Publications

Browse

Now showing 1 - 1 of 1

Research Methods in Social Relations-Quasi-Experimental Designs
(SPSSI, 1986) Kidder, Louis H.; Judd, Charles M.
Quasi-experimental designs provide a way to study some naturally occurring social treatments. They are a compromise between a true experiment that has high internal validity and the poor pre-experiments that have almost no internal validity at all. Quasi experiments enable us to rule out some threats to validity because they include more data points than the pre-experiments. The number of quasi-experimental designs that a creative researcher can construct is limitless. We have presented three types that are extensions of pre-experiments to show how the additional data points make a previously uninterpretable design interpretable. A determined researcher can design yet unthought-of quasi experiments by gathering data from enough subjects at enough times to rule out many threats to internal validity, so that even without random assignment, it will be possible to infer causes and effects.