Master this deck with 95 terms through effective study methods.
Imported from Quizlet
A clear, testable statement predicting the relationship between variables.
- Clear relationship - Testable - Operational definitions - Independent & dependent variables - Declarative statement
- Null hypothesis - Confounding variable hypothesis - Causal hypothesis
States there is no relationship between variables.
States another variable caused the results.
States the independent variable caused the dependent variable.
Are statistical tests accurate?
- Unreliable measures - Violations of statistical assumptions
- Valid measures - Equal sample sizes
Is theory the best explanation for results?
Alternative explanations
Strong theory and controls
Do results generalize to broader population?
- Unrepresentative sample - Overgeneralization
- Representative sample - Clear sample description
Did independent variable cause changes?
Confounding variables
Controls
Natural changes over time
Children grow smarter
Control group
Outside event affects results
Major event during study
Control group
Practice effects
Doing better second test
Different test versions
Measurement changes
Researcher grading differently
Standardized tools
Extreme scores move toward average
Worst students improve
Control group
Groups not equal at start
Smart students in one group
Random assignment
Participants drop out
Low motivation students leave
Equal dropout rates
Groups share information
Students talk to each other
Separate groups
Order affects performance
Fatigue
Counterbalancing
Participants behave differently because they're in study
Participants respond to cues about expectations
Improvement due to expectations
Researcher behavior influences participants
Researcher treats one group differently
Measure before and after treatment
Measures change
- Many threats to validity - Weak design
Two groups measured after treatment
- Stronger design - Better internal validity
No baseline measurement
Controls applied to nearly all research to reduce threats to validity.
Creating a controlled environment for research.
Quiet room with no distractions
- Reduce internal validity threats - Increase external validity (natural setting
Using reliable and valid measures.
Using standardized tests instead of homemade tests
Repeating a study to confirm results
Same study, same procedures
Small changes to extend theory
Different operational definitions
Participant doesn't know condition
Neither participant nor experimenter knows condition
Using computers to run study
- Reduces bias - Precise measurement - Less contact with participants
Observable, measurable behaviors
Reaction time instead of happiness rating
Reduce bias
Two researchers record behavior
Hiding study purpose from participants
Balanced placebo design
Fake treatment
Sugar pill
All possible participants
Group researcher wants to study
Group researcher can access
Participants chosen from population
Everyone has equal chance
Random sampling within subgroups
Sample includes same % males and females
Random sample from accessible population
Easy-to-get participants
Randomly assign participants to groups
Assign participants in blocks
Equal group sizes
Match participants then randomly assign
- Reduces bias - Controls unknown variables - Improves internal validity - Creates equivalent groups
- Control groups - Random assignment
Group that does not receive treatment
Makes groups equivalent