Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
13 Cards in this Set
- Front
- Back
quasi means
|
having resemblance to
|
|
quasi exp. are
|
approximations of experiments
"near experiments" |
|
like true experiments..
|
used to estimate effects of 1 or more treatments (IVs) on 1 or more DV
|
|
unlike true experiments..
|
no random assignment (naturally occurring groups)
conducted in "real world" (applied research, external validity is emphasized) less control, less internal validity |
|
internal validity
|
strength of inference that TV caused change in DV depends on type of design
|
|
When use quasi-exp?
|
1. when random assign not feasible
2. even if feasible..what happens in real world 3. ethical considers may preclude random assign |
|
Single-Group Pretest-Posttest
|
One group, pretest/posttest
OXO subject to numerous interal validity threats little/no control no comparison group, change from pre to post could be due to anything inference for effect of treatment is weak |
|
Nonequivalent Groups Pretest Only Group
|
2 groups
(1 groups recieves treatment, other doesnt. compare control and exp't groups scores) XO O some control potentially stronger inference due to comparison with control group (depends on similarity of 2 groups) |
|
Nonequivalent Groups Pretest-Posttest Design
|
2 groups
compare amount of change by comapring differences in groups' pretest and posttest scores OXO OO more control, stronger inference less potential for other factors causing differences in the DV |
|
Interrupted Time Series Design
|
1 group
multiple pretests and posttests look for trends leading up to and following treatment OOOOXOOOO even more control eliminates more alt. explanations less likely a difference could be attributed to other causes than if only 1 pretest and 1 postest were used |
|
Interrupted Time Series Design with Reversal
|
1 group
multiple pretests and posttests, introduce and remove treatment (removal=reversal) look at difference in trends before/after treatment is introduced and removed OOOOXOOO-XOOOO maximum control if DV score return to baseline (pre-treatment levels), stronger inference |
|
limitations to interrupted time series
|
often cant remove the treatment
treatment may have effects that remain even though treatment has been removed removal may create changes not due to treatment |
|
Multiple Time Series Design
|
2 groups (1 is control group)
compare trends of 2 groups OOOOXOOOO OOOO OOOO maximum control minimizes alt. explanations depends on extent 2 groups are similar and experience similar things after the treatment |