Ch 9: QUASI- EXPERIMENTS AND APPLIED RESEARCH

quasi-experimental

some systematic variation in exposure to different conditions has occurred between groups, but random assignment of participants to these groups was not, or could not be employed

Artifact

an artificial or spurious finding, a “ pseudo- effect” that results inevitably from the properties of the measuring instrument or from the method of data collection employed

propensity score matching

uses statistical procedures to match respondents on as many covariates

propensity score

represents the conditional probability of membership in one group ( e.g., the experimental group) over another ( e.g., the control group), given the pattern of the person’s status on the various covariates

counterfactual score

an outcome that could have happened under a different condition

case-control designs

based on data already in hand

autocorrelated errors

random events that affect the measurement obtained at one time is more likely to carry over and be correlated with measurements taken at temporally adjacent points than with observations separated by longer time delays between assessments

comparison time- series designs

combine features of an interrupted time-series design and a comparison group design

Regression discontinuity (RD) designs

test the existence of some systematic relationship between a pretest selection variable and a posttest measure

testing effects

threats to the experiment’s internal validity

maturational effects

Merely having the time to concentrate on reading speed even without experiencing the training may have led to improvements

sample selection bias

may already be motivated to improve

history effect

If some event, in addition to the treatment intervention, occurs between pre-test and post-test in one group only, then it is not clear what any group differences at post-test should be attributed to.

selection–maturation interaction

a threat to internal validity in research where comparison groups mature or change over time at different rates, making it look like a treatment worked when natural development actually caused the difference

contamination effects

despite your best efforts, the treatment and control groups influence each other in some way

time series design

having only one sample but taking measurements of the dependent variable on three or more occasions

interrupted time series design

the treatment intervention ‘interrupts’ an otherwise seamless time series of observations

instrumentation effects

changes in the accuracy or reliability of measurements over time

multiple time-series design

An extended series of data collection points are used with both the treatment group and the non-equivalent control

compensatory rivalry

control group members may compensate for not receiving the treatment by trying harder to perform well