Professional Documents
Culture Documents
• Analogical Theory
• Explains the relationships among
variables through analogies to something
that is well-understood
• E.g., Memory as a computer
Classifying Theories
• Fundamental Theory
• Proposes a new structure to
explain the relationships among
variables
• Not only describes behavior, it
explains behavior – not through
analogies but through a new
structure
• Highest level of theory
• E.g., Cognitive dissonance theory
Classifying Theories
• Confirmational Strategy
• Look for evidence to confirm
predictions from a theory
• Important part of theory testing,
but has limits
• Confirmation does not prove a
theory is correct
• Confirmation may occur when
predictions are too loosely defined
Testing Theories
• Disconfirmational Strategy
• Using a positive research result to
disconfirm a theory’s predictions
* Should use both strategies
• Begin with confirmational
• Can it explain the phenomena?
• Move to disconfirmational
• Do unexpected outcomes happen?
Strong Inference
• Theory is tested and modified based on
outcome of research and then tested
again
• Cycle of testing and modification
continues until theory adequately
accounts for behavior
• Several alternative explanations can
be tested with an experiment
• Some ruled out
• New experiment tests remaining
alternatives
• Continue until only one alternative remains
Good Questions
• Not too broad
• Operational definitions – describe
how a variable will be measured
− Limits generalizability
Good Questions
• Ask IMPORTANT questions
• Clarify theoretical or empirical issues
• Support one hypothesis/theory over
another
• Address important practical issues
• Probably unimportant if:
• Answer already firmly established
• Small effects of no theoretical interest
• No reason to believe that 2 variables are
related
Purposes of Literature
Review
• To fully describe the results from
prior research
• What is the “state of the knowledge”?
• To clearly state the purposes of the
study
• Purpose is to address some need
• To clearly state the hypotheses,
which should follow logically from
the literature review
Sources of Research
Info
• Primary vs. Secondary Sources
• A primary source includes a full
report of a research study, including
methodological details
* Primary sources are preferred because
secondary sources may be biased or
author may not have discussed
something you may think is important.
• A secondary source summarizes
information from a primary source
(e.g. review article)
• Exception – meta-analysis
Sources
• Books
• General textbooks or specialized
professional publications
• Anthologies - assemble papers that
an editor feels are important in a
field
• Possibility for editor’s bias
• Most useful in early stages of
literature search
• Should be used with caution - may
not undergo rigorous review; info
Sources Cont.
• Scholarly Journals
• Current research and theoretical
thinking
• Refereed vs. nonrefereed journal
* Prefer refereed sources
• You can evaluate the quality of a
journal by
• Consulting Journals in Psychology
• Consulting the Social Science
Citations Index
• Using the method of authority
Sources Cont.
• Conventions and Professional
Meetings
• Gatherings of researchers to
present findings
• Provide the most up-to-date info
• Advantages of attending
convention
• Information is at the frontiers of
science
• Meet others in your field and
exchange ideas
Evaluating a Research
Article: The Introduction
• Has relevant research been adequately
reviewed?
• Are assertions supported with the
appropriate citations?
• Are the purposes of the study clearly
stated?
• Are the hypotheses clearly stated, and
do they flow logically from the info in
the introduction?
Evaluating a Research Article:
The Method Section
• Was the nature of the subject sample
specified?
• Does the design of the study allow an
adequate test of the hypotheses?
• Are there any methodological flaws
that might affect the validity of the
results?
• Is sufficient detail presented to allow
one to replicate the study?
Evaluating a Research Article:
The Results Section
• Did the statistically significant effects
support or refute the hypotheses?
• Are the differences reported large or
small?
• Were the appropriate statistics used?
• Do the tables, figures, and text match?
Evaluating a Research Article:
The Discussion Section
• Do the conclusions presented match the
results reported?
• If the author speculates about
implications of results, does he or she
stray too far from the results reported?
• How well do the results mesh with
existing theory and empirical data?
• Does the author point the way to
directions for future research?
Factors Affecting the Quality
of Research Information
• Statistical Significance
• Journals typically do not publish
findings that do not meet the
minimum .05 level of statistical
significance
• Form of nonexperimental
research
• Goals
• Determine whether two variables
covary
• Establish direction, magnitude,
forms of the relationship
• No variables are manipulated
• Instead just observed
Correlational Research
Descriptive Purposes
• Determine whether a relationship
exists between 2 variables
Predictive Purposes
• Knowledge of value of one variable
can help us predict value of related
variable
• Predic to r v aria ble used to predict
the value of a Criterion variable
Correlation & Causality
• Correlational research ca nno t be
used to establish caus al
re lat ion ship s among variables –
correlation does not equal causation
• Causality: 3 conditions to say that A
causes B
• A precedes B
• A is related to B
• All alternative explanations have been
ruled out
Obstacles to Causal
Explanations
• Third variable problem
• An unmeasured variable may account
for changes in both variables
• Both observed variables may vary together
although they are not directly related
• Examples
• Aggressive video games and aggression –
aggressive personality
• Ice cream sales and crime – hot weather
• There are ways to statistically
account for the third variable problem
Obstacles to Causal
Explanations
• The directionality problem
• When a causal relationship does exist,
it is often hard to know the direction of
the effect
• Does A cause B, B cause A, or both?
• Examples
• Playing aggressive video games and
aggression
• Weight and frequency of exercise
When to Use Correlational
Research
• Gathering data in the early stages
of research
• Identify possible causal relationships
which can then be tested
experimentally
• Inability to manipulate variables
• Manipulating independent variable
may be impossible or unethical
• Relating naturally occurring
variables
Experimental Research
• Involves high degree of control over
variables
• Allows us to establish causal
relationships
• An Independent Variable is
manipulated
• A variable whose values are chosen
and set by the experimenter
• Participants must be exposed to at
least two levels of this variable
Experimental Research
• A Dependent Variable is measured
• The variable whose value you observe (the
outcome)
• The value of the dependent variable depends
on the participants’ behavior
• Sample used
• Results may apply only to subjects
representing a unique group (i.e.
college students)
Threats to External
Validity
• Reactive effects of
experimental arrangements
• Participants’ knowledge that they
are research subjects may affect
results (demand characteristics)