First Test Flashcards
Dependent Variable
(Response Variable, Measured Variable):
Variable measured to see if manipulation was successful.
measured across the experiment – different between conditions
Independent Variable
(Predictor Variable, Explanatory Variable):
Variable manipulated in order to produce changes in the dependent variable.
Often involves different levels (or conditions) which are then compared to evaluate manipulation.
Independent variables: manipulated between conditions – predict change in the DV
Dr. Chocula thinks that dark chocolate is superior to milk chocolate. He rounds up a sample of volunteers (n=20), and has them consume some dark chocolate. Afterwards, he asks them to rate how delicious the chocolate was on a 1 to 10 scale (x̄=10; s=1). He then asks them to consume and rate some milk chocolate (x̄=8.78; s=1.3).
What is DV and IV?
DV: Rating of deliciousness is being measured
IV: Type of chocolate is being manipulated.
In this case there are two levels within the independent variable (dark chocolate, milk chocolate).
Dr. Palmer wants to determine whether Steelers fans like beer more than Bengals fans. He administers an attitude scale that measures how much a person likes beer to a group of Steeler fans and to a group of Bengals fans.
What is the DV and IV?
DV: Fondness for beer is being measured
IV: Team affiliation is being manipulated.
In this case there are two levels within the independent variable (Steelers fans, Bengals fans).
Dr. Pharma wants to see if a new nicotine patch will help smokers stop smoking at a greater rate than nicotine gum, or no nicotine substitute. Her colleagues, who work in a smoking cessation clinic conduct a study to determine cessation rates at both two months and one year after an initial attempt to quit with each option.
What is the DV and IV?
DV: Cessation rates are being measured.
In this case we have two conditions for this dependent variable (two months, one year).
IV: Nicotine substitute is being manipulated.
In this case there are three levels within the independent variable (nicotine patch, nicotine gum, no substitute).
Methods and Stats
A set of tools we use to conduct a scientific psychology.
Our tools for asking and answering questions by testing hypotheses.
The “scientific method”
A way of asking and answering questions by testing hypotheses
Where do good research questions come from?
Research questions about human (or animal) behavior come from observations of behavior!
good research questions always take into account the results of past studies
literature searches on the topic of your interest are always necessary
psychological research never takes place within a vacuum
Clinical:
Abnormal or pathological functioning (e.g., PTSD, autism, etc.)
Social:
Influence of other individuals (e.g., attitudes, attraction, etc.)
Neuropsych:
Role of nervous system and brain (e.g., localization of function)
Industiral-organizatonal
Success and satisfaction in the workplace (e.g., training programs)
Cognitive-perceptual:
Internal mental processes (e.g., memory, attention, etc.)
Forensic
Understanding criminal behavior (e.g., expert witness, profiling, etc.)
Induction:
What is it?
what’s the problem of induction?
Reasoning from specific instances (observations) to general principles (theory)
thousands of observations of birds flying → conclude that all birds can fly
what’s the problem of induction? it only takes one . . .
Deduction:
What is it?
Reasoning from general principles (theory) to specific instances (observations)
start with a theory: all humans are mortal (given I am a human) → conclude I am mortal
Positive test bias:
Tendency to look for confirmatory evidence
Validation:
Attempt to gather evidence to support/confirm a hypothesis
Falsification:
Attempt to gather evidence to refute/disconfirm a hypothesis
Qualification:
Attempt to define the boundary conditions that support/refute hypothesis
What is research design?
Configuration of variables involved – how they are manipulated and/or measured
All good research designs should have . . .
1)
2)
3)
4)
1) Reliable and valid measurements of the dependent variable of interest
2) Good measurement/manipulation of the predictor or independent variable
3) Controls in place to eliminate (or assess the likelihood of) alternative explanations
4) Real opportunities to get patterns of results that dis-confirm the hypothesis (falsification)
An experiment allows a research to…
exert control to rule out alternative explanations.
Correlation does not imply…
causation
3 Important rules of correlation?
1) The cause must precede the effect
2) The cause must be present when the effect occurs
3) The cause must not be present when the effect does not occur (no “third variables”)
What are some “threats to internal validity”
practice, fatigue, demand characteristics, experimenter bias,
selection bias, history, maturation, differential attrition and many more
Extraneous variables:
Other variables – may or may not affect the dependent variable
Confound variables:
Other variables that vary systematically with the independent variable
create the possibility for alternative explanations of the effect
Between-subject designs:
Also what is a problem with this design?
Different groups experience different conditions
Problem: Differences between groups other than the experimental factor
Within-subject designs:
Also what is a problem with this design?
Same group experience different conditions
Problem: possibility of order or carry-over effects
Factorial/mixed designs:
Also what is a problem with this design?
Multiple variables manipulated (some between, some within)
Problem: gets complicated in a hurry, same problems as single factor designs and may reduce external validity
Random assignment:
Assign participants to experimental conditions at random (particularly important for between-subjects designs!)
Matching:
Pair participants in one group to participants in another on some control variable (particularly important for quasi-experimental designs!)
Counter-balancing:
Use all possible orders of treatment conditions (particularly important for within-subjects designs!)
