We are collaborating with Pollfish on some public opinion polling! Click here for the permanent home.

Key findings this week:
1) This survey was 21 days after the last one and there was a meaningful increase in the amount of people taking the strong Republican position on average, but no noticeable change in the amount of people taking the strong Democratic position on average.

2) The biggest movements were in global warming, where 3 pp more people take the strong GOP position and medicare where 3 pp take the strong Democratic position.

3) We saw a rightward shift on immigration with 3 pp loss for the strong Democratic position and 2 pp gain for the strong Republican position. There was no position where less people professed the strong Republican position.

Please click here for the permanent home with tables and charts, with details of all of the questions …

Technical Details:

Pollfish does in-application polling on third party applications. We are reporting the modeled and post-stratified results. First, we run hierarchical logit regressions of the answers 1-7 on age, gender, geography, education, race, and party identification. Second, we post-stratify the results to the target populations of likely voters in the 2016 presidential election based on our interpretation of TargetSmart‘s voter file. This is experimental work, but we want to highlight three things:

    1. Cheap: the retail cost of running this survey is just $1,000 per week. A similar survey with random digit live dialing (RDD) would probably cost between $10,000 and $20,000 per week.
    2. Fast: while we are trying to be consistent with the questions, we can add questions and have results in 1 or 2 hours. Further, the lower cost allows us to have high time granularty, which is prohibitively expensive with RDD data collection.
    3. Accurate: our research shows that this method is accurate. But, we are also going to emphasize shifts in sentiment as much as levels of sentiment, where shifts in sentiment reflect the within survey design changes. As long we are surveying from a similar sample frame each week, the shifts are real within that sample frame, regardless of any concerns over the exact level of the sentiment.

There are two really topical papers to reference for this work. First, David and Sharad Goel (along with Adam Obeng) have a paper called Non-Representative Surveys: Fast, Cheap, and Mostly Accurate which talks about the spectrum of data collection. Second, David and Sharad Goel (along with Wei Wang and Andrew Gelman) have a paper called Forecasting elections with non-representative polls which talks about the analytics of modelling and post-stratifying in more depth.