Meta-analysis of Lagged Regression Models: A Continuous-time Approach

Abstract

In science, the gold standard for evidence is an empirical result which is consistent across multiple studies. Meta-analysis techniques allow researchers to combine the results of different studies. Lagged effects models based on longitudinal data are increasingly the target for meta-analysis: However, in current practice, little attention is paid to the unique challenges of meta-analyzing these lagged effects. Namely, it is well-known that lagged effects estimates change depending on the time that elapses between measurement waves. This means that studies that use different uniform time intervals between observations (e.g., 1 hour vs 3 hours or 1 month vs 2 months) can come to very different parameter estimates, and seemingly contradictory conclusions, about the same underlying process. In this article, we introduce, describe, and illustrate a new meta-analysis method (CTmeta) which assumes an underlying continuous-time process, thereby offering a potential solution to the time-interval problem. We illustrate this method and compare it with the current best-practice in dealing with time-interval dependency in the meta-analysis literature.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 21 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.