Tuesday, July 29, 2008

KLRVU explains why you shouldn't believe their poll

The cleanup continues at KLRVU. We have already seen how Campaign Life's pollster KLRVU plagiarized a page explaining "why you should participate" in its polls (and here). I have already commented on the frightful mess that Allan Bruinooge's first rewrite of that page on July 25.

In the last days, Mr. Bruinooge has returned to the page and reworked it. Most of the plagiarism is gone and his text improved, which is a good thing. Many errors remain, which do not speak well for this pollster's eye for detail, but let that go.

We now have at least the three versions of KLRVU's page describing why you should participate in their polls. As you can see here, each version has gotten shorter.

One thing that has remained through all three versions of the document is this:
    If Klrv calls (July 24)
The shorter version of this: "Don't decline to participate, because if you decline, someone else's opinion will hold a greater sway in the poll". An earlier version of the page offered an illustration:
    Premier (July 25)
The point is that the result of a poll might be skewed if people don't participate. But this is precisely the problem with KLRVU's notorious Morgentaler poll. KLRVU claims to have called 157,115 Canadian households with the question

"Do you believe abortionist Henry Morgentaler deserves the Order of Canada?"

Of these 157,115, only 13,324 answered. But even if we accept KLRVU's numbers (which seem inflated), 144,000 must feel like the hypothetical premier-hater in KLRVU's example -- they have not had their opinion recorded.

But the question has to be asked. If non-participation removes from citizens their voice and thereby undermines the validity of a poll, shouldn't the 144,000 non-participants in KLRVU's Mortengaler poll also discredit its results?