exam 3 Flashcards
another word for between subjects
independent measures
another word for between subjects
repeated measure design
shared experiences
same person but random assignment in the order of conditions
advantages of within subjects
eliminates individual differences
no worrying about selection effects
fewer participants
disadvantages of within
experience variables
carryover effects (practice and fatigue)
how to rid of order effects
counterbalancing- flips the order to help reduce effects
complete counterbalancing
using every possible sequence
partial counterbalancing
use a subset of possibe sequences
maybe control 1st so you have a baseline then chnage the rest
latin square counterbalancing
blend between complete and partial
concurrent measures- within
doing the same thing at the same time so all levels are at the same time
threats to internal validity
confounds
selection effects
order effect
attrition
maturation
history
regression to mean
testing
instrumentation
observer bias
demand characteristics
placebo effect
confounds
did we control varaibles?
an additional variabel that systemaically varies
selection effects
random assignment?
having different types of people in conditions
order effect
carryover effects
counterbalancing
attrition
group makes up changes over time b/c participants group out
maturation
improvement over time because of natural development or improvement
is there a comparison group?
history
an external factor that affects most/all members og a group
comparison group
regression to the mean
when extreme scores at pretest change over time b/c the events that caused the scores do not recur the same way
comparison group
testing
a type of order effect: a group changes over time b/c of repeated testing
instrumentation
a group changes over time b/c the measurement instrument changes
codebook?
observer bias
researcher expectancy influences the measurments in one group
are they blind to purpose?
demand characteristics
participants guess what the study is about and change their behavior
placebo effect
improve b/c believe they are being helped
comparison group?
what is the purpose of external validity?
the extent to which the results generalize
do the results of the study apply beyond the study?
mundane realism
the extent to which the events in the study capture, mimic or happen in the real world
psychological realism
the extent to which the psychological processes are the same as they occur under normal circumstances
is internal or external validity often prioritized?
internal
can you have both internal and external validity?
yes but it’s hard to get both
exact replication
do everything the exact same
conceptual replication
still trying to answer the same research question but if i change one aspects of the EV will i get the same result?
replication plus extension
do study 1- copying it exactly
do study 2- slightly change it and see if it matches
pilot test
a study conducted to ensure that a new procedure works as planned
reasons not to ddraw the blanket conclusion
1) experiments need not to seem artificial
2) often conducted to learn about psychological processes that are likely to operate in a variety of people and situations
operationalization
the conversion from RQ ti experiemnt design
is a correlational experiment high or low in IV?
low
is an experimental deign high or low in IV?
high
most research subjects are….
WERID
western
educated
industrialized
rich
democratic countries
what causes biased samples?
1) easy to contact
2) able to contact
3) who invites themselves
b/c we tend to get extreme views
good EV
unbiased/probability
random
representative sample
unknown EV
biased samples
non-probability samples
unrepresentative
sampling characteristics
population
sampling frame
sample
representative sample- relfects the important characteristics of a pop.
non-probability sampling
convience
self-selected
quota- not random but chosen according to specific purposes/goals
snowball
probability sampling
simple random sampling-each member of pop. has equal chance of beung selected
stratified random
clsuter sampling- find ways ppl are already divided and then sample from those groups
what are the 2 types of stratified random sampling
1) proportioate- ensures the sample matches the proportion of that subgroup in the pop.
2) didproportionate- sampling a bit higher than the representation of the pop to ensure a big enough sample to draw conclusions about that aub group
what causes sampling bias?
response rate
what is response rate
% of cases who participate in a survet out of all those selected into the sample
can create non-reponses bia
what is non-response bias
attitudes/bleifs of responders differ frim thise who didnt respond
threats to external validity
sampling bias
history
experimenter/situation bias
testing effects- same as IV
sampling bias
the sample is not representative of the pop.
experimenter/situation bias
factors like setting, time of day, location etc. limits generalizability to findings
how to counter EV threats
replications- do it for other settings
field experiments-counter testing and situation effects by using natural contexts
probability sampling- counters selection bias
reasons a study might not replicate
1)null effects-there really is no effect, or first study got it wrong or maybe the thing is just too hard to measure
2) failure to adequtely control exprimental conditions
types of EV
population/sample
study, measures time, etc.
ecological- real world setting