Businessman Herman Cain enjoyed a boost in popularity after winning unscientific straw polls in Florida and Illinois. His legitimacy as a front runner has been further cemented by several polls conducted by news networks and polling agencies that suggest he is gaining steam among both conservative voters and the general population.
A pollconducted between October 3 and October 5 by IBOPE Zogby, brings apparent good news for Cain, and bad news for his Republican primary opponents and President Obama. Results show Cain with a 20 percentage point lead among likely Republican primary voters, and with a 2 percentage point lead over President Obama among a representative sample of the adult U.S. voters. Although this poll, combined with a general polling trend, indicates Cain is a serious contender for the Republican nomination, and possibly the presidency, some of the methods used to gather data may be misleading.
Although pollster John Zogby has in the past been criticized for pushing a conservative agenda, IBOPE Zogby International, which formed when Zogby International was acquired by the IBOPE group in March 2011, is an independent and non-partisan polling organization. No outside groups benefited from, sponsored or endorsed the Oct. 3 to Oct. 5, 2011 poll.
Was There Delivery Bias?
The poll data indicates the questions were presented to respondents online. IBOPE Zogby’s Frequently Asked Questions describes the organization’s polling process in more detail. IBOPE Zogby maintains a database of registered respondents that voluntarily sign up to be polled. Although respondents are recruited by email at random, the pool from which they are selected is limited since it only consists of those with internet access who have the know-how and desire to register for IBOPE Zogby’s panel. Data reported on IBOPE Zogby’s FAQ indicates 79 percent of adults use the internet, however this leaves more than 20 percent of adults without representation in these types of polls. Data included in the poll indicates results were weighted to better represent its target populations, but the National Council on Public Polls (NCPP) says the scientific quality of polls should be questioned when the sample consists of volunteers. Although it is impossible to obtain a perfect sample in polling, random telephone dialing of both home and cell phone numbers is the preferred method, since 95 percent of the adult population owns a telephone, and those contacted have not been pre-registered.
Zogby Poll Sample Analysis
IBOPE Zogby surveyed two sample groups for this poll. One group consisted of 1581 adult voters, and the other group consisted of 796 likely Republican primary voters. Respondents from both groups were invited to participate via email, and obliged. IBOPE Zogby did not release how many potential respondents were initially contacted, whether they recorded incomplete surveys, or whether duplicate submissions were prevented. IBOPE Zogby followed NCPP standards in recruiting samples consisting of adults who are likely voters, but poll data does not reveal the criteria used to define a likely voter. Although any attempt to predict whether the sample will turn out in elections is preferred over a generic sampling of adults, methods of selecting likely voters vary by poll, and are sometimes unreliable. The margin of error for the general sample was +/- 2.5 percent, and the margin of error for the smaller Republican primary voters sample was +/- 3.5 percent. These fall below 4 percent, which is the range defined by PBSas being consistent with that of most other national polls. However, the margin of error only accounts for the discrepancies that occur when attempting to approximate nationwide results based on a small sample. If members of the population are excluded from consideration due to narrow polling or their own refusal to cooperate, poll results may be more severely affected.
Wording of the Poll Questions
The Republican primary voter sample was asked one question, “If the Republican primary for president were held today, for whom would you vote?” The question was followed by a list of candidates. Cain was selected by 38 percent of the respondents, followed by Romney with 18 percent, Perry and Paul tied with 12 percent, Gingrich and Huntsman tied with 4 percent, Bachmann with 3 percent, Santorum with 1 percent, Johnson with less than 1 percent, “other” with less than 1 percent and “not sure” with 7 percent. There was no apparent bias in the way this question was worded or presented.
The sample of all likely voters was asked, in three separate questions, for whom they would vote if Cain, Romney or Perry were the nominee for President against Barack Obama. Cain received 46 percent favorability against Obama’s 44 percent, and Obama received 41 percent favorability against Romney’s 40 percent and 45 percent favorability against Perry’s 40 percent. While these questions themselves were not biased, other questions were presented to respondents including whether “the nation is on the right track or headed in the wrong direction” and “do you think President Obama deserves to be re-elected, or is it time for someone new?” These questions may have placed respondents in a negative mindset against Obama. According to the NPCC, questions such as these may skew results when asked prior to the questions on which polling percentages are based, however it is unclear where IBOPE Zogby placed these questions within the online survey.
Understanding the Poll Results
Herman Cain is a formidable opponent against his fellow Republican candidates and President Obama. Although this poll indicates a significant surge in his popularity, consider the degree to which his approval is increasing in the context of several polls that use a variety of samples and polling methods.
Allen, M. Zogby won’t duplicate Obama poll. Politico. Accessed October 9, 2011.
IBOPE Zogby International. IBOPE Zogby Poll: Cain Expands Lead Over GOP Field & Leads Obama, 46%-44%. Accessed October 9, 2011
IBOPE Zogby International. FAQ. Accessed October 9, 2011.
Decoding Science. One article at a time.