A construct validity study of the Holtzman Inkblot Technique: correlation with the MMPI-2 content scales