by Frank Farach, Staff Scientist

I recently returned from the annual meeting of the Association for Psychological Science in Washington, D.C., where I attended a presentation by Dr. Cuthbert, chair of the National Institute of Mental Health’s (NIMH) Research Domain Criteria (RDoC), on the development and future of RDoC. As promised, I’m back to share what I learned about this initiative and how it is likely to affect your research.

The purpose of RDoC is to “develop, for research purposes, new ways of classifying mental disorders based on dimensions of observable behavior and neurobiological measures” (NIMH, 2008). Dr. Cuthbert noted that RDoC research is fundamentally different than traditional research. Whereas the latter is often organized around diagnostic categories and only one or two levels of analysis, RDoC is formally agnostic to diagnostic categories and encourages the examination of constructs from multiple levels of analysis (see the RDoC matrix). He offered several tips to help researchers plan RDoC-style research:

  • Select a broad sampling frame and include any necessary control groups. A sampling frame encompassing an entire chapter of the DSM-5, plus controls, would be just about right. Example: all patients presenting in an anxiety disorders specialty clinic, plus a control group of community controls.

Broad Sampling Frame Data Management Practices

  • Study independent and dependent variables at different units of analysis from the RDoC matrix. Example: Examine how individual differences in startle reactivity to aversively conditioned stimuli relate to self-reported fear, genetic variability, and treatment response.
  • In treatment (therapy or drug) development, include potential mediators and moderators of treatment outcome that are aligned with RDoC constructs. Devise and study interventions intended to modify an RDoC construct, then evaluate the extent to which the intervention affects the construct and the extent to which change in the construct affects outcome. Also be sure to measure moderator variables that may be related to the magnitude or direction of the treatment effect (e.g., presence or absence of a risk factor). Such designs can help research move systematically from asking “does it work?” (efficacy) to “how does it work?” (mechanisms) to “for whom does it work best?” (moderators). This advice dovetails with the NIH’s focus on precision medicine.

These tips are useful, but they raise many new questions. In my next post, I’ll share my analysis of how the RDoC initiative will affect data management practices and suggest how you can become better prepared for them.