THE FACTS ABOUT INDIANA POLLING - Democratic Underground

LulaMay

(1000+ posts) Send PM | Profile | Ignore

Fri Apr-18-08 06:57 PM
Original message

THE FACTS ABOUT INDIANA POLLING

There's been a lot of misinformation at DU about the second survey usa poll done in Indiana that had Obama +5 compared to the one +16 for Hillary.

As pollster.com shows here in this statement from Survey USA, two completely different methodologies were used in the polling. I think the DU'ers need to learn the facts about why the polls were different.

http://www.pollster.com/blogs/poll_downs_centersurveyusa_ind.php

Update: Please note that unlike the survey conducted and released earlier this week by SurveyUSA, this poll survey used a different "sample frame" (one based on a registered voter list rather than random digit dialing). I asked SurveyUSA's Jay Leve to confirm and he sent along a detailed explanation (that appears after the jump) of all the methodological differences between the polls and what may or may not explain the differences in presidential preference on the two surveys.

Methodological differences between two different SurveyUSA polls in Indiana

SurveyUSA conducted two separate, independent, non-commingled election polls in the state of Indiana in the past week, one for SurveyUSA’s TV-news clients (the TV poll) and one for Indiana University Purdue University Fort Wayne (the University poll). On many questions, the two polls produce similar results. On one question, the two polls produce different results. It is not possible to know for certain if one of the two polls is more “right,” but it is possible to inventory the things that SurveyUSA knowingly did differently on the two surveys.

· The TV poll was conducted using “RDD” sample (Random Digit Dial, purchased from SSI of Fairfield CT), which is what SurveyUSA uses on many, but not all, of its statewide election polls.

· The University poll was conducted using “RBS” sample (Registration Based Sample, purchased from Aristotle, of Washington DC), which is what SurveyUSA uses on many, but not all, of its research for non-media clients.

· There are advantages to both sampling methods. Neither is always better. With RBS, a pollster begins with a list of known registered voters, so the likelihood of reaching a voter at any given phone number is higher. But a pollster may not be able to call a perfect cross-section of voters, since RBS suppliers do not have a working telephone number for every voter. In Indiana, 82% of voter records have a telephone associated with them, 18% do not. “Registered voters with available phone number” is the “universe” for the University poll. By contrast, RDD sample ensures that all have an equal opportunity for inclusion, but there is no guarantee that those you reach are in fact registered. Likelihood of reaching a voter is lower. Number of wasted phone calls is higher, since non-working phones, disconnected phones, fax machines and modems are included (by definition) in an RDD sample . “Likely Indiana residences” is the universe for the TV poll. Known cell phones are excluded from both samples. Listed business phones and institutional phones are excluded from the RDD sample.

· SurveyUSA has written about the differences between RDD vs RBS sample when polling on a contest that involves a white woman and a black man. That paper is available here:

· TV poll interviewing was conducted over 3 consecutive nights that included a weekend (Friday, Saturday, Sunday).

· University poll interviewing was conducted over 3 consecutive nights that included only weekdays (Monday, Tuesday, Wednesday).

· Respondents in the TV poll were told at the beginning that poll was sponsored by a TV news station. Respondents in the University poll were told at the beginning that the poll was sponsored by Indiana University Purdue University Fort Wayne. The actual poll questions and demographic questions for both surveys were voiced by the same professional announcer, and although some questions were common to both surveys, other questions were not.

· The TV questionnaire was shorter. A TV poll “likely voter” could have answered as few as 8 questions (of the 14 total questions) and be included.

· A University poll respondent needed to answer at least 26 questions (of the 51 total questions) to be included.

· The “who will you vote for in the Democratic primary for President of the united states” question was asked 3rd in the TV poll. The same question was asked 18th in the University poll.

· University poll respondents were asked if they had already voted. TV poll respondents were not.

· Before the University poll respondents were asked how they would vote in the Democratic Presidential Primary (in addition to confirming that the respondent was registered in Indiana and likely to vote in the general election in November), respondents were asked questions about …

a) whether Indiana was headed in the right direction,

b) their opinion of John McCain,

c) their opinion of Hillary Clinton,

d) their opinion of Barack Obama,

e) their opinion of President Bush,

f) their opinion of Indiana Governor Mitch Daniels, who is up for re-election in November,

g) the issue the next President should focus on first,

h) the issue the next Governor should focus on first,

i) who the respondent would vote for in the November general election for President if the Democratic nominee is Clinton,

j) who the respondent would vote for in the November general election for President if the Democratic nominee is Obama,

k) if Clinton is Democratic nominee, and if Clinton picks Indiana Senator Evan Bayh as running mate, does respondent change vote in November general election,

l) who respondent would vote for in November general election for Governor if Democratic nominee is Long Thompson, and

m) who respondent would vote for in general election for Governor if Democratic nominee is Schellinger.

· TV poll was balanced to most recently available U.S. Census targets for Indiana adults, on gender, age, race and 4 regions defined at the county level by SurveyUSA. To see which Indiana counties are in which region, go here.

· University poll was balanced to demographics of the complete voter file for gender, age and 3 regions, each of which region is a cluster of 3 Indiana Congressional Districts. Indiana CD 1, 2 and 3 are combined into a region labeled Northern Indiana. CD 4, 5, and 7 are combined into a region labeled Central Indiana. CD 6, 8 and 9 are combined into a region labeled Southern Indiana.

Any one of these methodological changes, or any combination of them, may contribute to the differences between certain findings in the TV poll and the University poll.