Ch. 13 - Designs, Quasi-experiments, Program Evaluation

Question 1

quasi-experimental designs

Answer 1

used when random assignment is not possible or ethical.
based on pre-existing groups such as race and gender.
no random assignment, possible manipulation of IV, possible control group/condition
generally, have good EXTERNAL validity

Question 2

one-group design

Answer 2

1. one group post-test only - measures behavior after an event or treatment. no control group.
2. one group pre-test/post-test - measurement of behavior before and after event/treatment. no control group.

Question 3

non-equivalent control group post-test-only design

Answer 3

one group WITH treatment, one group WITHOUT.
no pretest
cannot assume equivalence, and can’t test for equivalence.

Question 4

pre-test/post-test non-equivalent control group design

Answer 4

measurement before and after for both groups.
only one group receives treatment
can compare changes in behaviour between groups

Question 5

quasi-experiments: threats to internal validity

Answer 5

instrumental decay/change (researcher action), history, maturation, testing, participant attrition, self-selection to groups