02 Experimental Method and Statistical Reasoning in Psychology

Download Report

Transcript 02 Experimental Method and Statistical Reasoning in Psychology

Experimental Method and
Statistical Reasoning in
Psychology
The Scientific Method


The four basic goals of psychology are to (1)
describe, (2) explain, (3) predict, and (4)
control or influence behavior and mental
processes.
To achieve these goals, psychologists rely on
the scientific method. The scientific method
refers to a set of assumptions, attitudes, and
procedures that guide researchers in creating
questions to investigate, in generating
evidence, and in drawing conclusions.
The Scientific Method


Like all scientists, psychologists are guided by the
basic scientific assumption that events are lawful.
When this scientific assumption is applied to
psychology, it means that psychologists assume that
behavior and mental processes follow consistent
patterns.
Psychologists are also guided by the assumption that
events are explainable. Thus, psychologists assume
that behavior and mental processes have a cause or
causes that can be understood through careful,
systematic study.
The Scientific Method



In striving to discover and understand consistent
patterns of behavior, psychologists are open-minded.
They are willing to consider new or alternative
explanations of behavior and mental processes.
However, their open-minded attitude is tempered by
a healthy sense of scientific skepticism. That is,
psychologists critically evaluate the evidence for new
findings, especially those that seem contrary to
established knowledge.
And, in promoting new ideas and findings,
psychologists are cautious in the claims they make.
Empirical evidence


Like any science, psychology is based on
empirical evidence—evidence that is the
result of objective observation, measurement,
and experimentation.
As part of the overall process of producing
empirical evidence, psychologists follow the
four basic steps of the scientific method.
The Steps in the Scientific
Method




Formulate a specific question that can
be tested.
Design a study to collect relevant data.
Analyze the data to arrive at
conclusions.
Report the results.
Step 1. Formulate a Hypothesis That
Can Be Tested Empirically


Once a researcher has identified a question or an issue to investigate, he or
she must formulate a hypothesis that can be tested empirically. Formally, a
hypothesis is a tentative statement that describes the relationship between
two or more variables. A hypothesis is often stated as a specific prediction
that can be empirically tested, such as “psychological stress increases the
likelihood of physical illness.”
The variables contained in any given hypothesis are simply the factors that
can vary, or change. These changes must be capable of being observed,
measured, and verified. The psychologist must provide an operational
definition of each variable to be investigated. An operational definition
defines the variable in terms of how it is to be measured, manipulated, or
changed.
Operational definitions


Operational definitions are important because many
of the concepts that psychologists investigate—such
as memory, happiness, or stress—can be measured
in more than one way.
In providing operational definitions of the variables in
the study, the researcher spells out in very concrete
and precise terms how the variables will be
manipulated or measured. In this way, other
researchers can understand exactly how the variables
were measured or manipulated in a particular study.
Step 2. Design the Study and Collect the
Data Descriptive methods


Descriptive methods are research strategies for
observing and describing behavior, including identifying
the factors that seem to be associated with a particular
phenomenon. Descriptive methods answer the who,
what, where, and when kinds of questions about
behavior.
Who engages in a particular behavior? What factors or
events seem to be associated with the behavior? Where
does the behavior occur? When does the behavior
occur? How often?
The experimental method


The experimental method is used to show that one
variable causes change in a second variable. In an
experiment, the researcher deliberately varies one
factor, then measures the changes produced in a
second factor.
Ideally, all experimental conditions are kept as
constant as possible except for the factor that the
researcher systematically varies. Then, if changes
occur in the second factor, those changes can be
attributed to the variations in the first factor.
Step 3. Analyze the Data and
Draw Conclusions



Once observations have been made and measurements
have been collected, the raw data need to be summarized
and analyzed.
Researchers use the methods of a branch of mathematics
known as statistics to summarize, analyze, and draw
conclusions about the data they have collected.
Researchers rely on statistics to determine whether their
results support their hypotheses. They also use statistics
to determine whether their findings are statistically
significant.
If a finding is statistically
significant


If a finding is statistically significant,
it means that the results are not very
likely to have occurred by chance.
As a rule, statistically significant results
confirm the hypothesis.
Meta-analysis



A statistical technique called meta-analysis is
increasingly being used in psychology to analyze the
results of many research studies on a specific topic.
Basically, meta-analysis involves pooling the results
of several studies into a single analysis. By creating
one large pool of data to be analyzed, metaanalysis
can sometimes reveal overall trends that may not be
evident in individual studies.
Meta-analysis is especially useful when a particular
issue has generated a large number of studies, some
of which have produced weak or contradictory
results.
Step 4. Report the
Findings






For advances to be made in any scientific discipline,
researchers must publish or share their findings with
other scientists. In addition to reporting their results,
psychologists provide a detailed description of the study
itself, including the following:
Who participated in the study
How participants were selected
How variables were operationally defined
What procedures or methods were used
How the data were analyzed
What the results seem to suggest
The precise details of the
study


Describing the precise details of the study makes it
possible for other investigators to replicate, or
repeat, the study. Replication is an important part of
the scientific process. When a study is replicated and
the same basic results are obtained again, scientific
confidence that the results are accurate is increased.
Conversely, if the replication of a study fails to
produce the same basic findings, confidence in the
original findings is reduced.
Naturalistic Observation



