About 61 percent of Iowa registered voters, including overwhelming majorities of Democratic and no-party voters, support same-sex marriage rights, according to a new survey conducted by graduate students at Drake University. The same survey indicated that 62.5 percent of Iowa voters oppose a constitutional amendment that would ban same-sex marriage. Those are the highest levels of support for marriage equality I’ve seen in any Iowa polling. I am skeptical that an opinion poll conducted in a more traditional way would find similar numbers.

More results from the online survey and details about its methodology are after the jump.

The Iowa LGBT Marriage Study 2011 was a research project for graduate students in the master’s of public administration program at Drake this year. Summary findings are posted here, and full results including question wording and order are here (pdf). You can download the full report here (pdf). Pages 19 through 22 explain how researchers collected and weighted responses, so that the adjusted results corresponded to the demographics of Iowans who were registered to vote in November 2010 and remained registered to vote in June 2011. The web-based survey was open to the public for 12 days, during which 2,217 people attempted to fill it out. The researchers used 1,780 responses to tabulate results (excluding respondents who did not complete the survey, or don’t live in Iowa, or didn’t live in Iowa when the Supreme Court announced its Varnum v Brien decision in April 2009).

Originally, there was variance between the age group, gender, and political party affiliation distribution for the survey respondents and the standard population. Younger females who aligned with liberal beliefs were the most frequent responders (Table 1). However, the large sample size of this study resulted in a significant number of no-party or independent and conservative responses. After weighting, the distribution of the weighted variables was similar to the standard population.

The weighted population was 47.4 percent male and 52.6 percent female, containing 33 percent self-identified “liberals,” 30.7 percent self-identified “conservatives” and 36.4 percent self-identified “independents.” Pages 23 through 26 of the full report include more demographic information about the survey respondents’ geographic distribution, race, ethnicity, education level, marital status, religion.

Asked “Should same-sex couples be legally allowed to marry in Iowa?”, the weighted results indicated that 61.3 percent said yes, 34 percent said no and 4.7 percent were not sure. Note: this survey did not give respondents the option of supporting some legal recognition (like civil unions), but not full marriage rights, for same-sex couples. Several Iowa polls conducted during the past two years have shown majority support for some legal recognition of same-sex relationships, but less than 50 percent support for full marriage equality.

For reasons I don’t understand, the researchers broke down responses on this question by “political party affiliation.” They found that among what they call “Democrats” in Figure 4 (page 28), 97 percent said same-sex couples should be allowed to marry, while only 1.3 percent said no and 1.7 percent were not sure. Among “independents,” 69.1 percent supported same-sex marriage rights, 22.3 percent said no and 8.6 percent were unsure. Among “Republicans,” just 13.3 percent said same-sex couples should be allowed to marry, 83.3 percent said no and 3.5 percent were unsure.

I would take those results with a big grain of salt. As far as I can tell from reading the questionnaire, this survey did not ask respondents whether they are registered Democrats, Republicans or no-party voters. Instead, researchers adjusted the results so that the proportion of self-identified “liberals” completing the survey corresponded to the proportion of Democrats in Iowa’s registered voter population. Similarly, those who identified themselves as “conservatives” in the survey were considered to be Republicans, and those who said they were “independents” were considered to be no-party voters.

But not all Iowa Democrats would call themselves liberals–far from it. I would expect nearly all self-identified liberals to say yes to marriage equality, but I doubt that 97 percent of registered Iowa Democrats support same-sex marriage rights.

I am puzzled by the Drake researchers’ decision to conflate “liberals” with Democrats, “conservatives” with Republicans and “independents” with no-party voters. They do the same thing reporting results for other questions: about support for a constitutional amendment to ban same-sex marriage, about the impact same-sex marriage has had on the people of Iowa; about how the average Iowan views same-sex marriage.

The Des Moines Register conducted a statewide poll using traditional telephone interviews between June 19 and 22. So far the newspaper has reported only some results for the likely Republican caucus-goer subgroup, but a much larger sample of registered Iowa voters was surveyed. In the past two years, each statewide poll commissioned by the Register has included a few questions about marriage equality, so probably new findings will come out in the next couple of weeks. I will be interested to see how the Register poll’s numbers compare to this online survey. The last Register poll, conducted in February 2011, found Iowans “split three ways” on the Iowa Supreme Court’s marriage ruling: 32 percent of respondents favored or strongly favored the controversial ruling, 37 percent opposed or strongly opposed it and 30 percent said they don’t care much one way or the other. That was a shift from the Register’s September 2009 poll, which found 26 percent of respondents favored the Supreme Court ruling on marriage, 43 percent opposed it and 31 percent didn’t care or were unsure.

Share any relevant thoughts in this thread.

UPDATE: A large sample size does not automatically produce a low margin of error for a survey. That would be the case only if the researchers used random sampling methods to reach the large group of respondents. This survey was not sent to a randomly-selected group of Iowans.