Chapter 2: The Methods of Social
WHY DO SOCIAL PSYCHOLOGISTS DO RESEARCH? ____________________________________
Many situations that we encounter in life can contain surprises and pitfalls that social
psychological research can help us anticipate and avoid.
Research can teach us how to interpret and predict the outcomes of various social experiences,
as well as help us understand our own behaviour and the behaviour of others.
Our personal opinions about why we behave like we do are often mistaken. Many of the factors
that influence our behaviour are hidden from us; they are represented in nonverbal, non-
conscious forms that aren’t accessible to introspection.
A hindsight bias refers to people’s tendency to be overconfident about whether or not they
could have predicted a given outcome after the fact. When you predict an outcome before you
find out what the outcome is, then you avoid the hindsight bias.
HOW DO SOCIAL PSYCHOLOGISTS TEST IDEAS? _______________________________________
Hypotheses are predictions about what will happen under particular circumstances. Hypotheses
often test broader theories about behaviour.
A theory is a body of related propositions intended to describe some aspect of the world.
Theories generally have support in the form of empirical data and often lead to greater
discoveries. For example, the theory of evolution explains bacterial adaption to drugs, and the
theory of relativity led to the creation of the atomic bomb.
An example of a hypothesis born from a social psychological theory is the prediction that if
person A likes person B, and person B dislikes person C, then person A will either dislike person
C or come to dislike person B. this hypothesis is generated by the balance theory, the theory
that people like to have consistent thoughts and behaviour and will do substantial mental work
to achieve cognitive consistency.
There are a number of ways to research hypotheses, including observations, archival research,
surveys, correlations, and experiments.
The simplest level of research is to merely look at a phenomenon in some reasonably systematic
way and hypothesize why things are happening the way they are. Social psychologists usually
observe situations in a semiformal way and then design additional research to verify their
Participant observation involves observing some phenomenon at close range. A psychologist will
spend time, and even live with the participants and note their behaviours. Archival Research
This method involves researchers looking at evidence found in archives of various kinds – record
books, police reports, sports statistics, newspaper articles, and databases containing
ethnographic (anthropological) descriptions of people from different cultures.
Example of Archival Research: Playboy and Playgirl Ads
York University researchers analyzed the “playmates of the year” in Playboy magazines from 1977
to 1996. They found that over the years, body size has continued to decrease. By the last year of the
study, all of the women were underweight according to Canada’s health guidelines.
Similarly, they analyzed Playgirl magazines from 1986 to 1997 and found that body size increase
dramatically in muscle, not fat.
Example of Observation and Archival Research: The NBA Study
Researchers analyzed videos of all 30 NBA teams during the 2008-2009 season. They looked at
how the players interacted with each other and recorded how often they touched each other, as well as
the duration of the physical contact.
Researchers then compared their information with archived statistical records from the NBA
website. They found that teams who touched each other more often and for longer durations won more
games than teams who didn’t.
The most common type of study in social psychology involves asking people questions. This
allows us to research attitudes, behaviours, or feelings that may be difficult to observe. Surveys
can be conducted using interviews or written questionnaires.
Samples are taken from a population, a particular group you want to research about (ex. First
year university students in Canada). A random sample uses participants that are randomly
selected from a population, thereby getting unbiased results (ex. Choosing every twentieth
name from a directory and asking those students to participate). A convenience sample uses
participants in some available subgroup of a population and can yield biased results because it
contains too many of some kinds of people and too few of others. (ex. Choosing students from
the library to participate).
Example of a Biased Survey: The 1936 US Election
An American magazine, Literary Digest, sampled more than a million people and asked them who
they would vote for: Republican Alf Landon or Franklin Delano Roosevelt. Their survey predicted that
Landon would win, but in reality, Roosevelt won. Their sample was biased because the readers of this
magazine were more likely to be republican and therefore was not representative of the population. Correlational Research
In correlational studies, researchers want to find out if a relationship exists between two
variables. Correlations do not prove causation!
Positive correlations mean that as one variable increases, the other variable increases (ex. If
you’re taller, you will weigh more).
Negative correlations mean that as one variable increases, the other variable decreases (ex. The
more you skip class, the worse you will do on the test).
Correlation coefficients can range from -1.0 to +1.0, -/+1.0 being a strong correlation, and 0
being no correlation at all.
Correlation does not prove causation. We can’t be sure if variable A causes variable B, if B
causes A, or if a third variable C causes both A and B.
Correlational studies are useful when a topic of research may be difficult or unethical to
experiment on (ex. Researchers can’t randomly assign people to different categories like gender,
intelligence, and socioeconomic class).
A longitudinal study is a study that is conducted over a long period of time with the same
population, which is periodically assessed regarding a particular behaviour. In this type of study,
a researcher may analyze a person when they are a child and then reanalyze them when they
are an adult. Using a longitudinal study rules out the opposite direction of causality (B causes A)
because nothing you do when you’re 30 can affect what you did when you were younger.
Experiments randomly assign people to different conditions, or situations, and enables
researchers to make strong inferences about how these different conditions affect people’s
behaviour. Variables are controlled by the experimenter.
An independent variable is a variable that is manipulated and is hypothesized to be the cause of
a particular outcome. The dependent variable is the variable that is being affected and
measured. For instance, if you wanted to know the effect of alcohol on your ability to walk in a
straight line, the in