•The only way to ensure equivalency of groups is to use random assignment13. Thus
perhaps you ﬂip a coin to determine whether people are in the control group or the
•Differences will tend to average out when participants are assigned randomly so that
the groups are equivalent on average. This also helps balance out the known and
Correlational Designs Examine How Variables Are Related
•Many times it is not possible to manipulate the variables of interest, and you have to
deal with them as they naturally occur. A correlational study14 cannot show causation.
•In correlational studies, it is always possible that some extraneous factor is
responsible for the apparent relation between your variables (third-variable problem15).
•Another problem with correlational studies is in knowing the direction of the cause-
effect relation between variables. Suppose some researchers took a survey and found
that people who reported sleeping a great deal also reported having a lower level of
stress. Does reduced stress lead to longer and better sleep, or does sleep reduce
people!s stress levels? Both scenarios seem equally plausible, an ambiguity that is
known as the directionality problem.
•Correlational studies are widely used in psychological science because they provide
important information about how variables are naturally related.
•Some questions cannot be studied via an experiment because manipulating an
independent variable would be unfeasible, unethical, or both. In this case, you need to
conduct correlational research. Knowing the relation between variables allows us to
Descriptive Studies Observe and Classify Behavior
•Descriptive studies, sometimes called observational studies because of the manner in
which the data are typically collected, involve observing and noting behavior in order
to provide a systematic and objective analysis. Some studies may rely on observing
behaviors at regular time intervals.
•There are two types of descriptive studies: naturalistic observation16, and participant
13 The procedure for placing research participants into the conditions of an experiment in which each
participant has an equal chance of being assigned to any level of the independent variable.
14 A research method that examines how variables are naturally related in the real world, without any
attempt by the researcher to alter them.
15 When the experimenter cannot directly manipulate the independent variable and therefore cannot be
conﬁdent that another, unmeasured variable is not the actual cause of differences in the dependent
16 A passive descriptive study in which observers do not change or alter ongoing behavior.
17 A type of descriptive study in which the researcher is actively involved in the situation.