Horizontal and vertical stratification of occupational positions

Abstract

It is good practice to test the validity of a new tool or a new operationalization of an old tool before it is applied in actual research. Validating a measure consists of testing whether it reflects the underlying theoretical logic (criterion validity) and at the same time accounts for meaningful horizontal and vertical differences with regards to several socio-demographic outcomes (construct validity) in the United States and Germany (Evans & Mills, 1998). Despite what the testing language seemingly implies, validity cannot be proven once and for all in social sciences because, firstly, we work with one proxy for several latent properties and, secondly, both our measure and what should be measured may change over time or even through the very act of observation.