mt2 Flashcards
(52 cards)
what is a frequency claim
describes rate or degree of a SINGLE VARIABLE
what is a mediator
intermediate variable between cause and effect.
remove mediator = no effect
what is a moderator
variable that changes the sign or strength of IV on DV
what is a Confounder
Variable that is related to both IV and DV. Not part of causal mediation process.
usually causally related to DV but cannot be the effect of the IV but can cause IV
What is internal validity
ability to claim that DV change is a result of IV change only
What are the categories of threats to internal validity
threats from experimental procedures
threats from treatment or manipulation (researchers)
threats related to participants
What measures can be taken to ensure internal validity
RCT with double blind procedures
validated measures
identification of confounders
what is external validity
how well the results apply to those outside the original study.
what are the types of external threats to validity
selection of too specific of a group. difficult to generalize to others
setting is too difficult to replicate for the same results
The timing of the treatment creates a result that may not be possible in the future
Which types of validity are most important in a qualitative study
usually internal then external.
however, in applied research and clinical trials, external may be more valuable.
what are the 4 criteria to claim causation
1) correlation
2) ensure isolation using controls
3) effects following cause
4) manipulation of cause results in manipulation of effect
what is an association claim
correlation
what is a frequency claim
a count
what is the difference and similarity between pre experimental and quasi experimental design
both do not use random assignment but quasi uses a control group while pre does not
why might a pretest post-test randomized group design not be used
if pretest impacts outcome
what is the Solomon four group design
a study design that can examine the effect of pretest on the final outcome
why might a quasi experimental design be used over an experimental design
higher external validity, ethical and practical limitations
what is a pre experimental research design
examining only 1 group and providing an intervention. no control group
what is a sample
a representation if the population chosen to be studied
what is the law of large numbers
large numbers of trials will produce outcomes aligning with theoretical probability.
what is probability sampling
random sampling that is statistically likely to be representative of the population
what is non probability sampling
non random sampling that does not ensure that all possible subjects have an equal change of being chosen
What are the types of random probability sampling
simple random- using random numbers to select
systematic- selects using a regular interval (ex. every 5th person). This is more useful for large lists.
what is stratified sampling
overall group is divided into subgroups which are then selected from. This is to ensure a full distribution of certain characteristics
