By Andy Potter – June 2025, fresh off the Delft Plenary floor
Let’s talk acronyms. No, not ISO, not XML—though we’ll get there. I’m talking about DDI: the Data Documentation Initiative. If you work with metadata, long-term data stewardship, or records related to statistical or research activities, it’s time to put this one on your radar.
You might be thinking, "I’m a records manager, not a data architect. What does this have to do with me?" Stick with me.
A Standard Is Born—From the 1990s to Now
DDI has been around since the floppy disk era—1995 to be exact. What started as an academic metadata project at UC Berkeley (remember OSIRIS?) has evolved into a robust, multi-model framework managed by the DDI Alliance.
The original DDI-Codebook (hello, XML DTDs!) described single social science studies: variables, questions, methodologies—study-by-study. Fast-forward a few years, and national statistical offices needed more: reusability, lifecycle coverage, and cross-domain integration.
Enter DDI-Lifecycle and the newer DDI-CDI (Cross-Domain Integration). These models support the entire data lifecycle and play nicely with standards like ISO/IEC 11179 and UNECE’s GSBPM. We're talking serious metadata muscle.
Why Records Professionals Should Pay Attention
DDI speaks our language: data provenance, variable description, interoperability, preservation.
Records and information professionals deal with:
Describing and preserving context
Managing data across systems
Supporting reproducibility and transparency
Implementing model-driven architectures (hello, PIM and PSM!)
Sound familiar?
DDI offers a structured, standards-based approach to these challenges. Whether you’re managing data from scientific studies, government surveys, or machine-learning pipelines, DDI’s “variable cascade” (yes, it’s a thing!) gives you building blocks to model and explain datasets logically and consistently.
And now, it's getting serious attention at the international level.
Enter ISO: The PAS Proposal
At the recent ISO TC 46 Plenary in Delft, a proposal to standardize DDI through a Publicly Available Specification (PAS) took center stage.
Presented by metadata veterans Dan Gillman and Arofan Gregory, the PAS aims to harmonize the core, unchanging features shared across DDI-Codebook, DDI-Lifecycle, and DDI-CDI. These include the variable description model, data lifecycle alignment, and the divide between logical and physical modeling (a.k.a. PIM vs. PSM). The proposal is now at ISO Stage 20.00—formally registered and in early development.
Why push for ISO status? Visibility and trust. Just like SDMX (ISO TS 17963) before it, a DDI standard can help national governments, UN agencies, research institutions, and commercial data users adopt more consistent, interoperable practices.
What’s Next—and How You Fit In
If you’re working on digital preservation, metadata strategy, or building bridges between archives and data repositories, DDI might be just the toolkit you need.
Thinking about structured metadata for AI training data? DDI’s variable model and provenance tracking could help.
Trying to align archival systems with statistical workflows? DDI already reflects lifecycle models like GSBPM.
Frustrated by siloed datasets with no context? DDI thrives on context.
As the PAS moves forward in TC 46/SC 4 (Technical Interoperability), we can expect to see broader engagement across various sectors. Now’s the time for records professionals to bring their voices to the table. Because metadata isn’t just for databases—it’s the lifeblood of transparency, accountability, and reusability.
Let’s ensure this standard addresses our needs as well.
Curious to learn more? Drop me a line or reach out to the DDI Alliance.
And remember: it’s not just about the data. It’s about what the data means, where it came from, and how we make it last.