I want to examine whether three scales purported to measure the same
underlying construct actually do so (construct equivalence across
instruments).  My initial inclination would be to use a
Multitrait-Multimethod Matrix approach or Confirmatory Factor
Analysis; however, I would prefer to yield this question to more
seasoned researchers.

My question is:  What would be the best method to address this issue?

As a side note, the instruments chosen have exhibited evidence of
multidimensional factor structures from prior EFA analysis.