Chapter 13 Flashcards
cross-sectional comparison
comparisons to see if both comparisons lead to the same conclusion (comparing different points in time)
interrupted time-series design
introducing the stimulus during the period of assessment over time in order to evaluate its effect on the variables
matching
problem is that even if matched on pre-test, impossible to determine pre-existing group differences
one-group pretest-posttest design
based on within-individual treatment comparisons, treatment effect is average difference between pre+post test
panel design
“the same families were reinterviewed regularly” // possible to document variations
pretest-posttest nonequivalent control group design
same as static design, but each group is given pre+post test to measure dependent variable at multiple points in time
quasi-experimental designs
when one or more independent variables are manipulated but participants are not randomly assigned to levels of the manipulated variables
regression toward the mean
ahen the experimental manipulation is applied to groups of people, who are extreme on the variable of interest, and who subsequently tend to regress toward the mean
replicated interrupted time-series design
comparisons between treatment and control conditions can be made both within the same participants over time and between diff groups of participants over time
static-group comparison design
assignment by naturally occurring groups X or self-selection
undermatching
matching on variables known to be associated with the dependent variable fails because you can’t know if you matched on enough variables to represent the population
