• Hey Guest!
    British Car Forum has been supporting enthusiasts for over 25 years by providing a great place to share our love for British cars. You can support our efforts by upgrading your membership for less than the dues of most car clubs. There are some perks with a member upgrade!
    **Upgrade Now**
    (PS: Upgraded members don't see this banner, nor will you see the Google ads that appear on the site.)
Tips
Tips

Misleading graphics

Share this page

longbridgehealey said:
I was a senior operations manager for one of the two largest social science research companies in the country for 21 years, and I can tell you - What the client wants the survey to indicate, the statistics can verify if the analysts skew them appropriately. :devilgrin:
Click to expand...


That's a fact! I have seen polls, for example, that suggest things like "57% of the American people agree with or approve of XYZ." These are often the only results that are seen or reported by the media. Very often, when you dig deeper into the polling data, you find that the sample group is bias on way or the other (i.e., people in groups that would be inclined to agree with XYZ are over-sampled, while those who would be more likely to disagree are under sampled. Also, very often "how" a question is asked can influence the results and leave an inaccurate impression of how people really feel about a subject.
 
Also, very often "how" a question is asked can influence the results and leave an inaccurate impression of how people really feel about a subject. [/quote]

Is that ever true! :yesnod:
 
Another interesting aspect of "polls" - who is being polled?

Most web-based polls are being taken by - surprise, surprise - the people who frequent the website where the poll is based!

And with landline phones being used for polls more than cell phones, you can bet there's a difference in the sample groups. See:

https://pewresearch.org/pubs/1761/cell-phones-and-election-polls-2010-midterm-elections

One thing for sure - A poll usually yields the responses of the people who were polled, whether they understood (or agreed with) the question or not.

Please answer my poll: YES or NO ... Did you stop beating your dog?

Tom
 
Looks like it's using #votes, #yes and #no all as slices of the pie. The #votes and the #Yes are the large slices, the #No is the sliver. Darn software.
 
It all boils down to the fact that behind most of these pie charts stand well fed folks looking for another fat slice of YOUR PIE! :wall: --elrey
 
I'm sorry, but you're surprised <span style="font-style: italic">why</span>?
 
Back
Top