Experimental Design Flashcards
moderating variable
changes the intensity of a relationship
mediating variable
provide a link between two variables
in between the dependent and independent variables
cohort study
people are grouped based on some principle and followed overtime
longitudinal prospective analysis
cross sectional study
sample a population and measure various things about them then look for correlations
snapshot of a population at a certain time
ex. opinion polls
internal validity of a study
how well we can draw conclusions from the data
can we determine that this caused this
external validity of a study
how well the study predicts real world outcomes
generalizability
test validity
how well design tests what it intends to
what is the concept of represnetativeness
choose a sample that doesn’t deviate from the overall population
what is social desirability bias
tendency to reply in a way that seems more socially acceptable
what is acquiescence bias
tendency to answer yes to a. question by default
Karl LOL
longitudinal study how is it different from cross sectional
multiple measures are made over time - addresses the limitation of cross sectional being unable to provide info about change over time
ex. see if high blood pressure predicts heart disease later in life
what is the relation between internal and external validity
as we increase the internal validity by controlling for more confounding variable the external validity decreases as things are not that controlled in real life
and vice versa
what is construct validity
does our test assess the construct we have designed
ie if we want to test athletic ability does it assess that
what is content validity
does it cover the full scope of content
ie testing athletic ability we need to do more than just lift weights as there are many other aspects of athletes
what is criterion validlity
how well the test correlates with other well respected criterion
if our results of athletic ability don’t correlate well with performance of athletics then we have low criterion validity
what is predictive validity
how well our test will predict later scores
ie. hockey camp predicts how well they will do in the NHL if they are the top player at the camp and end up being a top player in the NHL it has good predictive validity
what is self reporting bias
consequence of allowing respondents choose their own answer such as social desirability bias or acquiescence bias
how is reliability related to precision
both mean consistent
reliability looks at the study as a whole producing similar results
how is accuracy related to validity
accuracy is the closeness to the actual value so we would expect a valid experiment to produce accurate results
how can a study be unreliable but valid
the average of the inconsistent results is close to the real value
ie one sky diver lands 10 ft to the right and the other 10 ft to the left the average will be at the right spot
difference between meta analysis and systematic review
A meta-analysis is a study in which previously-published data is pooled and re-analyzed.
A systematic review is a structured, critical analysis of the previously-published scientific literature on a given topic
what type of study is looking at medical records
retrospective
not cross sectional because it can be tracked over time
what is Operationalization
making a variable measurable
what are the most important ethical principles to be aware of for experiments
need for informed consent
research involves benefits vs risks - the study must have some sort of value and scientific validity
is it an ethical problem that harm come to the subjects
no this may be necessary as long as they provide consent and the results are worth the discomfort this is ok!!!
