The Gioia Data Structure: First-Order Concepts, Second-Order Themes, and Aggregate Dimensions

The three tiers

Gioia, Corley and Hamilton frame the first two tiers as a deliberate change of voice. First-order analysis is informant-centric — codes use terms the participants themselves would recognise, and at this stage there are usually far too many of them (fifty to a hundred is normal early on). Second-order analysis is theory-centric — the researcher steps back and asks what is happening conceptually, using constructs the informants would not necessarily use themselves. Aggregate dimensions then compress the second-order themes to the two-to-six overarching constructs that structure the eventual model.

A worked example

From a study of staff wellbeing during an organisational system change — one aggregate dimension of a larger structure:

Read it bottom-up and the logic is auditable: two informants describe unanswered feedback in their own words → the researcher names that pattern futile feedback channels → together with procedural exclusion, it describes something larger: erosion of professional voice. A full study repeats this discipline across every dimension.

Building one — the four moves

• 1. Code in informant terms — open-code transcripts staying close to participants' language. Do not compress yet — an unwieldy list is a feature of this stage, not a failure.
• 2. Reduce by similarity — collapse near-duplicates into a manageable set of first-order concepts (published studies typically land at roughly 25–40), keeping the informant voice in the labels.
• 3. Move to the theoretical voice — group first-order concepts into second-order themes by asking what is going on conceptually. This is the interpretive leap — and the layer examiners probe hardest.
• 4. Distil aggregate dimensions — compress second-order themes into the few overarching constructs that will structure your model. The data structure is the static snapshot; the grounded model you build from it shows how the dimensions relate dynamically.

The citation everyone gets wrong

The methods paper is: Gioia, D. A., Corley, K. G., & Hamilton, A. L. (2013). Seeking qualitative rigor in inductive research: Notes on the Gioia methodology. Organizational Research Methods, 16(1), 15–31.

It is routinely misattributed to the Academy of Management Journal — an easy mistake, since many of the best-known studies that use the method appeared in AMJ and its peer journals. If your reference list says AMJ 2013, fix it before your examiner does.

Common failure modes

• Second-order labels that are just summaries — if a theme label could have come from an informant, it hasn't made the move to the theoretical voice yet.
• Too many aggregate dimensions — ten dimensions is a category list, not an abstraction. Exemplars run two to six.
• A retrofitted structure — building the diagram after the findings are written, as decoration. Reviewers can tell — the tiers don't trace.
• No traceable path — every aggregate dimension must walk back down to real quotes. If a tier transition can't be evidenced, the structure is asserting, not showing.

Traceability is the method — so record it as you go

The Gioia methodology's core promise to reviewers is that the abstraction is auditable. That makes it a natural fit with a proper audit trail: each merge of first-order concepts, each theme grouping, each dimension decision — dated, with rationale. If AI tools assist with evidence retrieval anywhere in the process, that log matters even more, and pairs with an AI disclosure statement. For how the method compares with its grounded-theory roots, see the grounded theory guide.

Frequently asked questions