- Hi Ekos Politics, Poll Me This . . .

Submitted: August 19th, 2010 at 11:10 am
tories claw back small lead - august 19, 2010
defecting university educated voters propel grits into a much more competitive race, August 19, 2010
http://www.ekospolitics.com/index.php/2010/08/tories-claw-back-small-lead-august-19-2010/#comment-10077

Click here for the full_report_august_19
http://www.ekospolitics.com/wp-content/uploads/full_report_august_19.pdf


Hi Ekos Politics,

I have a few questions

- What about undecideds? I find that these numbers are generally left out, except some polls give result for second choice.

Your Detailed Table: National Federal Vote Intention: August 11 - 17 (week 2), at p. 8, seems to suggest that these results are only amongst decided voters (support for the parties ads up to 99.9%, not 100% presumably due to rounding). Am I reading this wrong and if not why do it this way.

How can we make any kind of analysis of voter trends without these numbers. You may have them in your full 17 page report but I don't have time to read it so carefully. Perhaps in your press releases, summaries and tables, you could set them out as well.

- If 50.4% feel the country is going in the right direction and 44% feel the government is going in the right direction, why would the Con's get only 32.5% for voter intention? Do you any suggestions as to what this is all about and why this is happening.

- The results for Ontario are essentially identical, not just statistically, as the results nationally, except for the Lib's and the Block. Is this just a co-incidence, a result of the poll design, or a reflection of actual voter intentions, or otherwise?

Also, your margins of error for all the other regions are double that for Ont & Que, where these two are within the normal margins of error for these types of polls. It seems your poll is very heavily weighted to Ontario and Quebec.

Why do you do this, especially given the perceived East-West schism? I understand that the pop of Ontario is so large that it has big influence on the result but is it really that pronounced and given this why allow the big difference in margins of error.

Lloyd MacILquham cicblog.com/comments.html
wlloydm@hotmail.com