Thus the study results may still be confounded by possible differences in those unobserved characteristics between filgrastim and pegfilgrastim groups.
This term is difficult to interpret because it gathers the effects of the unobserved differences in returns and differences in the means of unobserved variables and also of the differences in the returns on omitted variables, such as discrimination effects [ 17].
Gravelle et al. (2010) highlighted that the difference between the regression constants for males and females can arise from unobserved differences in returns, differences in the means of unobserved variables and differences in the returns on these omitted variables [ 17]. 2. This term is the difference between the estimated coefficients of men relative to the reference (male advantage if >0). 3.
Figure 2 Preferences and group differences in the means of unobserved variables.
8Note that nothing essential changes in our conclusions if we also allow for preference-based discrimination and/or perceived differences in the mean of unobserved variables.
A more problematic case, which is the focus of this paper, is statistical discrimination due to perceived group differences in the variance of unobserved variables.
The reason is that when employers act upon perceived group differences in the variance of unobserved variables the degree of discrimination depends on how the experimenter sets the level of qualifications in the job applications.
The next section explains the issue with perceived group differences in the variance of unobserved variables and the level of standardization (henceforth the HS critique, since it was first discussed in the seminal paper by Heckman and Siegelman, 1993).
Neumark applies his method to the data in the seminal correspondence study conducted by Bertrand and Mullainathan (2004) and finds indicative evidence of that the degree of discrimination depends on perceived group differences in the variance of unobserved variables.
In cases of imbalance, once the analyst conditions on a set of observable confounders, it is assumed that there are no differences in the distributions of unobserved confounders which are correlated with those that are observed [13].
More problematic is that, despite being an unbiased measure of what the law defines as discrimination, which includes the case when employers act upon perceived group differences in the variance of unobserved variables, it may not be very informative.
