Time-sliced Meta-analysis — Temporal Evidence Synthesis
Time-sliced Meta-analysis · Also known as: temporal meta-analysis, period-stratified meta-analysis, time-segmented meta-analysis, chronological meta-analysis
Time-sliced meta-analysis is a variant of standard meta-analysis in which the primary studies are partitioned into successive time periods (slices) and a separate pooled effect estimate is computed for each period. By comparing pooled effects across periods, researchers can detect whether an intervention's effectiveness, a relationship's magnitude, or a methodological consensus has shifted over time. This temporal lens transforms a static evidence summary into a longitudinal narrative of how scientific knowledge on a topic has evolved.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use time-sliced meta-analysis when you suspect that an effect or phenomenon has changed over time — for example due to treatment refinements, population shifts, methodological improvements, or publication-practice changes — and when the primary literature spans multiple decades with sufficient studies per period. It is especially valuable in fields undergoing rapid methodological evolution or policy change. Do not use it when the available primary studies are too sparse to produce stable period estimates (fewer than three or four studies per slice renders pooled estimates unreliable), when there is no substantive theory predicting temporal change, or as a purely exploratory fishing exercise without pre-specified slices.
Strengths & limitations
- Reveals temporal drift in effect sizes that a single pooled estimate would obscure.
- Helps distinguish genuine changes in a phenomenon from artefactual trends due to evolving study quality or publication bias.
- Integrates naturally with cumulative meta-analysis, making it easy to show when evidence first reached stability.
- Period-level forest plots and trend charts communicate findings accessibly to policymakers and practitioners.
- Compatible with standard meta-analytic software (R metafor, Stata metan, RevMan) requiring only a date-stratification step.
- Sparse slices produce wide confidence intervals, making period comparisons statistically underpowered.
- The choice of slice boundaries can strongly influence apparent trends; arbitrary or post-hoc segmentation inflates Type I error.
- Confounding between period and study characteristics (e.g., later studies use better designs) makes causal attribution of trends difficult.
- Requires a sufficient total number of primary studies; a corpus too small for a standard meta-analysis cannot support additional temporal partitioning.
Frequently asked
How many studies do I need per time slice?
There is no absolute minimum, but fewer than three or four studies in a slice produces a pooled estimate with very wide confidence intervals that is difficult to interpret. If your literature is sparse, consider using fewer, broader slices or shifting to a meta-regression approach with continuous publication year as a moderator rather than discrete windows.
Should I use fixed-effects or random-effects within each slice?
Random-effects models are almost always preferable because they account for heterogeneity across studies within each period. Critically, the same model should be applied to all slices so that period estimates are directly comparable. If one slice is homogeneous and another is not, the difference may reflect a real change in effect variance — document it rather than switching models.
How is time-sliced meta-analysis different from cumulative meta-analysis?
Cumulative meta-analysis adds studies one at a time in chronological order and plots how the pooled estimate stabilizes as evidence accumulates — it asks 'when did the evidence converge?' Time-sliced meta-analysis groups studies into discrete periods and compares period-level pooled effects — it asks 'has the true effect changed across eras?' Both are temporal approaches but they answer distinct questions.
Can I use meta-regression instead of slicing?
Yes, and it is often complementary. Meta-regression with publication year as a continuous moderator tests for a linear trend without imposing arbitrary boundaries. Time slicing is preferred when the researcher has a theoretical reason to expect distinct phases (e.g., before and after a guideline change). In practice, reporting both analyses strengthens transparency.
Do PRISMA guidelines apply to time-sliced meta-analysis?
Yes. A time-sliced meta-analysis is a meta-analysis with additional temporal stratification, so PRISMA 2020 reporting standards apply fully. The pre-specified slicing strategy should be documented in a prospectively registered protocol (e.g., PROSPERO) to guard against post-hoc boundary selection.
Sources
- Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Introduction to Meta-Analysis. Wiley. ISBN: 978-0470057247
- Lau, J., Antman, E. M., Jimenez-Silva, J., Kupelnick, B., Mosteller, F., & Chalmers, T. C. (1992). Cumulative meta-analysis of therapeutic trials for myocardial infarction. New England Journal of Medicine, 327(4), 248–254. DOI: 10.1056/nejm199207233270406 ↗
How to cite this page
ScholarGate. (2026, June 3). Time-sliced Meta-analysis. ScholarGate. https://scholargate.app/en/scientometrics/time-sliced-meta-analysis
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Bibliometric AnalysisScientometrics↔ compare
- meta-regression-based meta-analysisScientometrics↔ compare
- Scoping ReviewScientometrics↔ compare
- Systematic Literature ReviewScientometrics↔ compare
- Thematic Evolution AnalysisScientometrics↔ compare