aictab {RSA} | R Documentation |
Show a table of AIC model comparisons
Description
Show a table of AIC model comparisons
Usage
aictab(
x,
plot = FALSE,
bw = FALSE,
models = names(x$models)[!names(x$models) %in% c("absdiff", "absunc")],
digits = NA
)
Arguments
x |
An RSA object |
plot |
Should a plot of the AICc table be plotted? |
bw |
Should the plot be black & white? |
models |
A vector with all model names of the candidate set. Defaults to all polynomial models in the RSA object. |
digits |
The output is rounded to this number of digits. No rounding if NA (default). |
Value
- Modnames
Model names.
- K
Number of estimated parameters (including the intercept, residual variance, and, if present in the model, control variables).
- LL
Model log-likelihood.
- AICc
Akaike Information Criterion (corrected).
- Delta_AICc
Difference in AICc between this model and the best model.
- AICcWt
The Akaike weights, also termed "model probabilities" by Burnham and Anderson (2002). Indicates the level of support (i.e., weight of evidence) of a model being the most parsimonious among the candidate model set.
- Cum.Wt
Cumulative Akaike weight. One possible strategy is to restrict interpretation to the "confidence set" of models, that is, discard models with a Cum.Wt > .95 (see Burnham & Anderson, 2002, for details and alternatives).
- evidence.ratio
Likelihood ratio of this model vs. the best model.
- cfi
Comparative Fit Index (CFI).
- R2
Coefficient of determination (R-squared).
- R2.adj
Adjusted R-squared.
- R2.baseline
Only provided if the model contains control variables. Difference in R-squared as compared to the baseline model with intercept and control variables (= the model "null"). This R^2 increment will typically be of interest because it refers to the amount of variance explained by the two predictors X and Y (plus their squared and interaction terms) in the RSA model.
- R2.baseline.p
Only provided if the model contains control variables. p-value for the F-test of the model against the baseline model.
Note
This function is similar to the function aictab
in the AICcmodavg
package.
References
Burnham, K. P., & Anderson, D. R. (2002). Model selection and multimodel inference: A practical information-theoretic approach. Springer Science & Business Media.
Examples
## Not run:
data(motcon)
r.m <- RSA(postVA~ePow*iPow, motcon, verbose=FALSE)
aictab(r.m, plot=TRUE)
## End(Not run)