Proposed Standards for Variable Harmonization Documentation and Referencing: A Case Study Using QuickCharmStats 1.1

Abstract

Comparative statistical analyses often require data harmonization, yet the social sciences do not have clear operationalization frameworks that guide and homogenize variable coding decisions across disciplines. When faced with a need to harmonize variables researchers often look for guidance from various international studies that employ output harmonization, such as the Comparative Survey of Election Studies, which offer recoding structures for the same variable (e.g. marital status). More problematically there are no agreed documentation standards or journal requirements for reporting variable harmonization to facilitate a transparent replication process. We propose a conceptual and data-driven digital solution that creates harmonization documentation standards for publication and scholarly citation: QuickCharmStats 1.1. It is free and open-source software that allows for the organizing, documenting and publishing of data harmonization projects. QuickCharmStats starts at the conceptual level and its workflow ends with a variable recording syntax. It is therefore flexible enough to reflect a variety of theoretical justifications for variable harmonization. Using the socio-demographic variable ‘marital status’, we demonstrate how the CharmStats workflow collates metadata while being guided by the scientific standards of transparency and replication. It encourages researchers to publish their harmonization work by providing researchers who complete the peer review process a permanent identifier. Those who contribute original data harmonization work to their discipline can now be credited through citations. Finally, we propose peer-review standards for harmonization documentation, describe a route to online publishing, and provide a referencing format to cite harmonization projects. Although CharmStats products are designed for social scientists our adherence to the scientific method ensures our products can be used by researchers across the sciences.

Forschungsdatenmanagement in der Sekundäranalyse

“…Sollen relevante Forschungsdateien der Sekundäranalyse längerfristig aufbewahrt bzw. Dritten zur weiteren Nutzung verfügbar gemacht werden, stellt sich die Frage, wie die Generierung dieser Daten am besten beschrieben werden kann (Winters/Netscher 2016;Dickmann/Enke/Harms 2010). Generell können Sekundärforschende die Maßnahmen ihres Forschungsdatenmanagements in einem projektinternen Datenmanagementplan systematisch dokumentieren (vgl.…”

Pretrained Language Models for Semantics-Aware Data Harmonisation of Observational Clinical Studies in the Era of Big Data

“…Furthermore, the absence of standards, frameworks, and journal requirements for the reporting and sharing of data harmonisation outcomes results in loss of resources, time, and effort [7]. Often, variable names and descriptions are ambiguous and inconsistent across datasets which increases the difficulty of dataset harmonisation [8].…”

Software Application Profile: Opal and Mica: open-source software solutions for epidemiological data management, harmonization and dissemination

“… 19–22 However, little software has been specifically designed to facilitate data harmonization in multi-study projects. 23 , 24 Finally, to our knowledge, no integrated solution exists to address the dual aim of disseminating epidemiological metadata to the research community and supporting data inter-operability across studies.…”