26 CHAPTER 1
Both the wording of survey questions and the order in which they appear can
affect the outcome. It is difficult to find a wording that will be understood in exactly the
same way by all those who read the question. Questions can be worded in a way that
the desired answer becomes obvious (often resulting in courtesy bias–type answers). For
example, “Do you agree that the new procedures for registering for classes are too com-
plicated?” is obviously looking for a confirmation, while “What is your opinion of the
new procedures for registering for classes?” is much more open to differing responses.
Even the order of questions in a survey matters: A question about how much should be
spent on public safety might have a very different answer at the beginning of a survey
than after a long list of questions about crimes and criminal activity.Correlations: Finding Relationships
1.7 Explain how researchers use the correlational technique to study relation-
ships between two or more variables.
The methods discussed so far only provide descriptions of behavior. There are really only
two methods that allow researchers to know more than just a description of what has
happened: correlations and experiments. Correlation is actually a statistical technique,
a particular way of organizing numerical information so that it is easier to look for pat-
terns in the information. This method will be discussed here rather than in the statistics
appendix found at the back of this text because correlation, like the experiment, is about
finding relationships. In fact, the data from the descriptive methods just discussed are
often analyzed using the correlational technique.
A correlation is a measure of the relationship between two or more variables. A
variable is anything that can change or vary—scores on a test, temperature in a room,
gender, and so on. For example, researchers might be curious to know whether cigarette
smoking is connected to life expectancy—the number of years a person can be expected
to live. Obviously, the scientists can’t hang around people who smoke and wait to see
when those people die. The only way (short of performing a very unethical and lengthy
experiment) to find out if smoking behavior and life expectancy are related to each other
is to use the medical records of people who have already died. (For privacy’s sake, the
personal information such as names and social security numbers would be removed,
with only the facts such as age, gender, weight, and so on available to researchers.)
Researchers would look for two facts from each record: the number of cigarettes the per-
son smoked per day and the age of the person at death.
Now the researcher has two sets of numbers for each person in the study that go
into a mathematical formula, to Learning Objective A.6, to produce a number
called the correlation coefficient. The correlation coefficient represents two things: the
direction of the relationship and its strength.Direction? How can a mathematical relationship have a direction?Whenever researchers talk about two variables being related to each other, what
they really mean is that knowing the value of one variable allows them to predict the
value of the other variable. For example, if researchers found that smoking and life
expectancy are indeed related, they should be able to predict how long someone might
live if they know how many cigarettes a person smokes in a day. But which way does
that prediction work? If a person smokes a lot of cigarettes, does that mean that he or she
will live a longer life or a shorter one? Does life expectancy go up or down as smoking
increases? That’s what is meant by the direction of the relationship.
In terms of the correlation coefficient (represented by the small letter r), the number
researchers get from the formula will either be a positive number or a negative number.
If positive, the two variables increase in the same direction—as one goes up, the othercorrelation
a measure of the relationship between
two variables.
correlation coefficient
a number that represents the strength
and direction of a relationship exist-
ing between two variables; number
derived from the formula for measur-
ing a correlation.
© The New Yorker Collection 1994 Leo Cullum
from cartoonbank.com. All Rights Reserved.