TY - JOUR
T1 - Meta-Analysis of the RDoC Social Processing Domain across Units of Analysis in Children and Adolescents
AU - Clarkson, Tessa
AU - Kang, Erin
AU - Capriola-Hall, Nicole
AU - Lerner, Matthew D.
AU - Jarcho, Johanna
AU - Prinstein, Mitchell J.
N1 - Funding Information:
None of the authors have any conflicts of interest to disclose. Matthew D. Lerner, Ph.D was supported by NIMH R01 MH110585, the Simons Foundation (SFARI Number 381283), and a NARSAD Young Investigator Award (Number 24890) during the preparation of this manuscript. Johanna Jarcho, Ph.D. was supported by the National Alliance for Research in Schizophrenia and Affective Disorders (NARSAD) Young Investigator Award: Ellen Schapiro & Gerald Axelbaum Investigator during the preparation of this manuscript. Given that the NIMH has specifically directed more funding to ?RDoC-informed research,? it is important to evaluate whether funding by the NIMH impacts reported levels of correspondence (i.e., the likelihood to report findings in support of RDoC). If differences in levels of correspondence exist between NIMH-funded studies compared to studies funded by non-NIMH agencies, this could suggest a bias in reporting or an improved design of NIMH-funded studies for explicitly testing correspondence between RDoC units of analysis. Such result would be useful for determining the impact of NIMH?s funding mission in promoting research consistent with the aims of RDoC and guiding decisions of future funding priorities by the NIMH. Additionally, by adding funding as a moderator of interest, we investigate how NIMH funding priorities may affect correspondence between different units of analysis on the RDoC matrix which could help examine the usefulness of each unit pair within the RDoC matrix when correspondence is explicitly being tested. Funding by the NIMH or the NIH versus other agencies did not moderate correspondence between subjective-by-behavior, subjective-by- circuit, or subjective-by-gene unit pairs. Thus, despite the NIMH?s explicit goal of funding studies designed to assess RDoC correspondence, these studies do not find greater correspondence between units of analysis than studies funded by any other agencies. However, it is possible that only a limited number of studies funded under the RDoC initiative have been published given that the RDoC framework was only established in 2010. Therefore, future examination of the literature when RDoC study publications are more abundant should be conducted to replicate our null results. The authors would also like to acknowledge Dorothy McGee and Kristy Lee for their work conducting the literature for this manuscript.
Publisher Copyright:
©, Copyright © Society of Clinical Child & Adolescent Psychology.
PY - 2020/5/3
Y1 - 2020/5/3
N2 - Objective: This meta-analysis advances a framework to understand correspondence among units of analysis of the social processing construct within Research Domain Criteria (RDoC). Method: As requested for this special issue, eligible studies cited an RDoC-initiative paper or mentioned RDoC in the abstract, title, or keywords were empirical and peer reviewed, and described a correlation or regression analysis (r, β, or odds ratio) between two different units of analysis in the social processing domain in youth. We examined the frequency (descriptive statistics) and magnitude of correspondence between unit-pairs (random effects models), and predefined moderators (meta-regression). Results: Eight of the twenty-eight possible unit-by-unit pairs were identified, with subjective-by-behavior units being the most common. Of those, only subjective-by-circuit had significant correspondence between units. Moderator analysis revealed that the age and diagnosis of generalized anxiety disorder moderated correspondence between subjective-by-circuit units of analysis, and that a diagnosis of autism spectrum disorder moderated correspondence between subjective-by-gene units of analysis. Younger ages and inclusion of either diagnostic group reduced correspondence. Conclusions: These findings indicate that the RDoC initiative has generated limited research within the social processing domain across units of analysis in youth to date. Moreover, National Institute of Mental Health (NIMH)-funded studies do not appear to be biased toward supporting the RDoC framework. However, the limited number of included studies precludes the generalizability of these findings and underscores the need for further research. Despite this, results suggest that the NIMH model for providing standard batteries of measurement tools may effectively reduce spurious correlations between subjective-by-behavior units of analysis.
AB - Objective: This meta-analysis advances a framework to understand correspondence among units of analysis of the social processing construct within Research Domain Criteria (RDoC). Method: As requested for this special issue, eligible studies cited an RDoC-initiative paper or mentioned RDoC in the abstract, title, or keywords were empirical and peer reviewed, and described a correlation or regression analysis (r, β, or odds ratio) between two different units of analysis in the social processing domain in youth. We examined the frequency (descriptive statistics) and magnitude of correspondence between unit-pairs (random effects models), and predefined moderators (meta-regression). Results: Eight of the twenty-eight possible unit-by-unit pairs were identified, with subjective-by-behavior units being the most common. Of those, only subjective-by-circuit had significant correspondence between units. Moderator analysis revealed that the age and diagnosis of generalized anxiety disorder moderated correspondence between subjective-by-circuit units of analysis, and that a diagnosis of autism spectrum disorder moderated correspondence between subjective-by-gene units of analysis. Younger ages and inclusion of either diagnostic group reduced correspondence. Conclusions: These findings indicate that the RDoC initiative has generated limited research within the social processing domain across units of analysis in youth to date. Moreover, National Institute of Mental Health (NIMH)-funded studies do not appear to be biased toward supporting the RDoC framework. However, the limited number of included studies precludes the generalizability of these findings and underscores the need for further research. Despite this, results suggest that the NIMH model for providing standard batteries of measurement tools may effectively reduce spurious correlations between subjective-by-behavior units of analysis.
UR - http://www.scopus.com/inward/record.url?scp=85075947860&partnerID=8YFLogxK
U2 - 10.1080/15374416.2019.1678167
DO - 10.1080/15374416.2019.1678167
M3 - Article
C2 - 31799882
AN - SCOPUS:85075947860
SN - 1537-4416
VL - 49
SP - 297
EP - 321
JO - Journal of Clinical Child and Adolescent Psychology
JF - Journal of Clinical Child and Adolescent Psychology
IS - 3
ER -
