… a satisfactory fit of a latent variable model does not establish the existence of ‘real’ underlying scales. It merely shows that the scales we construct by this means are capable of explaining the associations we find in the data.
Bartholomew, D. J. B. (1998). Scaling Unobservable Constructs in Social Science. Journal of the Royal Statistical Society: Series C (Applied Statistics), 47(1), 1–13.
Validity refers to the degree to which evidence and theory support the interpretations of test scores entailed by proposed uses of tests … It is the interpretations of test scores required by proposed uses that are evaluated, not the test itself.
(Messick, 1989, p.9)
Messick, S. (1989). Validity. In R. L. Linn (Ed.), Educational Measurement (3rd ed., pp. 13–103). New York: American Council on Education and Macmillan.
Apart from any mathematical reason for such an assumption, there are psychological or substantive reasons. The proposition is that the latent quantities are the only important factors and that once these are determined behavior is random (in the sense of statistical Independence). In another terminology, the set of individuals with specified latent characteristics are “homogeneous”.
(Anderson, 1959, p.11)
Anderson, T.W. (1959). Some scaling models and estimation procedures in the latent class model. In O. Grenander (Ed.), Probability and statistics (The Harold Cramer Volume) (pp. 9-38). New York: Wiley.
Unfortunately, the meaning of alignment has been cheapened somewhat when test publishers show that all of their multiple-choice items can be matched to the categories of a state’s content standards, even if altogether they tap only a narrow subset of the intended standards.
(Shepard, 2006, p. 629)
Shepard, L. A. (2006). Classroom assessment. In R. L. Brennan (Ed.), Educational Measurement (4th ed., pp. 623–646). Westport, CT: ACE/Praeger.
