill News Service/Ipsos survey are based on online interviews conducted November 7-19, 2012 among a sample of 1,686 American citizens aged 18 and older residing in the United States, including 516 who voted in the recent Presidential election and 1,170 who did not vote.
For the sample, Ipsos utilized its Cortex and Ampario systems. Cortex is Ipsos’ field management platform that automates and simplifies online data collection. While Cortex is mostly known for routing (i.e., allocating in real-time the right respondent to the right study), it also facilitates sample blending. In addition to Ipsos’ own panel consisting of nearly 1 million active members, Cortex combs over 45 sample sources, including other approved panels, social networks, and reward systems. Ampario is Ipsos’ arm that reaches out to “non-panelists.” Ampario connects to people within social networks, applications, ad networks and other publishers traditionally outside of the research realm, using surveys as a means of monetization. The incentive for our Ampario sample is a virtual currency system as a reward for completing surveys.
The combination of a blended population sample with Cortex and the Ampario sample allows Ipsos to create a perpetual sample flow that “pulls” from the general population in large numbers and efficiently directs the right respondents to the right surveys. In addition, Ipsos has a standard quality practice to ensure the identity of respondents. This practice includes de-duplication with existing panel and recruitment campaigns based on email, address, name, surname, etc., along with de-duplication with an Ipsos blacklist (including emails of clients, competitors, and Ipsos employees). The de-duplication process also includes using RELEVANTID, a digital fingerprinting technology. Please let us know if you would like even more detail regarding Ipsos’ quality assurance practices.
The sample outgo was balanced by a number of demographic variables, including age, gender, and region. Each eligible respondent was asked whether they were an American citizen and if they voted in the 2012 Presidential election. Those who said that they did vote in the election were classified as “voters;” those who indicated that “something prevented them from voting” or that they “chose not to vote” were classified as “non-voters.”
The questionnaire was largely the sample for both target audiences, though the non-voters were asked some additional questions focusing on the reasons why they did not vote. Both groups were asked about their past voting behavior; attitudes about voting, the voting system, and the role of government; political and civic engagement; news consumption; knowledge and understanding of current events and candidates’ policy positions; and a variety of demographic questions.
Once data collection was complete, the data were weighted by gender, age and ethnicity to the post-election voter profile based on data from the Current Population Survey (CPS).
To facilitate out exploration of both the similarities and differences among non-voters, a typology was constructed by using a cluster analysis, a statistical technique that classified respondents into the most homogeneous and meaningful groups possible based on their reasons for not voting; past voting behavior; attitudes about voting, the voting system, and the role of government; political and civic engagement; and news consumption.
Statistical margins of error are not applicable to online polls. The precision of the Ipsos online polls is measured using a credibility interval. In this case, the poll has a credibility interval of plus or minus 4.9 percentage points for voters and 3.3 percentage points for non-voters.
All sample surveys and polls may be subject to other sources of error, including, but not limited to coverage error and measurement error. Figures marked by an asterisk (*) indicate a percentage value of greater than zero but less than one half of one per cent. Where figures do not sum to 100, this is due to the effects of rounding.