Technical Appendix: Workflow of `do_boot()`

In manymome: Mediation, Moderation and Moderated-Mediation After Model Fitting

library(DiagrammeR)

Goal

This technical appendix describes how do_boot() from the package manymome (Cheung & Cheung, 2023) works to generate bootstrap estimates to be used by other functions to form confidence intervals.

do_boot()

mermaid("
flowchart TD
  classDef default fill:#EEEEFF;
  classDef errornode fill:#FFDDDD;
  classDef startend fill:#DDFFDD;
  classDef mcnode fill:#FFFFDD;
  classDef bootnode fill:#DDFFFF;
  classDef lavnode fill:#FFDDFF;
  classDef lmnode fill:#FFDDDD;
  classDef subnode fill:#FFFFDD;

  ZZ([\"Start\"])
  Z{{\"How was the model fitted?\"}}
  ZZ:::startend --> Z
%%  Zla[\"Fitted by lavaan()\"]
%%  Zlm[\"Fitted by lm()\"]
%%  Z -- lm --> Zlm
%%  Z -- lavaan --> Zla
  Z -- lavaan --> A
  Z -- lm --> B

  subgraph lavaan [ ]
  A{{\"Are bootstrap estimates stored?\"}}
  AYes[[\"Call fit2boot_out()\"]]
  ANo[[\"Call fit2boot_out_do_boot()\"]]
  end

%%  Zla --> A
  A -- Yes --> AYes:::subnode
  A -- No --> ANo:::subnode

  subgraph lm [ ]
  B{{\"Parallel processing?\"}}
  BYes[[\"Call lm2boot_out_parallel()\"]]
  BNo[[\"Call lm2boot_out()\"]]
  end

%%  Zlm --> B
  B -- Yes --> BYes:::subnode
  B -- No --> BNo:::subnode

  C([\"Return a boot_out-class object\"])
  AYes --> C:::startend
  ANo --> C
  BYes --> C
  BNo --> C
", height = 400, width = 800)

lavaan

fit2boot_out()

It retrieves stored bootstrap estimates.

mermaid("
flowchart TD
  classDef default fill:#EEEEFF;
  classDef errornode fill:#FFDDDD;
  classDef startend fill:#DDFFDD;
  classDef mcnode fill:#FFFFDD;
  classDef bootnode fill:#DDFFFF;
  classDef lavnode fill:#FFDDFF;
  classDef lmnode fill:#FFDDDD;
  classDef subnode fill:#FFFFDD;

  ZZ([\"Start\"])
  ZZ:::startend --> A

  A[\"boot_est <- boot2est()\"]
  B[\"boot_implied <- boot2implied()\"]
  B2[\"Combine boot_est and boot_implied\"]
  C([\"Return a boot_out-class object\"])
  A --> B
  B --> B2
  B2 --> C:::startend
", height = 400)

fit2boot_out_do_boot()

mermaid("
flowchart TD
  classDef default fill:#EEEEFF;
  classDef errornode fill:#FFDDDD;
  classDef startend fill:#DDFFDD;
  classDef mcnode fill:#FFFFDD;
  classDef bootnode fill:#DDFFFF;
  classDef lavnode fill:#FFDDFF;
  classDef lmnode fill:#FFDDDD;
  classDef subnode fill:#FFFFDD;

  ZZ([\"Start\"])
  ZZ:::startend --> A

  A[\"Call gen_boot_i() to create<br>a function, boot_i(), for doing one bootstrap replication\"]
  B[\"Generate indices (ids) locally\"]
  A --> B
  C[\"Call boot_i() R times\"]
  D[\"Keep only replications with valid values\"]
  E([\"Return a boot_out-class object\"])
  B --> C
  C --> D
  D --> E:::startend
", height = 500)

lm

lm2boot_out_parallel()

mermaid("
flowchart TD
  classDef default fill:#EEEEFF;
  classDef errornode fill:#FFDDDD;
  classDef startend fill:#DDFFDD;
  classDef mcnode fill:#FFFFDD;
  classDef bootnode fill:#DDFFFF;
  classDef lavnode fill:#FFDDFF;
  classDef lmnode fill:#FFDDDD;
  classDef subnode fill:#FFFFDD;

  ZZ([\"Start\"])
  ZZ:::startend --> A

  A[\"Call merge_model_frame() to reconstruct the data\"]
  B[\"Generate indices (ids) locally\"]
  A --> B
  C[\"Call lm_boot2est_i() R times\"]
  E([\"Return a boot_out-class object\"])
  B --> C
  C --> E:::startend
", height = 450)

lm2boot_out()

mermaid("
flowchart TD
  classDef default fill:#EEEEFF;
  classDef errornode fill:#FFDDDD;
  classDef startend fill:#DDFFDD;
  classDef mcnode fill:#FFFFDD;
  classDef bootnode fill:#DDFFFF;
  classDef lavnode fill:#FFDDFF;
  classDef lmnode fill:#FFDDDD;
  classDef subnode fill:#FFFFDD;

  ZZ([\"Start\"])
  ZZ:::startend --> A

  A[\"Call merge_model_frame() to reconstruct the data\"]
  C[\"Call lm_boot2est_i() R times\"]
  E([\"Return a boot_out-class object\"])
  A --> C
  C --> E:::startend
", height = 380)

Notes

Extracting Point Estimates

When the point estimates are needed, they will be extracted internally using functions developed for the fit object, which can be a lavaan-class object, a list of the outputs from stats::lm(), or a lavaan.mi-class object generated by fitting a model to several datasets using multiple imputation.

Reference

Cheung, S. F., & Cheung, S.-H. (2023). manymome: An R package for computing the indirect effects, conditional effects, and conditional indirect effects, standardized or unstandardized, and their bootstrap confidence intervals, in many (though not all) models. Behavior Research Methods. https://doi.org/10.3758/s13428-023-02224-z

manymome documentation built on Oct. 4, 2024, 5:10 p.m.