Custom Search

Tuesday, November 9, 2010

Quasi-experimental method

Quasi-experimental method
In quasi-experimental studies the independent variable is not (or cannot be) manipulated as such, and so assignment to experimental groups cannot be random. The fact that no manipulation occurs interferes dramatically with our ability to make conclusive causal inferences. Examples of independent variables that cannot be manipulated by an experimenter include gender and age. Obviously experimenters cannot change the gender or age of participants, but they can compare the responses of groups of people with different ages or of different genders. Compared to the experimental method, there is no real control over the independent variable, so we cannot conclude that it is necessarily responsible for any change in the dependent variable. On this basis, as we will see, the quasi-experimental method actually has more in common with survey methodology than with
the experimental method. It has all the weaknesses of the experimental method, but it lacks the main strength. In practice, it is often conducted in conjunction with the experimental method. For example, in our learning study we might compare the effect of the new training method on both men and women.

No comments: