A Consultant's Casebook
The Measurement of Change
One of the more difficult research tasks in the social sciences is the measurement of change. In controlled, laboratory studies, measuring the effects of change is relatively easy. Outside the laboratory, however, where natural variations are the norm, it is far more difficult to decide whether any of the variables that co-vary with observed changes are a cause of the change, a result of the change, or simply co-vary with the change.
Many statisticians will tell you that regression models can include multiple variables and that the regression weights can be interpreted to develop an understanding of which variables affect the observed change(s) and to what extent they influence the change. Well, it may be possible. But in my experience, it has been difficult to determine the direction of causality. And, the statements regarding percent of variance accounted for by this or that variable leave me cold.
All material is copyrighted! Not to be used without the written permission of Andres Inn. ©2000.
Send mail to Andres.Inn@mail.ee with questions or comments about this web site.