Simple Experiments. Causal Claim Boldest claim a scientist can make Verbs such as “associated...
-
Upload
jacob-hensley -
Category
Documents
-
view
214 -
download
0
Transcript of Simple Experiments. Causal Claim Boldest claim a scientist can make Verbs such as “associated...
Simple Experiments
Causal Claim
• Boldest claim a scientist can make• Verbs such as “associated with” and “related
to” replaced with “causes, influences, affects or makes”
• Must be based on sound experimental research
Experiments: The Basics
• Very specific meaning• Manipulation of at least one variable • Measurement of at least one variable• Control of possible threatening variables
Independent Variables
• Manipulated• At least two levels• Assign participants to at least one of the levels
(condition)• Plotted on the X axis• Examples:– Color, parenting type, amount of caffeine, minutes
of exercise a day
Dependent Variables
• Measured – Depends on IV
• Behavioral, physiological, self-reports, attitudes
• Determines kinds of statistics employed• Plotted on the Y axis
Control Variables
• Researchers need to be sure they are manipulating one variable at a time the IV
• Must hold all other factors/variables constant• Confound variables- may be the cause of the
change in the dependent variable. – Influences “internal validity”– Possible explanation other than the IV
Comparison Groups
• Control groups• Level of the IV that is intended to represent
“no treatment” or a neutral condition• AKA “placebo grou”
Other Important Factors
• Random selection
• Random assignment– Except when “matching “on variable
• Design
Design
• Refers to how the subjects are placed with regards to IV conditions/levels
• Depends on concerns for confound variables such as:– Fatigue– Practice– Variables known to be a potential confound
Within-Subjects Design
• All P’s exposed to all levels of the IV• Same P’s in each level• Advantage:– Less variability between groups
• Disadvantage:– Practice effect, fatigue, order effects; demand
characteristics• Use of counterbalancing present the levels
of the IV to in different orders
Counterbalancing
• Split participants into groups; each group receives one of the condition orders
• Example with IV that has 3 levels– ABC BCA– ACB CAB– BAC CBA
• Partial counterbalancing an option when too many levels
Between-Subjects Design
• Different P’s in each of the IV levels• Advantage:– No practice effects– No fatigue– No order effect
• Disadvantage:– Too much variability– Overcome by random selection and assignment– Large sample size
Matched-Groups Design
• Used when researcher concerned about known confounding variable (gender, IQ etc)
• Used to equate the groups so the effect of the IV is clearer