Skip to contents

Package Overview

mars-package
MARS: Meta Analysis and Research Synthesis

Core MARS Models

Fit univariate, multivariate, multilevel, and related MARS models.

mars()
Fit MARS Meta-Analytic Models
summary(<mars>)
Summarize an ordinary MARS fit
mars_repeated()
Repeated-effects multivariate meta-analysis
summary(<mars_repeated>)
Summarize repeated-effects models with optional coefficient inference
mars_dta()
Bivariate meta-analysis of diagnostic test accuracy
mars_dta_paired()
Pool diagnostic accuracy for two paired tests
summary(<mars_dta_paired>) print(<summary.mars_dta_paired>)
Summarize pooled accuracy and paired diagnostic contrasts
mars_dta_thresholds()
Pool diagnostic accuracy at reported thresholds
summary(<mars_dta_thresholds>) predict(<mars_dta_thresholds>) plot(<mars_dta_thresholds>)
Summarize and plot threshold-specific diagnostic accuracy
dta_sroc()
Summary ROC curve for diagnostic accuracy models
predict(<mars_dta>) coef(<mars_dta>) vcov(<mars_dta>) summary(<mars_dta>) print(<mars_dta>) print(<summary.mars_dta>) plot(<mars_dta>)
Summarize, predict, and plot diagnostic accuracy
mars_inference_options
Coefficient inference options in mars
coef(<mars>)
Extract MARS coefficient estimates
vcov(<mars>)
Extract a model-based or cluster-robust covariance matrix
confint(<mars>)
Confidence intervals for MARS coefficients
robust_wald()
Cluster-robust Wald test for MARS coefficients
estimation()
Master Estimation Function
mixed_corr_meta()
Mixed Correlations of Study-Level Effect Sizes and Moderators
corr_to_mars()
Convert a Mixed Correlation Object to a mars-Compatible Object
masmr() summary(<masmr>) print(<summary.masmr>)
Methodology-Adjusted Substantive Meta-Regression (MASMR)
masmr_contrast()
Substantive contrasts under common methodological conditions
masmr_support()
Diagnose support for a MASMR substantive contrast
masmr_departures()
Specify candidate response-surface departures
masmr_bounds()
Record scientifically interpretable sensitivity bounds
masmr_bound_difference()
Calibrate an interaction bound from two methodological shifts
masmr_sensitivity()
Sensitivity of a substantive contrast to bounded methodological departures
model_fit()
Fit Path Models to a Correlation Matrix
path_model()
Path Model Function
path_effects()
Derived Effects from a Synthesized Path Model
print(<mars_path>)
Print a fitted path model
olkin_siotani()
Olkin & Siotani variance-covariance matrix
tau_u_computation()
Tau U effect size
hsmd() print(<hsmd>)
Harmonized Standardized Mean Differences
summary(<hsmd>)
Summarize fitted SMD and HSMD models
hsmd_effect_sizes()
Compute HSMD effect-size columns

Effect-Size Computation and Transformation

es_methods()
Available effect-size computation routes
compute_es() print(<mars_es>) as.data.frame(<mars_es>)
Compute effect sizes from reported study statistics
aggregate_es_subgroups()
Compute SMDs after combining disjoint subgroups within arms
compute_es_covariance()
Construct sampling covariance for dependent effect sizes
dta_paired_effects()
Prepare paired diagnostic effects and sampling covariance
dta_threshold_effects()
Prepare diagnostic effects at multiple thresholds
transform_es() print(<mars_es_transform>) as.data.frame(<mars_es_transform>)
Transform effect estimates and propagate uncertainty

Data Preparation and Example Data

df_to_corr()
Helper function to read in data
bcg
BCG Vaccine Trial Data
becker09
Becker 2009 Correlation Data
hsmd_example_1
HSMD Example 1 Data
hsmd_example_2
HSMD Example 2 Data
math_correlations
Math Correlations Data
school
Modified School Calendar Meta-analysis Data
smd_data
Simulated Multivariate Standardized Mean Difference Data
teacher_expectancy
Teacher Expectancy Data

