The MMPI-2 restructured clincial scales (RC) and restraints to innovation - summary of an article by Stephen and Kamphuis

Stephen E. Finn & Jan H. Kamphuis (2006) The MMPI–2 Restructured Clinical (RC) Scales and Restraints to Innovation, or "What Have They Done to My Song?", Journal of Personality Assessment, 87:2, 202-210

Introduction

The RC scales are a set of nine non-overlapping scales designed to measure the common factor (Demoralisation) and unique components of the eight original Clinical Scales.

The RC scales are robust.

Conceptual and methodological flaws?

Comparing like with like: multivariate scales and ‘syndromal fidelity’

A loss of ‘multivariate structure’ is a critique of the RC scales.

There are several reasons why unifactorial scales may be preferred over syndromal scales: 1) syndromes are generally not true traxa and go through changes in definitions over time. Unifactorial scales have better chance of being useful over time 2) Syndromes consist of subdomains, which are better assessed by multiple-separate measures than by one composite amalgam.

Confusing the Nature of the clinical scales, content scales, RC scales and the DSM-IV

The DSM-IV criteria define a disorder, whereas the MMPI-2 clinical scale items are heterogeneous fallible indicators of syndromes.

The main thing the RC and content scales have in common is their high internal consistency and subsequent face validity.

Item composition and the ‘credibility’ of the RC scales

The RC scales were not a priori meant to be the same as the clinical scales. The idea was to use factor analyses to reshuffle the MMPI-2 items such that eight non-overlapping scales would emerge that captured unique components of the original Clinical scales.

Unfortunate omissions?

Other purely empirical ways to solve the Clinical scale covariation problem

There are other, possibly superior ways to attack Clinical scale covariation. It is an empirical question whether these are better.

On the nature of DEM and Construct Drift

Dem is the theoretically inspired and depressively biased marker that simultaneously overextracts and underetracts unwanted variance from the Clinical Scales, in turn resulting in construct drift in the next step of the RC Scales’ derivation. Construct drift is relative to some other target construct. The RC scales describe related but different constructs than the original clinical scales.

Redundant? Restructured? Clinical?

The RC scales show meaningful relations to the Clinical scales but are not identical nor ‘proxies’ for the Clinical scales nor were they meant to be. They aim to represent the substantive core component.