. 6
( 115 .)


variable increases whereas
An excellent example of a correlational study is one conducted by Del Barrio, Aluja,
the value of a second
and Garcia (2004). Del Barrio et al. investigated the relationship between personality
characteristics and an individualʼs capacity to feel empathy for someone in need. Del
Barrio et al. administered a measure of empathy and personality inventory measuring the
“Big Five” personality dimensions (energy, friendliness, conscientiousness, emotional
stability, and openness) to Spanish adolescents. Del Barrio et al. found that “friendli-
ness” correlated most strongly with empathy for both boys and girls. High scores on
the “friendliness” dimension related to higher empathy scores. They also found that
“energy,” “conscientiousness,” and “openness” all positively correlated with empathy
for girls and boys, although not as strongly as “friendliness.” “Emotional stability” did
not signi¬cantly correlate with empathy.
Based on this brief summary, you can see that six variables were measured: ¬ve per-
sonality dimensions and empathy. However, notice that Del Barrio and her colleagues did
not manipulate any of the variables. Therefore, there were no independent variables.
Although correlational research does not demonstrate causal relationships, it does
play an important role in science. Correlational research is used in situations where it
is not possible to manipulate variables. Any study of individual characteristics (age,
sex, race, and so on) is correlational. After all, you cannot manipulate someoneʼs age
or sex. Correlational research is also used when it would be unethical to manipulate
variables. For example, if you were interested in how alcohol consumption affects the
human fetus, it would not be ethical to expose pregnant women to various dosages of
alcohol and see what happens. Instead, you could measure alcohol consumption and
the rate of birth defects and look for a correlation between those two variables. Finally,
correlational research is useful when you want to study variables as they occur natu-
rally in the real world.

Settings for Social Psychological Research
Social psychological research is done in one of two settings: the laboratory or the ¬eld.
Laboratory research is conducted in a controlled environment created by the researcher;
participants come into this arti¬cial environment to participate in the research. Field
research is conducted in the participantʼs natural environment; the researcher goes to
the participant, in effect taking the study on the road. Observations are made in the
participantʼs natural environment; sometimes, independent variables are even manipu-
lated in this environment.
Chapter 1 Understanding Social Behavior 17

Figure 1.4 Scatterplots
Laboratory Research
showing correlations of
Most research in social psychology is conducted in the laboratory. This allows the
different directions and
researcher to exercise tight control over extraneous (unwanted) variables that might
strength: (a) correlation
affect results. For example, the researcher can maintain constant lighting, temperature, of 0 indicated by dots
humidity, and noise level within a laboratory environment. This tight control over the randomly arrayed;
environment and over extraneous variables allows the researcher to be reasonably con¬- (b) strong positive
dent that the experiment has internal validity”that is, that any variation observed in the correlation; (c) perfect
dependent variable was caused by manipulation of the independent variable. However, positive correlation (+1)
that tight control also has a cost: The researcher loses some ability to apply the results indicated by the dots
beyond the tightly controlled laboratory setting (external validity). Research conducted lined up perfectly, sloping
from bottom left to upper
in highly controlled laboratories may not generalize very well to real-life social behav-
right; (d) strong negative
ior, or even to other laboratory studies.
correlation; (e) perfect
negative correlation
Field Research
indicated by the dots lined
Field research comes in three varieties: the ¬eld study, the ¬eld survey, and the ¬eld
up perfectly, sloping from
experiment. In a ¬eld study, the researcher makes unobtrusive observations of the par-
upper left to lower right.
ticipants without making direct contact or interfering in any way. The researcher simply
watches from afar. In its pure form, the participants should be unaware that they are
¬eld study A descriptive
being observed, because the very act of being observed tends to change the participantsʼ
research strategy in which
behavior. The researcher avoids contaminating the research situation by introducing any
the researcher makes
changes in the participantsʼ natural environment. unobtrusive observations
Jane Goodallʼs original research on chimpanzee behavior was a ¬eld study. Goodall of the participants without
investigated social behavior among chimpanzees by observing groups of chimps from making direct contact or
interfering in any way.
a distance, initially not interacting with them. However, as Goodall became more
Social Psychology

