Interpreting Your Poll
Sampling works
Organizations often spend too much on their survey research.
They often exhaust their limited polling budget on samples
that are too large. Or have suppliers conduct too many focus
groups. In this case, more is not better. Better would be
more frequent surveys and smaller samples. Remember, samples
of any size have some degree of precision (described by the
famous margin of sampling error).
After all, you wouldn’t read a daily paper or visit
a website just once every six months. Regular fresh updates
are better than big expensive one-time surveys with information
that you know will go stale before you can afford to poll
again.
Beware… pollsters who say a larger
sample is more reliable than a smaller sample. Samples of
any size—so long as they are truly random or “scientific”—have
some degree or accuracy.
Beware… pollsters who say you can
confidently use a “representative focus group”
to reach a decision on your strategy. Focus group participants
represent no one but themselves. They are not representative
(they show up for the money and the refreshments!). That doesn’t
mean focus groups are not useful. But you would no more tally
percentages of focus group participants who support your ideas
than ask them how they intend to vote and think that’s
an election trial heat.
Beware… pollsters who say their organization’s
findings are more reliable or accurate than other polling
services. No one knows if a poll is accurate or not. A survey
is a sample, not a census. And even if every customer, member
or voter were surveyed, there would likely still be errors
due to the problem of accurately tabulating the masses of
information solicited.
At Vector Research, it is our responsibility as the "pollster"
to handle these problems. Marc Zwelling, the president of
The Vector Poll™, will analyze the poll and provide
an interpretation of your poll so that you have a clear sense
of direction that lets you move
forward with confidence.
|
