Download Research Methods and Statistics: Survey - Wiki-cik

Research: Analyzing the Data
• Now that data has been gathered from a
correlational, descriptive, or experimental
research method, it’s time to analyze it!
• Off the top of your head estimates are often
misleading. Big, round, undocumented numbers
are misleading and should be investigated!
– Ex: One percent of Americans-2.6 million-are
homeless. (Or is it 300,000, as estimated by the
government?)
– We ordinarily use only 10 % of our brain. (Or is it
closer to 100%? Which 90% would you be willing to
sacrifice?)
Analyzing the data
• Measures of central tendency: help us summarize
the data for quick analysis.
• Mode: most frequently occurring score
• Mean: Arithmetic average
• Median: The middle score
Analyzing the data
• A few abnormally large or small numbers
can throw off the mean in statistical data.
Always note which measure of central
tendency is being reported.
Measures of Variation
• How similar or diverse
are the scores in the
data?
• Low variability= more
reliability
• Range of scores: the gap
between the lowest and
the highest scores
provides only a crude
estimate of variation
because a couple of
extreme scores in an
otherwise uniform group,
such as $475000 and
$710,000 will create a
deceptively large range.
• Standard deviation: how
much the scores deviate
from one another.
When is a difference reliable?
• Representative samples better than biased samples.
– Keep in mind what population a study has sampled
• Less-variable observations are more reliable than those
that are more variable.
– An average is more reliable when it comes from scores with low
variability.
• More cases are better than fewer.
– Averages based on many cases are more reliable (and less
variable) than averages based on only a few cases.
• Don’t be overly impressed by a few anecdotes.
Generalizations based on a few unrepresentative cases
are unreliable.
When is a difference significant?
• Data must be reliable before being judged for
their significance.
• Statistical significance: the sample averages are
reliable and the difference between them is
relatively large
– i.e. the difference observed is probably not due to
chance
• Remember: statistical significance indicates the
likelihood that a result will happen by chance, it
does not indicate the importance of the result.