View article

The time series quasi-experiment was proposed initially by Campbell (1963) as a means of assessing the causal effect of a discrete intervention on a social science time series. Although the design had been widely used in the physical sciences, social science applications presented idiosyncratic inferential problems. Addressing these problems as threats to internal and external validity, Campbell (1963) concluded that the time series design was the strongest of the common quasi-experimental designs. Addressing threats to statistical conclusion and construct validities as well as internal and external validities, Cook and Campbell (1979) subsequently concurred with the optimistic validity assessment for the design.