PSYC1003 Study Guide - Final Guide: Sample Size Determination, Design Issues, Confidence Interval
Research Methods
Independent variable: the manipulation introduced by the researcher, the
treatment variable
Dependent variable: the outcome measured by the researcher, the variable that is
dependent upon the treatment
Random assignment: participants randomly placed into groups
Control versus experimental group: control experiences no differences whereas in
the experimental group the IV is manipulated
Confounds: an unintended manipulation of another variable related to the IV
Correlation causation
The Case Study
Method
The Survey (or
Correlational)
Method
The Quasi-
Experimental
Method
The Experimental
Method
What is it - The
researcher
intensively
observes
one person
or group of
people
- The researcher
collects
information
about different
variables from a
number of
people, to
examine the
relationship
between the
variables
- Participants
are assigned
into one of
two or more
groups
based on a
particular
variable,
and the
dependent
variable is
measured
- Participants are
randomly
assigned to one
of two or more
groups
- The researcher
manipulates a
theoretically
relevant variable
- The effect of this
manipulation on
an outcome
variable (SV) is
measured)
When to
use
Describe
- When what
we want to
study is
rare,
unique, or
new
Describe + predict
- Used when we
are interested
in naturally
occurring
patterns in the
real world
Describe +
Predict +
(sometimes)
explain
- Used when
we can’t
achieve
random
allocation of
subject to
treatment
groups
(impossible
or
unethical)
- The IV isn’t
or can’t be
manipulate
d
Describe + predict +
explain
Strengths - Rare, unique - Provide a - Allows - Experiments give
find more resources at oneclass.com
find more resources at oneclass.com
situations
- For new
situations-
observation
s from case
studies can
be useful as
‘agenda
setters’ for
future
research
convenient way
of identifying
‘real-world’
issues and
problems
- Relationships
found can be
useful ‘agenda
setters’ for
future research
- They can be
useful for
making
predictions
based on
observed
relationships
- If two variables
are not related,
we can rule out
the possibility
of a causal link
researchers
to study
variables
that are not
able to be
manipulate
d
researchers
control of IV
- If there is
significant
change in the
DV, we can make
the logically
appropriate
casual inference
that this must
have been due
to our
manipulation of
the IV
- With sufficiently
large samples,
the results will
generalise to the
population at
large, providing
that our sample
is representative
of the
population in
relevant respects
Weaknesse
s
- What may
be true for
one person
may not be
true for
another
- Difficult to
tease apart
the
variables/
confounds
- Ethical
issues
- Just because
two things are
related, doesn’t
mean that one
is caused by the
other
- Assignment
to groups
isn’t
random-
interferes
with our
ability to
make valid
causal
inferences
- If not done well,
results may not
be valid due to
confounding
- Experimental
manipulation is
not always
possible
Experimental Design
find more resources at oneclass.com
find more resources at oneclass.com
Our objectives:
oReliability- the confidence that a given finding can be reproduced
oValidity- the confidence that a given finding shows what we believe it shows
Internal validity- does the outcome really reflect the experimental
manipulation?
External validity- how well can we generalise?
Construct validity- correspondence between the theory (construct)
and the measurement used
Designing the experiment
oChoosing the IV and DV
Operationalisation- translating our concept of interest into something
that is observable and measurable
An effective IV/ DV is valid, ethical, and a balance (relevance-
sensitivity trade-off)
oBetween vs within subjects design
Between-subjects design: there are two different groups of
participants in the control and experimental groups, and the
experimental manipulation occurs between these two groups.
Within-subjects design: the same group of participants are both the
control and experimental groups. The experimental manipulation
occurs within the same group.
When it goes wrong:
oThreats to internal validity
In within-subjects designs due to time- practice effects, fatigue
effects, maturation effects, history effects
oThreats to external validity
If the groups are not representative of the population
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Independent variable: the manipulation introduced by the researcher, the treatment variable. Dependent variable: the outcome measured by the researcher, the variable that is dependent upon the treatment. Control versus experimental group: control experiences no differences whereas in the experimental group the iv is manipulated. Confounds: an unintended manipulation of another variable related to the iv. The researcher intensively observes one person or group of people. The researcher collects information about different variables from a number of people, to examine the relationship between the variables. Participants are assigned into one of two or more groups based on a particular variable, and the dependent variable is measured. When what we want to study is rare, unique, or new. Used when we are interested in naturally occurring patterns in the real world. Used when we can"t achieve random allocation of subject to treatment groups (impossible or unethical) The iv isn"t or can"t be manipulate d.
