Research Methods • Experimental Method – only method that can establish “cause and effect” – important components include: 1. random sampling 2. random assignment 3. hypothesis 4. control and experimental groups 5. independent variable and dependent variable 6. control of subject and experimenter bias
21
Embed
Research Methods Experimental Method –only method that can establish “cause and effect” –important components include: 1.random sampling 2.random assignment.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Research Methods
• Experimental Method
– only method that can establish “cause and effect”
– important components include:
1. random sampling
2. random assignment
3. hypothesis
4. control and experimental groups
5. independent variable and dependent variable
6. control of subject and experimenter bias
Control
• Manipulation of the independent variable - altering a condition of the experiment believed to influence the dependent variable.
• Constancy of conditions - maintaining conditions that may influence behavior at levels we choose or believe them to be (e.g., all Ss tested at or about the same time of day).
• Elimination of extraneous variables - identifying those variables which may influence behavior and exclude them from the research situation (e.g., environmental distractions).
The ability to establish cause and effect with experiments isdue to the degree of control possible in true experiments. There are three major means by which control is achieved:
Independent VariablesAn independent variable is one under the direct control of the experimenter (e.g., level of background white noise) or, if not under direct control, has been selected for inclusion in the experiment (e.g., gender) and is believed to influence behavior in a predictable manner.
Independent variables are manipulated in a variety of ways:• present different stimuli (e.g., complex vs. simple designs, drug
dose, highlighter vs. no highlighter)
• vary content in scenarios (e.g., three vs. five “friends”)
• vary characteristics of “paper people” (e.g., male vs. female)
• etc.
Dependent VariablesA dependent variable is a direct or indirect measure of behavior or mental process.
There a several categories of dependent variables:
• frequency of response
• length of response
• strength of response
• number of correct responses/number of errors
• reaction time (RT)
• rating scales
• test scores
Threats to Internal Validity
• history - specific events (extraneous variables) occurring between presentation of the independent variable and measurement of the dependent variable or between a first and second measurement (i.e., pretest-posttest), other than the independent variable.
• maturation - processes within the Ss that change as a function of the passage of time (e.g., growing older, more hungry, tired, etc.).
We have seen how manipulation of the independent variable functions in research. The other two means of control are directed toward establishing internal validity - ensuring all plausible rival hypotheses are eliminated (i.e., “confounding”).
There are a number of common threats to internal validity:
Threats to Internal Validity• testing - the effects of taking a test upon the scores of a second
test.
• instrumentation - changes in calibration of measuring instruments or changes in scorers/observers.
• differential selection - groups selected in such a way that may result in group biases (e.g., one group is more intelligent, more motivated, better educated, etc.).
• experimental mortality - differential loss of Ss from treatment groups (e.g., illness, time constraints, death, etc.).
Pre-Experimental DesignsThere are many instances in which true experiments are not possible. In those cases, pre-experimental or quasi-experimental designs are sometimes used. There are, however, threats to internal validity in each that must be acknowledged.
The following are three pre-experimental designs:
One-Shot Case Study
X O
One-Group Pretest-Posttest Design
O X O
Static-GroupComparison
X O O
Basic True Experimental Design
Population RandomSample
RandomSampling
Ideally, we would like to have a small group or “sample” to study which represents the entire population. That can be accomplished if we use “random sampling” to select our sample.
Basic True Experimental Design
Convenience
Sample
Random Assignment
Unfortunately, we are seldom able to obtain such a sample. We must, therefore, often rely on using subjects who are readily available -- a “convenience sample” -- and split them into a minimum of two groups using a technique called “random assignment.”
Control GroupExperimental
Group
Basic True Experimental Design
Control GroupExperimental
Group
PresentIndependent Variable
MeasureDependent Variable
CompareGroups
Basic Experimental Design
ConvenienceSample:
Students in Gen. Psych. class
Control GroupExperimental
Group
Random Assignment
Have you ever wondered whether those “Highlighters” help you study? Let’s see how we could develop an experiment to test the following hypothesis:
Highlighters facilitate memory of facts read from textbooks.
All subjects will be given several pages to read. After they have done so, they will bedismissed and asked to return to the experimental lab the next day.
Basic Experimental Design
Control Group Experimental Group
Present Independent VariableAvailability of Highlighter
MeasureDependent VariableNumber of correctlyrecalled facts on quiz
CompareGroups
No Highlighter Highlighter Available
Each subject is given five pages from an Intro Psych text and told to read thepages carefully because they will be tested on the material. The subjects aredismissed after they finish reading and asked to return the next day.
Expanding the Basic DesignThe basic experimental design is limited to comparing twogroups. We can, however, increase its versatility by using“factorial designs.”
Factorial designs use a slightly different terminology than what we have been using:
• Factor - an independent variable
• Levels - different values of the independent variable
• Treatment condition/group - a group exposed to one of the levels of the independent variable
Simple-Randomized Design
Aa1 a2 a3
Factor
S1
.
.
.
.
.n1
S1
.
.
.
.
.n2
S1
.
.
.
.
.n3
Levels
Between- vs. Within-Subjects
Aa1 a2 a3
Aa1 a2 a3
S1
.
.
.
.
.n1
S1
.
.
.
.
.n2
S1
.
.
.
.
.n3
S1
.
.
.
.
.n
Two-Factor Designs(completely crossed designs)
Aa1 a2
b1
B
b2
a1b1
a1b2
a2b1
a2b2
a3b1
a3b2
a3
This is simply a combination of two simple-randomized designs.
Mixed Designs
Aa1 a2 a3
b1
B
b2
S1...nS1...n
Between-subjectFactor
Within-subjectFactor
Higher-Order Factorial Designs
a1 a2 a3
A
b1
B
b2
c2
c1 C
a1b1c2
a3b2c1
Main Effects and Interactions
In addition to investigating several variables at one time to test for main effects, factorial designs permit us to investigate interactions as well.
Main effect - a significant influence of a factor on thedependent variable.
Interaction - the simultaneous effects of two factors on the dependent variable that may be different than the effects of the factors individually.
Graphing Main Effects
Aa1 a2
b1
B
b2
Main effect for Factor AMain effect for Factor BMain effects for Factors A and B