PSYC 201W Chapter Notes - Chapter 11: Needs Assessment, Internal Validity, Measuring Instrument
Document Summary
What is a quasi-experiment: control defining feature of any experiment, control: If the best designs are used, reasonably strong causal inferences can be drawn when certain data patterns occur. If weak designs are used, clear causal conclusions can not be drawn. Diagramming quasi-experimental designs: system for diagramming quasi-experimental designs, x = treatment, o = observation (i. e. , measurement period, temporal order proceeds from left to right. Improved economic, employment, or job conditions: maturation, spontaneous remission, testing, taking pretest may have affected scores on the posttest. Instrumentation: pretest to posttest changes in the observer"s criteria for rating, regression to the mean. If participants" pretest scores were high, then on average, regression to the mean in their posttest scores would be expected: attrition. In a study where participants are not selected because they are above or below average on some attribute, regression to the mean is less likely to be a plausible alternative explanation for its results.