When Does a Statistical Time Series Stop Being the Same Series?
A source-linked pilot audit of methodological discontinuities in MENA official statistics.
Research question
When an official statistical series changes concept, base, weights, classification, survey design, sampling frame, compilation method or historical treatment, what determines whether observations on either side of that change can still be compared directly?
Contribution
The paper treats method version as part of the statistical observation. Each documented discontinuity is classified by what changed, whether the producer supplies a usable bridge or backcast, and the action a researcher should take before joining the series across regimes.
Pilot result
Producer-supplied reconstructed, revised or overlapping history.
Some backcast or overlap exists, but does not fully settle the transition.
The researcher must preserve or explicitly resolve the break.
Interpretation: these are counts within a purposively selected 15-case pilot. They are not estimates of how common bridge provision or method breaks are across MENA statistical systems.
Main methodological finding
A statistical discontinuity does not imply one universal response. Some transitions should be handled with an official reconstructed history. Others require a provisional backcast, base alignment, a classification concordance, or explicit preservation of the break because no official bridge has been verified.
Claim boundary
The paper does not rank statistical agencies and does not treat revisions, rebasing or survey redesign as evidence of poor quality. Method changes can improve measurement. The research object is what those changes imply for downstream comparability and reproducibility.
Review state
All current records are source-verified against producer material. Independent non-author reproduction is pending. The release will remain a candidate until a non-author reviewer reconstructs at least five records from the published first-party sources without asking the author for the intended answer.
Suggested citation
Aloreidi, L. (2026). When Does a Statistical Time Series Stop Being the Same Series? A source-linked pilot audit of methodological discontinuities in MENA official statistics. MENA Open Data & Evidence Lab Working Paper 001, v0.2.0-rc1.