accepted by the chimps, she began to interact with them, even to the point of feeding
them. Can we be sure that Goodallʼs later observations are characteristic of chimp
behavior in the wild? Probably not, because she altered the chimpsʼ environment by
interacting with them.
¬eld survey A descriptive In the ¬eld survey, the researcher directly approaches participants and asks them
research strategy in which questions. For example, he or she might stop people in a shopping mall and collect
the researcher directly information on which make of car they plan to buy next. The ubiquitous political polls
approaches participants and
we see all the time, especially during election years, are examples of ¬eld surveys.
asks them questions.
Field studies and surveys allow us to describe and catalogue behavior. Political polls,
for example, may help us discover which candidate is in the lead, whether a proposition
is likely to pass, or how voters feel about important campaign issues. However, they
cannot tell us what causes the differences observed among voters, because we would
need to conduct an experiment to study causes. Fortunately, we can conduct experi-
ments in the ¬eld.
The ¬eld experiment is probably the most noteworthy and useful ¬eld technique
¬eld experiment for social psychologists. In a ¬eld experiment, the researcher manipulates independent
A research setting in which variables and collects measure of the dependent variables (the participantʼs behavior).
the researcher manipulates In this sense, a ¬eld experiment is like a laboratory experiment. The main difference
one or more independent
is that in the ¬eld experiment, the researcher manipulates independent variables under
variables and measures
naturally occurring conditions. The principal advantage of the ¬eld experiment is that
behavior in the participant™s
it has greater external validity”that is, the results can be generalized beyond the study
natural environment.
more legitimately than can the results of a laboratory experiment.
As an example, letʼs say you are interested in seeing whether the race of a person
needing help in¬‚uences potential helpers. You might consider a ¬eld experiment in
which you have someone, a confederate of yours (a confederate is someone working
for the experimenter), pretend to faint on a subway train. In the experiment, you use
two different confederates, one a black male, the other a white male. The two are as
alike as they can be (in age, dress, and so on) except, of course, for skin color. You
then observe how many people help each man and how quickly they do so. Such an
experiment would be very realistic and would have a high degree of external validity.
Consequently, the results would have broad generality.
A disadvantage of the ¬eld experiment is that the researcher cannot control extra-
neous variables as effectively as in the laboratory. Thus, internal validity may be com-
promised. In the subway experiment, for example, you have no control over who the
participants are or which experimental condition (white or black confederate) they will
walk into. Consequently, the internal validity of your experiment”the legitimacy of the
causal relationship you discover”may suffer. The experiment also poses some ethical
problems, one of which is that the people who purchased a ride on the subway did not
voluntarily agree to participate in an experiment. We discuss the ethics of research in
a later section of this chapter.

The Role of Theory in Social Psychological Research
theory A set of interrelated
propositions concerning the On many occasions throughout this book, we refer to social psychological theories. A
causes for a social behavior theory is a set of interrelated statements or propositions about the causes of a particu-
that helps organize research
lar phenomenon. Theories help social psychologists organize research results, make
results, make predictions
predictions about how certain variables in¬‚uence social behavior, and give direction to
about the in¬‚uence of certain
future research. In these ways, social psychological theories play an important role in
variables, and give direction
helping us understand complex social behaviors.
to future social research.
Chapter 1 Understanding Social Behavior 19

There are a few important points to keep in mind as you read about these theo-
ries. First, a theory is not the ¬nal word on the causes of a social behavior. Theories
are developed, revised, and sometimes abandoned according to how well they ¬t with
research results. Rather than tell us how things are in an absolute sense, theories help us
understand social behavior by providing a particular perspective. Consider attribution
theories”theories about how people decide what caused others (and themselves) to
act in certain ways in certain situations. Attribution theories do not tell us exactly how
people assign or attribute causality. Instead, they suggest rules and make predictions
about how people make such inferences in a variety of circumstances. These predic-
tions are then tested with research.
The second important point about social psychological theories is that often, more
than one theory can apply to a particular social behavior. For example, social psycholo-
gists have devised several attribution theories to help us understand how we make deci-
sions about the causes for behaviors. Each theory helps provide a piece of the puzzle of
social behavior. However, no single theory may be able to account for all aspects of a
social behavior. One theory helps us understand how we infer the internal motivations
of another individual; a second theory examines how we make sense of the social situ-
ation in which that individualʼs behavior took place.

Theory and the Research Process
Theories in social psychology are usually tested by research, and much research is
guided by theory. Research designed to test a particular theory or model is referred
basic research Research
to as basic research. In contrast, research designed to address a real-world problem
that has the principal aim of
is called applied research. The distinction between these two categories is not rigid,
empirically testing a theory or
however. The results of basic research can often be applied to real-world problems, and
a model.
the results of applied research may affect the validity of a theory.
applied research Research
For example, research on how stress affects memory may be primarily basic research,
that has a principal aim to
but the ¬ndings of this research apply to a real-world problem: the ability of an eye-
address a real-world problem.
witness to recall a violent crime accurately. Similarly, research on how jurors process
evidence in complex trials (e.g., Horowitz & Bordens, 1990) has implications for pre-
dictions made by various theories of how people think and make decisions in a variety
of situations. Both types of research have their place in social psychology.

