Page 49 - Statistics II for Dummies

Chapter 2: Finding the Right Analysis for the Job 33

Measuring Precision with Margin of Error

Precision is the amount of movement you expect to have in your sample
results if you repeat your entire study again with a new sample. Precision
comes in two forms:

✓ Low precision means that you expect your sample results to move a lot
(not a good thing).
✓ High precision means you expect your sample results to remain fairly
close in the repeated samples (a good thing).
In this section, you find out what precision does and doesn’t measure and
you see how to measure the precision of a statistic in general terms.
Before you report or try to interpret any statistical results, you need to have
some measurement of how much those results are expected to vary from
sample to sample. This measurement is called the margin of error. You always
hope, and may even assume, that statistical results shouldn’t change much
with another sample, but that’s not always the case.

Up close and personal: Survey results

The Gallup Organization states its survey results The first sentence of the quote refers to how
in a universal, statistically correct format. Using the Gallup Organization collected the data,
a specific example from a recent survey it as well as the size of the sample. As you can
conducted, here’s the language it uses to report guess, precision is related to the sample size,
its results: as seen in the section “Measuring Precision
with Margin of Error.”
“These results are based on telephone
interviews with a randomly selected The second sentence of the quote refers to the
national sample of 1,002 adults, aged 18 precision measurement: How much did Gallup
years and older, conducted June 9–11, expect these sample results to vary? The fact
2006. For results based on this sample, one that Gallup is 95 percent confident means that
can say with 95 percent confidence that if this process were repeated a large number
the maximum error attributable to sampling of times, in 5 percent of the cases the results
and other random effects is ± 3 percent- would be wrong, just by chance. This incon-
age points. In addition to sampling error, sistency occurs if the sample selected for the
question wording and practical difficulties analysis doesn’t represent the population —
in conducting surveys can introduce error not due to biased reasons, but due to chance
or bias into the findings of public opinion alone. Check out the section “Avoiding Bias”
polls.” to get the info on why the third sentence is
included in this quote.

06_466469-ch02.indd 33 7/24/09 9:31:40 AM