No abstract

No abstract

In a variety of problems, it is desirable to have a single coefficient summarize the causal effects of a set of variables when other variables are controlled.The "sheaf coefficient" presented here does this and can be employed meaningfully in the context of path analysis models.AUTHOR'S NOTE: For helpful comments and suggestions concerning an earlier version of this paper, I am grateful to Arthur S. Goldberger and also to Downloaded from [148] ometimes indicators are determinants of an abstract construct, and observations on the indicators must be used to assess the relations between the construct and other variables. Psychometric theory is of little value in such problems because it is based largely on the assumption that indicators are determined by a latent construct instead of being determinants of it. Blalock (1971) considered the matter in detail and concluded that serious analytic problems may arise when indicators are determinants of the construct they are meant to measure (see also Land, 1970). In fact, Blalock gave evidence that when the problem exists in a completely uncompromised form, meaningful analyses cannot be done at all. Some special models of analysis are being devised to deal with such problems. Blalock (1971) showed that solutions sometimes are possible when at least one traditional indicator is used to measure the latent construct (i.e., an indicator that is determined by the construct rather than a determinant of the construct). Hauser and Goldberger ( 1971 ) developed a solution for problems in which an unmeasured variable is related to a dependent variable (or to several dependent variables), and it can be assumed that the unmeasured variable is uncorrelated with any of the other determinants of the dependent variable(s). The presentation here extends some existing ideas which are reviewed in a later section to give another solution applicable under special conditions to three common analytic problems, illustrated by the following examples.

No abstract

No abstract