FINAL: Chapter 10 Simple Experiemnts Flashcards
What are the three criteria for causation?
Covariation, Temporal Precedence, Internal Validity
Covariation in criteria for causation
The measured variable (our “outcome variable”) must change as the causal variable changes
Temporal Precedence in criteria for causation
The causal variable must come first and the outcome variable after
Internal Validity in criteria for causation
Alternative causal explanations must be ruled out
What are the 3 characteristics of experiments?
Empirical/Objective approach, Manipulation of variables, and keeping other factors constant
What is the Empirical/Objective approach?
Gathering info through the use of your senses during systematic hypothesis testing. Hypotheses are specific predictions that can be tested.
Manipulation of variables (in reference to characteristics of experiments)
In experiments, the hypothesized causal variable is manipulated by the researcher and the resulting outcome is measured
In the characteristics of experiments, what does it mean by “keeping all other factors constant”?
(internal validity: the ability to rule out alternative explanations for a causal relationship) (control variables: variables that the experimenter holds constant on purpose. Potential third variable that the researcher wants to rule out)
Which variable to we manipulate in an experiment? Which do we measure?
We manipulate the causal/independent, and measure the dependent
How do experiments show internal validity?
Show this by manipulating the IV while holding all other factors constant
How do experiments show temporal precedence?
Manipulating the hypothesized cause and measuring its subsequent effect on outcomes
How do experiments show covariation?
The IV is systematically manipulated, and then we look to see if the DV is also changed systematically
Do we always need an IV in an experiment? If so, how many?
At least one
Do we always need a manipulated IV? If so, how many?
At least one
What type of IV is not manipulated?
Subject Variable
What is a subject variable?
Characteristics of the subjects themselves that determine the assignment to a particular group can only be measured. Must have at least one other IV that is manipulated
What is a between-subjects design?
Different groups in each level of the IV
What keywords indicate that there is a between-subjects design?
“randomly assigns __ to either..” “ to __ or __”
What is a within-subject design?
Same groups experience each level
What keywords indicate that there is a within-subjects design?
“__ and __”
What are the 3 types of manipulations?
Environmental, Instructional, Invasive
What is an environmental type of manipulation?
Changing the social/physical environment (EX: Do people perform differently when others are watching?)
What is an Instructional type of manipulation?
Changing instruction given to one vs another (EX: Does venting your anger decrease aggressive behavior?)
What is an Invasive type of manipulation?
Creating physical changes in participants (EX: Does this pill make you smarter?)
What is a manipulation check?
It is used to see if the manipulation of IV worked. It is concerned with construct validity.
What makes a good DV?
Reliable and valid, sensitive, no ceiling/floor effects
What does it mean if a DV is reliable and valid?
How well the researchers measured the DVW
What does it mean if a DV is sensitive?
Does the measure pick up on subtle differences in the outcome of interest (EX: 5kg vs. 1kg vs. 0.5kg)
What is a ceiling effect?
Everyone scores on the high end of the DV. Usually means the test or measure was too easy or not sensitive enough
What is a floor effect?
Everyone scores on low end of the DV. Usually means
What are control variables?
They are variables that the experiment holds constant on purpose potential third variable research wants to rule out
What are comparison groups?
Groups in an experiment whose levels of the IV differ from those of another group in some meaningful way
What is a control group?
The group intended to represent “no treatment”
What is a placebo group?
A group that believes they receive something when they don’t
What is a design confound?
Confound introduced due to design of an experiment
What are selection effects?
When the kinds of participants in one level of the IV are systematically different from those in the other (introduces confounds)
How do we avoid selection effects?
Random Assignment, Matched group design
What is random assignment?
Each participant has an equal chance of being in each of the IV conditions
What is a matched group design?
Design that attemps to measure a potential confounding variable in participants, match up participants w/ similar levels on that variable, randomly puts one subject from each pair in separate groups
What are the two types of between-subject designs?
Post-test Only + Pretest/Posttest
What is a post-test-only design in a between-subject design?
Participants are randomly assigned to IV groups, then DV is measured once
What is a pretest/posttest design in a between-subject design?
Participants are randomly assigned to IV conditions and DV is measured TWICE! Once before exposure, once after
What are the two types of within-subjects design?
Repeated measures + Concurrent measures
What is a repeated measure in a within-subject design?
Participants are measured on DV more than once (after exposure to each level of IV)
What is a concurrent measure in a within-subject design?
Participants exposed to all levels of IV at roughly the same time. The DV is only measured once
What are the advantages of using a within design over a between?
1). You know subjects in thte 2 groups are the same in terms of subject variables, because they are the same subjects! 2). Less statical noise (due to difference in people) -> more precise estimates -> more powerful. 3). Requires fewer participants (EX: You have a w/n subjects IV w/ 3 levels. You want 50 participants in each level. How many part. do you need total? 50! If it was b/w it would be 150)
What are the disadvantages of using a w/n subjects design?
Not always practical or possible, order effects
What are order effects?
When being exposed to one condition changes how participants perform on later conditions. Prevent by counterbalancing (preventing levels on IV in different sequences)