Theory and Application
Application of basic theoretical ideas may take many forms. Consider, for example, the
idea that it is healthy for individuals to confront and deal directly with psychological
traumas from the past. Although various clinical theories have made this assumption,
evidence in support of it was sparse.
In one study, social psychologist Jamie Pennebaker (1989) measured the effects of
disclosure on mind and body. The research showed that when the participants confronted
past traumas, either by writing or talking about them, their immunological functioning
improved and their skin conductance rates were lowered. This latter measure re¬‚ects a
reduction in autonomic nervous system activity, indicating a lessening of psychological
tension. In other words, people were “letting go” as they fully revealed their feelings
about these past traumas. Those who had trouble revealing important thoughts about
the event”who could not let go of the trauma”showed heightened skin conductance
rates. Pennebakerʼs work shows that the act of con¬ding in someone protects the body
from the internal stress caused by repressing these unvoiced traumas. Thus, this is an
example of basic research that had clear applications for real-life situations.
Social Psychology

What Do We Learn from Research in Social Psychology?
Two criticisms are commonly made of social psychological research. One is that social
psychologists study what we already know, the “intuitively obvious.” The other is that
because exceptions to research results can nearly always be found, many results must
be wrong. Letʼs consider the merits of each of these points.

Do Social Psychologists Study the Obvious?
William McGuire, a prominent social psychologist, once suggested that social psy-
chologists may appear to study “bubba psychology””things we learned on our
grandmotherʼs knee. That is, social psychologists study what is already obvious and
predictable based on common sense. Although it may seem this way, it is not the
case. The results of research seem obvious only when you already know what they
hindsight bias are. This is called hindsight bias, or the “I-knew-it-all-along” phenomenon (Slovic
Also known as the “I-knew- & Fischoff, 1977; Wood, 1978). With the bene¬t of hindsight, everything looks
it-all--along” phenomenon; obvious. For example, after the attacks on 9/11, some commentators asked why
shows that with the bene¬t of
President Bush or the CIA did not “connect the dots” and see the attacks coming.
hindsight, everything looks
Unfortunately, those dots were not so clear in the months and years leading up to the
attacks. In hindsight, the signs seemed to point to an attack, but before the incident,
things were not so clear. In fact, the 9/11 Commission pointed out that hindsight can
bias our perceptions of events:

Commenting on Pearl Harbor, Roberta Wohlstetter found it “much easier after the
event to sort the relevant from the irrelevant signals. After the event, of course,
a signal is always crystal clear; we can now see what disaster it was signaling
since the disaster has occurred. But before the event it is obscure and pregnant
with con¬‚icting meanings.” As time passes, more documents become available,
and the bare facts of what happened become still clearer. Yet the picture of how
those things happened becomes harder to reimagine, as that past world, with its
preoccupations and uncertainty, recedes and the remaining memories of it become
colored by what happened and what was written about it later. (9/11 Commission
Report, 2004)

Although the results of some research may seem obvious, studies show that when
individuals are given descriptions of research without results, they can predict the
outcome of the research no better than chance (Slovic & Fischoff, 1977). In other words,
the results were not so obvious when they were not already known!

Do Exceptions Mean Research Results Are Wrong?
When the ¬ndings of social psychological research are described, someone often points
to a case that is an exception to the ¬nding. Suppose a particular study shows that a
person is less likely to get help when there are several bystanders present than when
there is only one. You probably can think of a situation in which you were helped with
many bystanders around. Does this mean that the research is wrong or that it doesnʼt
apply to you?
To answer this question, you must remember that in a social psychological
experiment, groups of participants are exposed to various levels of the independent
variable. In an experiment on the relationship between the number of bystanders and
the likelihood of receiving help, for example, one group of participants is given an
opportunity to help a person in need with no other bystanders present. A second group
of participants gets the same opportunity but with three bystanders present. Letʼs say
Chapter 1 Understanding Social Behavior 21

Table 1.1 Results from a Hypothetical Study of Helping Behavior

Participant Number No Bystanders Three Bystanders

1 No help No help
2 No help No help
3 Help No help
4 Help Help


. 6
( 115 .)