Page 235 - Between One and Many The Art and Science of Public Speaking

202 Part 3 Putting Theory Into Practice

audience believes to be true, we fi rst have to convince them that ours are
more reliable if we are to have any success.

Numerical Data

Numerical summaries of data, such as percentages, ratios, and averages, are
valuable when used judiciously in our speeches. These can be a rich source of
information; yet they can also be confusing and misleading. For example, an
American automobile manufacturer announced a survey showing that its cars
were preferred overwhelmingly to foreign cars. However, it turns out that the
company included only 200 people in its survey, none of whom even owned a
foreign car. 4
We are constantly bombarded by numbers that seem authoritative but are of
dubious value. Some questions to ask about numerical data are the following:
• Is the source reliable and unbiased? The tip-off to the problem with the
survey on foreign versus American cars is that it was sponsored by an
American car company. Numbers found through general searches of com-
mercial, individual, or organizational Internet sites are often suspect. On
the other hand, those found in offi cial sources, such as www.fedstats.gov,
are less likely to be biased, because this site collects offi cial government
data.
• Are the numbers based on a poll? A meaningful poll calls participants, not the
other way around. Based on sophisticated sampling techniques and ran-
dom selection, a national poll can predict a presidential election with about
a four-percentage-point margin of error. But when our Internet provider,
local television station, or newspaper conducts an “unscientifi c poll,” in
which people record their views, the results are meaningless. Only people
who are interested in the topic will respond, and there is nothing to prevent
someone from responding a hundred times. In short, such polls are worse
than worthless because they undermine confi dence in legitimate polls.
• Were unbiased questions asked? A poll asking whether disposable diapers
should be banned was preceded by a statement that disposable diapers ac-
count for only 2 percent of trash in landfi lls. Not surprisingly, 84 percent of
those polled felt disposable diapers should not be banned. 5
• Was the sample representative? A representative sample is absolutely neces-
sary for a poll to be reliable and valid. A representative sample is one made
up of people who possess the same attributes as the people in the popula-
tion from which the sample is drawn. A speech class, for example, could
be representative of the student body at a college. But unless the class was
drawn randomly from the entire student body, we do not know for certain.
There are many ways to obtain a representative sample, but the most
common way is to randomly select people from the population in which we
are interested: for example, college students between the ages of 18 and 25;
all single mothers in the state; or members of a state bar association. Gen-
erally speaking, the larger the sample randomly drawn from a population,
the more representative the sample.
The complexities of sampling theory are beyond the scope of this book.
Even so, we want to emphasize that the value of any poll depends on sam-

8/24/07 3:20:52 PM
M4344.indb 202 8/24/07 3:20:52 PM
M4344.indb 202