Higher-order factors
are not commonly discussed in the literature, but it is likely that
you will run across discussion of this topic at some point. We included
this chapter because we feel that it is not discussed enough in research
methods, and too much in the content literature.
Higher-order factors are tricky. They are
not difficult to empirically evaluate once you figure out how to perform
the analyses. The trickier issue is deciding whether there truly is
a higher-order factor or whether the original analysis should have
specified a single-factor structure and, finally, whether there is
a meaningful difference between the two. We remain convinced that
these exploratory techniques are fun, but relatively low in value
without confirmatory techniques to evaluate the competing hypotheses
in a more rigorous manner, and replication in independent samples.
Also, if you are committed to performing this type of analysis within
an EFA framework, Thompson (2004) has described in detail a more modern
methodology for presenting this information.