Single-Case Effect Sizes

Compute effect sizes and sampling variances from single-case phase data.

bc_smd_computation()
Between-case standardized mean difference for replicated AB designs
log_response_ratio_computation()
Log response ratio for single-case phase contrasts
nap_computation()
Nonoverlap of all pairs for single-case phase contrasts
ird_computation()
Improvement rate difference for binary single-case outcomes
level_slope_computation()
Segmented level-change and slope-change effects for single-case data

Within-Study Covariance and Heterogeneity

profile_likelihood()
Profile Likelihood for mars Models
profile_random_effects()
Profile Likelihood for Random-Effects Variance Parameters
plot(<profile_likelihood>)
Plot Profile-Likelihood Curves

Diagnostics and Publication Bias

residuals(<mars>)
Residuals for MARS Models
residual_diagnostics()
Residual Diagnostics for MARS Models
plot_residual_panel()
Plot residual diagnostics panels for MARS models
influence_diagnostics()
Influence Diagnostics for MARS Models
publication_bias()
Publication-Bias Selection Models and Small-Study Diagnostics
egger_test() print(<mars_egger>)
Egger-Type Small-Study-Effects Test
trim_fill() print(<mars_trimfill>)
Trim-and-Fill Sensitivity Analysis
funnel_plot()
Funnel Plot for Publication-Bias Models
selection_plot() plot(<pub_bias>)
Selection-Function Plot for Publication-Bias Models

Alternative Estimation and Resampling

bootstrap_mars()
Bootstrap one independent-unit sample
replicate_bootstrap()
Replicate the bootstrap computations
sample_jackknife()
Jackknife mars wrapper
replicate_jackknife()
Replicate the jackknife computations
permutation_mars()
Permutation mars wrapper
replicate_permutation()
Replicate the permutation computations
mars_alt_estimation()
Primary main alternative estimation function
compute_alt_stats()
Compute summaries from alternative estimation replications
plot(<mars_boot>)
Plot alternative-estimation replications

Network Meta-Analysis

network_meta()
Joint coherent network meta-analysis
network_meta_coherent() predict(<nma_coherent>)
One-stage coherent network meta-analysis
print(<nma_coherent>) summary(<nma_coherent>)
Print and summarize joint network models
print(<nma_output_table>) summary(<nma_output_table>)
Print network diagnostic tables without rounding stored results
coef(<nma_coherent>) vcov(<nma_coherent>)
Coefficients and covariance of coherent network models
confint(<nma_coherent>)
Confidence intervals for coherent network models
network_arm_data()
Construct network contrasts and covariance from independent arms
network_nodesplit()
Independent direct and indirect evidence by node-splitting
network_nodesplit_plot()
Plot direct and indirect node-split effects side by side
network_incoherence() print(<nma_incoherence>) summary(<nma_incoherence>)
Direct versus indirect network incoherence
network_incoherence_plot()
Plot covariance-aware network incoherence assessments
network_transitivity()
Summarize effect modifiers across network comparisons
network_covariance_sensitivity()
Sensitivity to specified sampling covariance assumptions
print(<nma_heterogeneity>) summary(<nma_heterogeneity>) network_heterogeneity()
Within-design network heterogeneity
network_heterogeneity_profile()
Profile likelihood interval for common network heterogeneity
network_influence() print(<nma_influence_summary>) summary(<nma_influence_summary>)
Study or design influence on a joint network fit
network_meta_repeated()
Repeated-effects network meta-analysis from arm-level continuous data
predict(<nma_mars>)
Predict treatment contrasts from a fitted network_meta model
network_pairwise_data()
Convert arm-level study data to contrast-level network meta-analysis data
network_smd_data()
Construct dependent Hedges SMDs from arm summaries
network_league_table()
Create a formatted league table from a fitted network meta-analysis
network_rankings() print(<nma_ranking_table>) summary(<nma_ranking_table>)
Recalculate network rankings with an explicit benefit direction
network_prediction_intervals() print(<nma_prediction_intervals>) summary(<nma_prediction_intervals>)
Prediction intervals for all network treatment contrasts
print(<nma_q_decomposition>) summary(<nma_q_decomposition>) network_q_decomposition()
Decompose network residual Q into within-design and between-design variation
network_design_incoherence() print(<nma_design_incoherence>) summary(<nma_design_incoherence>)
Higgins design-by-treatment incoherence tests
network_design_detachment() print(<nma_design_detachment>) summary(<nma_design_detachment>)
Between-design disagreement after detaching individual designs
network_component_meta()
Additive component network meta-analysis
network_complex_effect()
Estimate arbitrary complex interventions from a component NMA
network_bias_test()
Small-study asymmetry test for network meta-analysis
network_graph_plot()
Base R Network Graph for Network Meta-analysis
network_forest_plot()
Forest Plot for Network Meta-analysis Effects
network_forest_overlay_plot()
Overlay Forest Plot for Direct, Indirect, and Total Network Effects
network_funnel_plot()
Comparison-adjusted funnel plot for network meta-analysis
network_contribution_heatmap()
Heatmap for Network Contribution Matrix
network_rank_plot()
Plot the full distribution of network treatment ranks
network_trajectory_plot()
Plot network treatment effects over numeric follow-up times
network_plot_export()
Export a network plot at explicit publication dimensions

