Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
33 Cards in this Set
- Front
- Back
- 3rd side (hint)
Experimental Method
|
-cause and effect based on random sampling.
-Hypotheses: null and alternative. -Manipulate IV(s) to affect DV to determine a possible causal relationship |
-Problems: extraneous variables, environment, etc.
|
|
Control Group
|
-Get the inactive part of the IV
-Comparison for other levels/groups |
|
|
IV vs DV
|
-IV: causal variable, at least two levels (not always enough)
-DV: result (effect) variable, sensitivity to change is important. |
-IV is controlled to measure the impact of IV on the DV.
|
|
Between Groups Design
|
-Separate groups for each level of the IV, the results are compared.
|
-2 group design=1 IV and 2 levels
|
|
Within Groups Design
|
-Every group gets each level of the IV
|
-the results are usually aggregated.
|
|
Simple Randomized Design
|
-1 IV, more than two levels
|
-McKenna
-study the effects of noise on studying -dorm noise, TV, rock 'n roll, birds/nature, silent. -Subjects studied text for 30 min, later quizzed -How well written/how informative was the text? Like text? Meet author? |
|
Correlational Research Design
|
-Use data that already exists to find a statistical correlation.
-Never know if you're missing variables -Measuring what exists |
-Bushman and Anderson (2001)
-Refuting viewing violence and acting violently -If viewing violence caused more violence, society would show it -Violent crime, non-violent, MDK RW: 13% 87% 2% TV: 87% 13% 50% |
|
Pseudo-Experimental Design
|
-A group is exposed to a treatment or condition and then observed for changes which are attributed to the treatment.
-IV exists but is not under control of researcher -1 group pre-post test |
-Static Group Comparison Design
-No premeasure only post |
|
Quazi-Experimental Design
|
-Experimental design with selected subjects based on real world exposure to variables at hand.
-Allow for more control with a manipulated IV, also at least one non-manipulated grouping variable. -Factorial in nature -Lack random assignment -no causal results |
|
|
Simple Time Series Design
|
Period of time between points of measurement.
|
-quazi-experimental, within subjects design
|
|
Equivalent Time Sample Design
|
-Equivelant times of operations (ABAB design)
-Observe - Manipulate - Observe - M (control) (experiment) (c) (e) |
-quazi-experimental
|
|
Strong Internal Validity
|
What you manipulate is affecting the outcome.
|
|
|
History Effect
|
-2 points of measurements, within subjects design
-Something historical occurs between pre/post measure |
-Ex: Study begins 9/9/2001 and ends 9/15/2001.
|
|
Maturation
|
Short term, internal to individual.
|
-Ex: getting bored, tired, hungry, etc.
|
|
Instrumentation
|
-Affects only the DV
-As a function of measuring DV, you change it -Attitude can be changed by answering the question |
-Ex: the pressure from a blood pressure cuff may affect the subject's performance.
|
|
Statistical Regression
|
-At least 2 measurements using same tool
-The tendency for extreme scores to regress toward the mean by the 2nd measurement. |
|
|
Selection Bias
|
-Anything in the selection process skewing the results?
|
-Ex: time of day, location, etc.
|
|
Mortality (attrition)
|
Any change due to subject loss from various comparison groups.
|
|
|
Experimenter Bias
|
The experimenter's own ideas influence the data.
|
|
|
Confirmation Bias
|
The experimenter only looks for data to support hypothesis(es).
|
|
|
Sequencing
|
The order of conditions may be problematic.
|
|
|
Interaction of problems
|
There may be multiple threats to internal validity.
|
-Ex: History, maturation, instrumentation, statistical regression, selection bias, mortality (attrition), subject bias, experimenter bias, and/or sequencing.
|
|
Subject Bias (Demand Characteristics)
|
The subject gives what they think the experimenter wants.
|
|
|
Latin Sqaure (counterbalancing)
|
-Helps break down every possible condition.
-Treatments are assigned at random within rows and columns, with each treatment once per row and once per column. |
-Usefull when the experimenter desires to control variation in two different directions.
|
|
Between Subjects Design
|
-The values of the DV for one group of participants (used) are compared with the values of another group of participants (never used).
|
-Ex: Used Never Used
10 ----------------4 8------------------6 14----------------5 (---- = 'compared to') |
|
Within Subjects Design
|
-Each participant gets all levels of IV, individual scores are compared
|
-Ex: Before After
SubjA: 10------------------4 SubjB: 8-------------------6 SubjC: 14------------------5 (---- = 'compared to') |
|
Non-Equivalent Control Group
|
-Find a group as close as possible to the treated group to be used as a control group.
|
-Can never be sure they are compatible
|
|
Questions to ask when reading an abstract
|
-What were the authors trying to figure out?
-What are the IV's and DV's? -were the authors trying to show causality? -How many groups did they use? -How did they assign subjects to groups? -What would worry the IRB? -Threats to internal validity? |
|
|
Instructional Independent Variable
|
-Manipulative
-Need attention of subjects -Eye contact, appropriate pauses -Leave room for questions -keep as simple/clear as possible -Impact: move subjects to respond -Avoid more than one manipulation at a time -People come in, say, "I need you to..." -Manipulation takes place during instruction |
|
|
Event IV
|
-Something happens, everything unfolds
-Any confederate is one -As long as it is played out in front of the subject it is one -create realism (such as fake newspaper which looks real) -Generally have more impact -Keeps people from guessing what the study is about because of the "surprise." -False feedback is between instructional and event IV. |
|
|
Standardization
|
-Important for consistency; every subject gets the same exposure
|
Live or recorded?
-Live gives the opportunity to make sure that people understand -Recorded gives uniformity, but not everyone processes information the same way. |
|
Cross-Lagged Panel Technique
|
-Establishes correlations over time
|
|
|
What would worry the IRB?
|
-Consent
-Risk -What is the Purpose/objective? -Debriefing -Benefits -Procedure -Subjects (recruitment, how many?) |
|