Conceptual comparison of constructs as first step in data harmonization: Parental sensitivity, child temperament, and social support as illustrations

Marije Verhage, Carlo Schuengel, Annaleena Holopainen, Marian J. Bakermans-Kranenburg, Annie Bernier, Geoffrey L. Brown, Sheri Madigan, Glenn I. Roisman, Mette S. Vaever, Maria S. Wong, Marian J. Bakermans-Kranenburg, Lavinia Barone, Kazuko Y. Behrens, Johanna Behringer, Annie Bernier, Ina Bovenschen, Geoffrey L. Brown, Rosalinda Cassibba, Jude Cassidy, Gabrielle CoppolaAlessandro Costantini, Mary Dozier, Karin Ensink, R. M. Pasco Fearon, Brent Finger, Airi Hautamaki, Nancy L. Hazen, Elena Ierardi, Inês Jongenelen, Simo Køppe, Francesca Lionetti, Sheri Madigan, Sarah Mangelsdorf, Mirjam Oosterman, Cecilia S. Pace, K. Lee Raby, Cristina Riva Crugnola, Glenn I. Roisman, Carlo Schuengel, Alessandra Simonelli, Gottfried Spangler, George M. Tarabulsy, Mette S. Væver, Marije Verhage, Maria S. Wong, Bronia Arnott, Heidi Bailey, Patrick J. Brice, Karl-Heinz Brisch, Germana Castoro, Elisabetta Costantino

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

This article presents a strategy for the initial step of data harmonization in Individual Participant Data syntheses, i.e., making decisions as to which measures operationalize the constructs of interest - and which do not. This step is vital in the process of data harmonization, because a study can only be as good as its measures. If the construct validity of the measures is in question, study results are questionable as well. Our proposed strategy for data harmonization consists of three steps. First, a unitary construct is defined based on the existing literature, preferably on the theoretical framework surrounding the construct. Second, the various instruments used to measure the construct are evaluated as operationalizations of this construct, and retained or excluded based on this evaluation. Third, the scores of the included measures are recoded on the same metric. We illustrate the use of this method with three example constructs focal to the Collaboration on Attachment Transmission Synthesis (CATS) study: parental sensitivity, child temperament, and social support. This process description may aid researchers in their data pooling studies, filling a gap in the literature on the first step of data harmonization.

•Data harmonization in studies using combined datasets is of vital importance for the validity of the study results.
•We have developed and illustrated a strategy on how to define a unitary construct and evaluate whether instruments are operationalizations of this construct as the initial step in the harmonization process.
•This strategy is a transferable and reproducible method to apply to the data harmonization process.
Original languageEnglish
Article number101889
Pages (from-to)1-17
Number of pages17
JournalMethodsX
Volume9
Early online date26 Oct 2022
DOIs
Publication statusPublished - 2022

Funding

This work was supported by a grant from Stichting tot Steun Nederland to M. Oosterman and C. Schuengel, a grant from the Social Sciences and Humanities Research Council Canada (No. 430-2015-00989 ) to S. Madigan, a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD; R01 HD102035 ) to G. I. Roisman, and a Veni grant by the Dutch Research Council (No. 451-17-010 ) to M. L. Verhage. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the grant agencies.

FundersFunder number
Stichting tot Steun Nederland
Eunice Kennedy Shriver National Institute of Child Health and Human DevelopmentR01 HD102035
Social Sciences and Humanities Research Council of Canada430-2015-00989
Nederlandse Organisatie voor Wetenschappelijk Onderzoek451-17-010

    Fingerprint

    Dive into the research topics of 'Conceptual comparison of constructs as first step in data harmonization: Parental sensitivity, child temperament, and social support as illustrations'. Together they form a unique fingerprint.

    Cite this