Network Resampling

bootstrap_network_meta()
Bootstrap network meta-analysis
replicate_network_bootstrap()
Replicate network meta-analysis bootstraps
jackknife_network_meta()
Jackknife network meta-analysis
replicate_network_jackknife()
Replicate network meta-analysis jackknife omissions
network_alt_estimation()
Primary network meta-analysis alternative-estimation function
compute_network_alt_stats() summary(<nma_resamples>) summary(<nma_boot>) print(<nma_boot>)
Compute summaries from network meta-analysis resampling replications

Correlation Synthesis Reports and Factor Models

efa_from_synthesis() summary(<efa_mars>) print(<efa_mars>)
Exploratory Factor Analysis After Correlation Synthesis
cfa_from_synthesis() print(<cfa_mars>)
Confirmatory Factor Analysis After Correlation Synthesis
synthesis_factor_report() summary(<synthesis_report>) print(<synthesis_report>) plot(<synthesis_report>)
Standardized Report for Correlation Synthesis and Latent Factors
as_markdown() as_html() as_docx()
Export Synthesis Reports

Penalized Meta-Regression

mars_lasso() mars_lasso_univariate() mars_lasso_multilevel() mars_lasso_multivariate()
Penalized Meta-Regression with Study-Grouped Cross-Validation
coef(<mars_lasso>) predict(<mars_lasso>)
Coefficients and Prediction for Penalized MARS Models
summary(<mars_lasso>) print(<summary.mars_lasso>) print(<mars_lasso>) plot(<mars_lasso>)
Summarize and Plot a Penalized MARS Model

Exploratory Random-Forest Meta-Analysis

mars_rf()
Random Forest Meta-Analysis Models
mars_rf_univariate()
Univariate Random Forest Meta-Analysis
mars_rf_multivariate()
Multivariate Random Forest Meta-Analysis
mars_rf_multilevel()
Multilevel Random Forest Meta-Analysis
rf_importance()
Random Forest Variable Importance
mars_rf_validate()
Grouped Validation for Random Forest Meta-Analysis
predict(<mars_rf>)
Predict from a Random Forest Meta-Analysis

Review Graphics

forest_plot()
Forest Plot for Meta-Analysis Results
gap_map_plot()
Evidence Gap Map Plot
prisma_diagram()
PRISMA Flow Diagram (Base R)
risk_of_bias_plot()
Risk-of-Bias Traffic Light and Aggregate Bar Plot

Deprecated Functions

find_reg_coef()
Find regression coefficients
find_B()
Estimate Regression Coefficients from a Correlation Matrix
Mul_R2()
Mul R2
c_mat_ft()
Build Implied Correlation Matrices for Observed Path Models
conduct_permutation()
Deprecated permutation wrapper