Why the Same Report Returns Three Different Numbers — and How to Reconcile Them

How consistent definitions and structured research data can make institutional reporting more reproducible, reconcilable, and defensible.

A familiar research administration problem: two people run what appears to be the same report and get different numbers.

One report shows 147 active researchers. Another shows 153. A third, generated a few months later, shows 150.

The difference may not be an error. It may result from different reporting windows, definitions, filters, dates, or changes to the underlying data.

When those rules are unclear or live in spreadsheets and email threads, teams can spend significant time comparing records just to understand why the totals do not match.

Key takeaway: A report is only as consistent as the data and reporting rules behind it. Reproducible reporting requires both.

The real issue isn't the number, it's the definition

Even a straightforward reporting question can hide several decisions.

Consider:

How many active researchers were affiliated with a program during the reporting period?

Before producing a defensible number, you need to determine:

• What qualifies as active?
• Which start and end dates determine inclusion?
• If someone leaves during the reporting period, do they still count?
• How are gaps in membership handled?
• Are secondary affiliations included?
• What happens when dates are corrected after the report is generated?

The same issue appears across funding, publications, memberships, trainees, collaborations, and other research data that changes over time.

Best practice: define the population, reporting period, and inclusion rules before generating the report, not after conflicting numbers appear.

Why the same report can return different totals

Differences typically come from one or more areas:

Sometimes, neither number is technically wrong. The reports may simply be answering slightly different questions.

Reconcile the records, not just the totals

When two reports disagree, comparing 147 vs. 153 does not explain the difference.

The more useful question is:

Which six records are different, and why?

The difference often reveals the source of the mismatch:

• Missing or incorrect dates
• Membership status changes
• Duplicate records
• Different program affiliations
• Different inclusion criteria
• Records added or updated since the earlier report
• Different source data

A clearer reconciliation path looks like this:

Reported total → included records → reporting criteria → underlying data → validated total

This is particularly valuable for recurring institutional and CCSG-related reporting, where figures may need to be reproduced or explained after the original report was generated.

Keep reporting logic from living in spreadsheets

Many research organizations have reporting rules that are understood internally but not consistently documented.

"Remove these records before submitting."

"For this report, use membership dates, not current status."

"Manually add these awards after the export."

These decisions may be appropriate, but when reporting logic lives in spreadsheets, email threads, or individual staff knowledge, reproducibility depends on institutional memory.

A more sustainable reporting path looks like this:

Structured research data → defined reporting rules → consistent filtering → reconciliation → defensible reporting

Research Logix in practice

Research Logix helps centralize and structure the research data behind institutional reporting, creating a clearer path from individual records to reported totals.

• Make reporting scope clearer — use record status and structured criteria to define what belongs in a report.
• Preserve dates and history — maintain dated information so reporting can reflect the appropriate period rather than only current status.
• Support record-level reconciliation — trace reported totals back to the records and details contributing to them.
• Reduce duplicate and rollup confusion — use structured identifiers and relationships to maintain consistent record identity.
• Reuse structured reporting data — apply consistent filters instead of rebuilding reporting populations in spreadsheets.
• Support recurring reporting — maintain a consistent data foundation for institutional, leadership, CCSG-related, and other reporting needs.

When two numbers disagree

Before rebuilding the report, check five things:

1. Confirm the reporting window and date field driving inclusion.
2. Confirm the population and status types included.
3. Confirm the record state, such as reviewed versus pending.
4. Confirm the counting unit, such as project vs. budget period, or publication vs. publication-author relationship.
5. Compare the included record lists, not just the totals.

The bottom line

"Three different numbers" is often a symptom of undefined scope, inconsistent date logic, or exceptions that are difficult to trace — not necessarily a broken report.

When reporting rules are defined up front and the underlying data is structured and traceable, discrepancies become easier to identify, reconcile, and explain.

Research Logix connects reported numbers to the records behind them, supporting reporting that is more consistent, reproducible, and defensible.

Schedule a demo · Contact our team