Apply the Same Weighting Across Different Analyses principle

SWADA (Same Weighting Across Different Analyses) is a consistency principle, not a statistical model parallel to da(), ad(), or cams(). da() (Difference of Averages) and ad() (Average Difference) are different constructions (Panaro, Röver & Friede 2026) and can assign studies different effective contributions. SWADA requires the same study weights for the subgroup-mean analyses and for the within-trial interaction, so that \(\mu_B - \mu_A = \gamma\).

Usage

swada(
  data,
  reference,
  modifier = NULL,
  trial = "study",
  subgroup = "level",
  yi = "yi",
  vi = "vi",
  weights = c("interaction"),
  method = "REML",
  ...
)

Arguments

data: A data frame with trial, subgroup, yi, and vi.
reference: Reference subgroup label.
modifier: Column of the two levels. Overrides subgroup when set. The default column remains "subgroup".
trial, subgroup, yi, vi: Column names. subgroup remains accepted.
weights: Weighting scheme. Currently "interaction" (recommended).
method: Estimator for the interaction RE-MA ("REML" default, or "ML").
...: Passed to ad() / metafor::rma() for the interaction fit.
engine: Computational backend. Currently "metafor" only. Bayesian SWADA is not implemented; Bayesian inference does not add CAMS IF functionality.
approach: Inferential approach. SWADA is frequentist only; "bayes" is an error.

Value

An object of class "interactmeta_swada".

Details

The default (weights = "interaction") is Interaction RE-weights SWADA (Panaro, Röver & Friede 2026): take the ad() weights \(w_j \propto 1/(v_{g j}+\hat\tau_g^2)\) and apply them to both subgroup series. Under that common vector the DA interaction is algebraically identical to AD: \(\sum_j w_j(y_{Bj}-y_{Aj})=\sum_j w_j g_j\). The interaction estimate itself is the AD/rma of \(g_j\); SWADA does not introduce a third interaction model.

References

Panaro RV, Röver C, Friede T (2026). Subgroup comparisons within and across studies in meta-analysis. Research Synthesis Methods (in press). arXiv:2508.15531. https://doi.org/10.48550/arXiv.2508.15531