Skip to contents

Create Model Choice Simulation Design

Source: R/modelchoice_design.R
create_modelchoice_design.Rd

Build a design grid for model choice analysis: multiple DGPs (pop_model_x) with per-DGP population parameters, and a fixed set of candidate analysis models fit to each generated dataset.

Usage

create_modelchoice_design(
  analysis_models,
  pop_values_x,
  pop_model_x,
  pop_generator = NULL,
  ...,
  .labels = NULL,
  .seed = NULL
)

Arguments

analysis_models

Named list of lavaan scripts (length >= 2). Names become analysis_model_labels. Order: most constrained (simplest) first.

pop_values_x

Named list keyed by pop_model_x. Each element is expand_pop_values_x-compatible (fixed, conditions, cross factors). Uni structure: loadings, residuals, coresids. Cor structure: loadings_p1, loadings_p2, residuals_p1, residuals_p2, coresids, lv_cov.

pop_model_x

Character vector of DGP levels (e.g. c("Uni", "Cor")). Names of pop_values_x must contain all of these.

pop_generator

Optional named vector mapping pop_model_x to "Uni" or "Cor" for data generation. If NULL, inferred from parameter names (loadings -> Uni, loadings_p1 -> Cor).

...

Other factors (e.g. n_dyads, n_items_x) to expand.

.labels

Optional named list for expand_pop_values_x labels (e.g. list(conditions = c("weak","strong"))).

.seed

Base seed; seed = .seed + row_number() per scenario.

Value

Tibble with one row per scenario: analysis_models list-column, analysis_model_labels, pop_model_x, pop_values_x, pop_generator, and other design columns.

Details

Model order: analysis_models must be ordered simplest (most constrained) first. For chi2_diff_startsimple and chi2_diff_startcomplex, the LRT compares nested models; if the complex model is listed first, the test logic is wrong. Example for loading invariance: list(loading_inv = script_load, configural = script_none).

See also

Other design: create_design(), expand_pop_values_x()

Examples

dvn <- dySEM::build_dvn(n_items_x = 3, x_stem = "x", x_delim2 = "_",
  distinguish_1 = "A", distinguish_2 = "B")
#> Error: 'build_dvn' is not an exported object from 'namespace:dySEM'
script_uni <- dySEM::scriptUni(dvn, scaleset = "FF", lvname = "X",
  constr_dy_meas = c("loadings", "intercepts", "residuals"),
  constr_dy_struct = "none")
#> Error: object 'dvn' not found
script_cor <- dySEM::scriptCor(dvn, scaleset = "FF", lvname = "X",
  constr_dy_meas = c("loadings", "intercepts", "residuals"),
  constr_dy_struct = c("variances", "means"))
#> Error: object 'dvn' not found
design <- create_modelchoice_design(
  pop_model_x = c("Uni", "Cor"),
  analysis_models = list(uni = script_uni, cor = script_cor),
  pop_values_x = list(
    Uni = list(loadings = c(.75,.8,.55), residuals = c(.44,.36,.7),
               coresids = c(.1,.12,.08)),
    Cor = list(loadings_p1 = c(.75,.8,.55), loadings_p2 = c(.75,.8,.55),
               residuals_p1 = c(.44,.36,.7), residuals_p2 = c(.44,.36,.7),
               coresids = c(.1,.12,.08), lv_cov = 0.3)
  ),
  n_dyads = c(50, 100),
  n_items_x = 3,
  .seed = 12345
)
#> Error: object 'script_uni' not found