Research Methods-Exam 1 Questions
With Well Elaborated Solutions
Advantages of pure experimental - ANSWER -Ability to establish cause and effect
relationships
Causality: IV & DV are correlated, no confounding variables.
Change in A is the only cause in change B.
Change in B cannot change without A changing first
Experimental Manipulation - ANSWER -Independent variable
eliminates confounding variables
Random Assignment - ANSWER -All participants in a sample have an equal
chance of being in any of the conditions of the study
-done by using some kind of random process
-controls for confounding variables/threats to internal validity
-should produce equivalent groups(but doesn't always)
Group Equivalence - ANSWER -Determine if theres major differences in age,
gender, SES, etc. across all groups created
The bigger the group, the better the chance they're equivalent
always investigate if groups are equal after randomly assigning conditions
Matching - ANSWER -to ensure equal distribution of characteristics across
groups
participants are matched based on presentation of characteristics and then random
assignment is performed
happens BEFORE the study begins
match things like gender, sleep, age, etc.
EX- study on 2 new teaching methods. kids matched according to reading level,
then each randomly assigned to condition, so testing is equal.
Between subject Design - ANSWER -participants can be a part of the treatment
group or the control group (but not both)
Within Subject Design - ANSWER -Each person acts as their own
control/baseline
, requires fewer participants than between
you need to measure DV at least 2 times (longitudinal)
the more the DV is measured, the more you know about the nature of change over
time.
each person experiences all IV's
Threats to within subject design - ANSWER -Testing/practice effect
Fatigue
Attrition
Maturation Effect
Order/ Sequence Effect
Testing/practice effect - ANSWER -Taking tests more than once can skew results
because participant has become familiar with the test
fix: fewer measurements or measure the same construct with different measures
Fatigue - ANSWER -Too tired to give accurate results
fix: fewer measures or more time between tasks
Attrition - ANSWER -Participant drops out before the study is over
fix: biased sample, find out reasons they quit, longer studies have higher drop out
rates
Maturation effect - ANSWER -natural changes in behavior as participant ages
you don't know if the results are because of natural changes or because of the
experimental manipulation
ex- did the child read better because he's older or because of the study?
Order/sequence effect - ANSWER -the arrangement(order) of treatment impact
responses
Potential confounding variable in study,where subject is exposed to multiple
conditions.
fix: Counter balancing: random order for each participant
Latin Square: assures a subset of treatment order possibilities are presented to an
equivalent number of participants
ex- ABC, CAB, BCA
Mixed Design - ANSWER -Combines aspects of between & within subject
designs
With Well Elaborated Solutions
Advantages of pure experimental - ANSWER -Ability to establish cause and effect
relationships
Causality: IV & DV are correlated, no confounding variables.
Change in A is the only cause in change B.
Change in B cannot change without A changing first
Experimental Manipulation - ANSWER -Independent variable
eliminates confounding variables
Random Assignment - ANSWER -All participants in a sample have an equal
chance of being in any of the conditions of the study
-done by using some kind of random process
-controls for confounding variables/threats to internal validity
-should produce equivalent groups(but doesn't always)
Group Equivalence - ANSWER -Determine if theres major differences in age,
gender, SES, etc. across all groups created
The bigger the group, the better the chance they're equivalent
always investigate if groups are equal after randomly assigning conditions
Matching - ANSWER -to ensure equal distribution of characteristics across
groups
participants are matched based on presentation of characteristics and then random
assignment is performed
happens BEFORE the study begins
match things like gender, sleep, age, etc.
EX- study on 2 new teaching methods. kids matched according to reading level,
then each randomly assigned to condition, so testing is equal.
Between subject Design - ANSWER -participants can be a part of the treatment
group or the control group (but not both)
Within Subject Design - ANSWER -Each person acts as their own
control/baseline
, requires fewer participants than between
you need to measure DV at least 2 times (longitudinal)
the more the DV is measured, the more you know about the nature of change over
time.
each person experiences all IV's
Threats to within subject design - ANSWER -Testing/practice effect
Fatigue
Attrition
Maturation Effect
Order/ Sequence Effect
Testing/practice effect - ANSWER -Taking tests more than once can skew results
because participant has become familiar with the test
fix: fewer measurements or measure the same construct with different measures
Fatigue - ANSWER -Too tired to give accurate results
fix: fewer measures or more time between tasks
Attrition - ANSWER -Participant drops out before the study is over
fix: biased sample, find out reasons they quit, longer studies have higher drop out
rates
Maturation effect - ANSWER -natural changes in behavior as participant ages
you don't know if the results are because of natural changes or because of the
experimental manipulation
ex- did the child read better because he's older or because of the study?
Order/sequence effect - ANSWER -the arrangement(order) of treatment impact
responses
Potential confounding variable in study,where subject is exposed to multiple
conditions.
fix: Counter balancing: random order for each participant
Latin Square: assures a subset of treatment order possibilities are presented to an
equivalent number of participants
ex- ABC, CAB, BCA
Mixed Design - ANSWER -Combines aspects of between & within subject
designs
