Aggregates individual-level data from simulate_meta_did() to the four-cell summary
(pre/post × control/treatment) required by meta_did(). The pre-post
correlation \(\rho\) is estimated empirically from matched pairs within
each arm and then averaged.
Arguments
- sim
A data frame produced by
simulate_meta_did().
Value
A data frame with one row per study in the format required by
validate_summary_data() for design "did".
Examples
sim <- simulate_meta_did(n_studies = 3, seed = 1)
as_summary_did(sim)
#> # A tibble: 3 × 13
#> study_id design n_control n_treatment mean_pre_control mean_post_control
#> <chr> <chr> <int> <int> <dbl> <dbl>
#> 1 study_1 did 100 100 0.463 0.438
#> 2 study_2 did 100 100 0.441 0.424
#> 3 study_3 did 100 100 0.452 0.434
#> # ℹ 7 more variables: sd_pre_control <dbl>, sd_post_control <dbl>,
#> # mean_pre_treatment <dbl>, mean_post_treatment <dbl>,
#> # sd_pre_treatment <dbl>, sd_post_treatment <dbl>, rho <dbl>