Statistics

Statistics
As explained previously, statistical techniques can be used to describe data, compare

two
or more data sets, determine if a relationship exists between variables, test hypotheses,
and make estimates about population characteristics. However, there is another aspect
of
statistics, and that is the misuse of statistical techniques to sell products that dont work
properly, to attempt to prove something true that is really not true, or to get our
attention
by using statistics to evoke fear, shock, and outrage.
There are two sayings that have been around for a long time that illustrate this point:
There are three types of lieslies, damn lies, and statistics.
Figures dont lie, but liars figure.
Objective
Explain how statistics
can be used and
misused.
7
Just because we read or hear the results of a research study or an opinion poll in the
media, this does not mean that these results are reliable or that they can be applied to
any
and all situations. For example, reporters sometimes leave out critical details such as
the
size of the sample used or how the research subjects were selected. Without this
information,
you cannot properly evaluate the research and properly interpret the conclusions
of the study or survey.
It is the purpose of this section to show some ways that statistics can be misused. You
should not infer that all research studies and surveys are suspect, but that there are
many
factors to consider when making decisions based on the results of research studies and
surveys. Here are some ways that statistics can be misrepresented.
Suspect Samples
The first thing to consider is the sample that was used in the research study. Sometimes
researchers use very small samples to obtain information. Several years ago,
advertisements
contained such statements as Three out of four doctors surveyed recommend
brand such and such. If only 4 doctors were surveyed, the results could have been
obtained by chance alone; however, if 100 doctors were surveyed, the results might be
quite different.
Not only is it important to have a sample size that is large enough, but also it is
necessary to see how the subjects in the sample were selected. Studies using
volunteers
sometimes have a built-in bias. Volunteers generally do not represent the population at
large. Sometimes they are recruited from a particular socioeconomic background, and
sometimes unemployed people volunteer for research studies to get a stipend. Studies
that require the subjects to spend several days or weeks in an environment other than
their home or workplace automatically exclude people who are employed and cannot
take time away from work. Sometimes only college students or retirees are used in
studies. In the past, many studies have used only men, but have attempted to
generalize
the results to both men and women. Opinion polls that require a person to phone
or mail in a response most often are not representative of the population in general,
since only those with strong feelings for or against the issue usually call or respond
by mail.
Another type of sample that may not be representative is the convenience sample.
Educational studies sometimes use students in intact classrooms since it is convenient.
Quite often, the students in these classrooms do not represent the student population of
the entire school district.
When results are interpreted from studies using small samples, convenience samples,
or volunteer samples, care should be used in generalizing the results to the entire
population.
Ambiguous Averages
In Chapter 3, you will learn that there are four commonly used measures that are
loosely
called averages. They are the mean, median, mode, and midrange. For the same data
set,
these averages can differ markedly. People who know this can, without lying, select the
one measure of average that lends the most evidence to support their position.
Changing the Subject
Another type of statistical distortion can occur when different values are used to
represent
the same data. For example, one political candidate who is running for reelection
Section 15 Uses and Misuses of Statistics 17
117
might say, During my administration, expenditures increased a mere 3%. His

opponent,
who is trying to unseat him, might say, During my opponents administration,
expenditures have increased a whopping $6,000,000. Here both figures are correct;
however, expressing a 3% increase as $6,000,000 makes it sound like a very large
increase. Here again, ask yourself, Which measure better represents the data?
Detached Statistics
A claim that uses a detached statistic is one in which no comparison is made. For
example,
you may hear a claim such as Our brand of crackers has one-third fewer calories.
Here, no comparison is made. One-third fewer calories than what? Another example is a
claim that uses a detached statistic such as Brand A aspirin works four times faster.
Four times faster than what? When you see statements such as this, always ask
yourself,
Compared to what?
Implied Connections
Many claims attempt to imply connections between variables that may not actually
exist.
For example, consider the following statement: Eating fish may help to reduce your
cholesterol. Notice the words may help. There is no guarantee that eating fish will
definitely
help you reduce your cholesterol.
Studies suggest that using our exercise machine will reduce your weight. Here the
word suggest is used; and again, there is no guarantee that you will lose weight by
using
the exercise machine advertised.
Another claim might say, Taking calcium will lower blood pressure in some
people. Note the word some is used. You may not be included in the group of some
people. Be careful when you draw conclusions from claims that use words such as may,
in some people, and might help.
Misleading Graphs
Statistical graphs give a visual representation of data that enables viewers to analyze
and
interpret data more easily than by simply looking at numbers. In Chapter 2, you will see
how some graphs are used to represent data. However, if graphs are drawn
inappropriately,
they can misrepresent the data and lead the reader to draw false conclusions. The
misuse of graphs is also explained in Chapter 2.
Faulty Survey Questions
When analyzing the results of a survey using questionnaires, you should be sure that
the questions are properly written since the way questions are phrased can often
influence
the way people answer them. For example, the responses to a question such as
Do you feel that the North Huntingdon School District should build a new football
stadium? might be answered differently than a question such as Do you favor
increasing
school taxes so that the North Huntingdon School District can build a new football
stadium? Each question asks something a little different, and the responses could be
radically different. When you read and interpret the results obtained from questionnaire
surveys, watch out for some of these common mistakes made in the writing of the
survey questions.
In Chapter 14, you will find some common ways that survey questions could be
misinterpreted by those responding and could therefore result in incorrect conclusions.

Statistics

Uploaded by

Copyright:

Available Formats

Statistics

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Statistics

Uploaded by

Copyright:

Available Formats

Statistics

As explained previously, statistical techniques can be used to describe data, compare

might say, During my administration, expenditures increased a mere 3%. His

You might also like