Understanding Psychology (10th Ed)

 
Module 5 
Conducting Psychological Research 
45
As the participants in each group were holding their discussion, they suddenly 
heard through the intercom one of the other participants—the confederate—having 
what sounded like an epileptic seizure and then calling for help. 
The participants’ behavior was now what counted. The dependent variable was 
the time that elapsed from the start of the “seizure” to the time a participant began 
trying to help the “victim.” If six minutes went by without a participant’s offering 
help, the experiment was ended. 
As predicted by the hypothesis, the size of the group had a signifi cant effect 
on whether a participant provided help. The more people who were present, the 
less likely it was that someone would supply help, as you can see in Figure 4 
(Latané & Darley, 1970). 
Because these results are straightforward, it seems clear that the experiment con-
fi rmed the original hypothesis. However, Latané and Darley could not be sure that 
the results were truly meaningful until they determined whether the results repre-
sented a
signifi cant outcome .
Using statistical analysis, researchers can determine 
whether a numeric difference is a real difference or is due merely to chance. Only 
when differences between groups are large enough that statistical tests show them 
to be signifi cant is it possible for researchers to confi rm a hypothesis (Cohen, 2002; 
Cwikel, Behar, & Rabson-Hare, 2000).
signifi cant outcome
Meaningful 
results that make it possible for 
researchers to feel confi dent that they 
have confi rmed their hypotheses.
Research Method
Description
Advantages
Shortcomings
Descriptive and
Researcher observes a 
Offers insight into 
Cannot
correlational research
previously existing 
relationships between 
determine
situation but does not 
variables
causality 
make a change in the 
situation 
Archival research
Examines existing data 
Ease of data collection 
Dependent on
to confirm hypothesis because data already
availability of data 
exist
Naturalistic
Observation of 
Provides a sample of
Cannot control the 
observation
naturally occurring 
people in their natural “natural habitat” being
behavior, without
environment 
observed 
making a change in the
situation
Survey research
A sample is chosen to 
A small sample can be Sample may not be
represent a larger
used to infer attitudes representative of the
population and asked a and behavior of a 
larger population; 
series of questions
larger population