When psychologists systematically observe and
record behaviors as they occur in their natural
settings, they are using the descriptive method called
naturalistic observation.
Usually, researchers engaged in naturalistic
observation try to avoid being detected by their
subjects, whether people or nonhuman animals.
The basic goal of naturalistic observation is to detect
the behavior patterns that exist naturally—patterns
that might not be apparent in a laboratory or if the
subjects knew they were being watched.
A Case Study



A case study is an intensive, in-depth investigation of an individual
or a small group of individuals. Case studies involve compiling a
great deal of information, often from a variety of different sources,
to construct a detailed picture of the person.
The subject may be intensively interviewed, and his or her friends,
family, and co-workers may be interviewed as well. Psychological
records, medical records, and even school records may be
examined. Other sources of information can include extensive
psychological testing and observations of the person’s behavior.
Clinical psychologists and other mental health specialists routinely
use case studies to develop a complete profile of a psychotherapy
client. Case studies are also used to investigate rare, unusual, or
extreme conditions.
Surveys



A direct way to find out about the behavior, attitudes, and opinions of
people is simply to ask them. In a survey, people respond to a
structured set of questions about their experiences, beliefs, behaviors,
or attitudes.
One key advantage offered by survey research is that information can
be gathered from a much larger group of people than is possible with
other research methods.
Typically, surveys involve a carefully designed questionnaire in a paperand-pencil format that is mailed to a select group of people. Computerbased or Internet-based surveys have become increasingly more
common. And, surveys are still often conducted over the telephone or in
person, with the interviewer recording the person’s responses.
Surveys


As with paper-and-pencil surveys, the interviewer
usually asks a structured set of questions in a
predetermined order. Surveys are seldom
administered to everyone within the particular group
or population under investigation.
Instead, researchers usually select a sample—a
segment of the larger group or population. Selecting
a sample that is representative of the larger group is
the key to getting accurate survey results. A
representative sample very closely parallels, or
matches, the larger group on relevant characteristics,
such as age, sex, race, marital status, and
educational level.
IQ and Classification of Intelligence (DSM-IV and
WAIS-R)
<20—25
Profound mental retardation
20—25 to 35—40
Severe mental retardation
35-40 to 50—55
Moderate mental retardation
50—55 to 70
Mild mental retardation
70-79
Borderline
80—89
Low average/dull normal
90-109
Average/normal
110-119
High average/bright normal
120-129
Superior
>130
Very superior
The Correlational Study


A correlational study examines how strongly two
variables are related to, or associated with, each other.
Correlations can be used to analyze the data gathered
by any type of descriptive method.
The participants also filled out a questionnaire on past
delinquent behavior and completed tests designed to
measure different personality characteristics. Finally,
each respondent’s cumulative grade should point
average. Once the data were collected from their survey
participants, psychologists used statistical procedures to
calculate a figure called a correlation coefficient. A
correlation coefficient is a numerical indicator of the
strength of the relationship between two factors.
The Correlational
Study


A positive correlation is one in which the two factors vary
in the same direction. That is, the two factors increase or
decrease together. For example, Anderson and Dill found that
there was a positive correlation of .22 between the amount of
time spent playing violent video games and aggressive
personality characteristics. That is, as the amount of time
spent playing violent video games increased, aggression
scores on personality tests increased.
In contrast, a negative correlation is one in which the two
variables move in opposite directions: As one factor
decreases, the other increases. For example, Anderson and
Dill found that there was a negative correlation of .20
between the amount of time spent playing video games and
academic achievement, as measured by cumulative college
grade-point average. As the amount of time spent playing
video games increased, college grade-point average
decreased.
Experimental method
The experimental method is a research method used to
demonstrate a cause-and-effect relationship between changes in
one variable and the effect that is produced on another variable.
Conducting an experiment involves deliberately varying one
factor, which is called the independent variable. The
researcher then measures the changes, if any, that are produced
in a second factor, called the dependent variable. The
dependent variable is so named because changes in it depend on
variations in the independent variable.
To the greatest degree possible, all other conditions in the
experiment are held constant. Thus, when the data are analyzed,
any changes that occur in the dependent variable can be
attributed to the deliberate variations of the independent variable.
In this way, an experiment can demonstrate a cause-and-effect
relationship between the independent and dependent variables.
Limitations of Experiments


The strength of a well-designed experiment is that it can
provide convincing evidence of a cause-and-effect
relationship between the independent and dependent
variables.
Experiments do have limitations, however. Because
experiments are often conducted in highly controlled
laboratory situations, they are frequently criticized for
having little to do with actual behavior. That is, the artificial
conditions of some experiments may produce results that
do not generalize well, meaning that the results cannot be
applied to real situations or to a more general population
beyond the participants in the study.
Limitations of Experiments


In order to make experimental conditions less
artificial, experiments are sometimes
conducted in a natural setting rather than in
a laboratory.
Another limitation of the experimental
method is that even when it is possible to
create the conditions that the researchers
want to study, it may be unethical to do so.
The most spread
personality tests




Minnesota Multiphasic Personality
Inventory (MMPI)
Eysenck Personality questionnaire
(EPQ):
Rorschach Inkblot test
Thematic Apperception Test (TAT)