% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/bayesfactor_parameters.R
\name{bayesfactor_parameters}
\alias{bayesfactor_parameters}
\alias{bayesfactor_pointull}
\alias{bayesfactor_rope}
\alias{bf_parameters}
\alias{bf_pointull}
\alias{bf_rope}
\alias{bayesfactor_parameters.numeric}
\alias{bayesfactor_parameters.stanreg}
\alias{bayesfactor_parameters.brmsfit}
\alias{bayesfactor_parameters.emmGrid}
\alias{bayesfactor_parameters.data.frame}
\title{Bayes Factors (BF) for a Single Parameter}
\usage{
bayesfactor_parameters(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
...
)
bayesfactor_pointull(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
...
)
bayesfactor_rope(
posterior,
prior = NULL,
direction = "two-sided",
null = rope_range(posterior),
verbose = TRUE,
...
)
bf_parameters(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
...
)
bf_pointull(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
...
)
bf_rope(
posterior,
prior = NULL,
direction = "two-sided",
null = rope_range(posterior),
verbose = TRUE,
...
)
\method{bayesfactor_parameters}{numeric}(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
...
)
\method{bayesfactor_parameters}{stanreg}(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
effects = c("fixed", "random", "all"),
component = c("conditional", "zi", "zero_inflated", "all"),
parameters = NULL,
...
)
\method{bayesfactor_parameters}{brmsfit}(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
effects = c("fixed", "random", "all"),
component = c("conditional", "zi", "zero_inflated", "all"),
parameters = NULL,
...
)
\method{bayesfactor_parameters}{emmGrid}(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
...
)
\method{bayesfactor_parameters}{data.frame}(
posterior,
prior = NULL,
direction = "two-sided",
null = 0,
verbose = TRUE,
...
)
}
\arguments{
\item{posterior}{A numerical vector, \code{stanreg} / \code{brmsfit} object, \code{emmGrid}
or a data frame - representing a posterior distribution(s) from (see 'Details').}
\item{prior}{An object representing a prior distribution (see 'Details').}
\item{direction}{Test type (see 'Details'). One of \code{0}, \code{"two-sided"} (default, two tailed),
\code{-1}, \code{"left"} (left tailed) or \code{1}, \code{"right"} (right tailed).}
\item{null}{Value of the null, either a scaler (for point-null) or a a range
(for a interval-null).}
\item{verbose}{Toggle off warnings.}
\item{...}{Currently not used.}
\item{effects}{Should results for fixed effects, random effects or both be returned?
Only applies to mixed models. May be abbreviated.}
\item{component}{Should results for all parameters, parameters for the conditional model
or the zero-inflated part of the model be returned? May be abbreviated. Only
applies to \pkg{brms}-models.}
\item{parameters}{Regular expression pattern that describes the parameters that
should be returned. Meta-parameters (like \code{lp__} or \code{prior_}) are
filtered by default, so only parameters that typically appear in the
\code{summary()} are returned. Use \code{parameters} to select specific parameters
for the output.}
}
\value{
A data frame containing the Bayes factor representing evidence \emph{against} the null.
}
\description{
This method computes Bayes factors against the null (either a point or an interval),
based on prior and posterior samples of a single parameter. This Bayes factor indicates
the degree by which the mass of the posterior distribution has shifted further away
from or closer to the null value(s) (relative to the prior distribution), thus indicating
if the null value has become less or more likely given the observed data.
\cr \cr
When the null is an interval, the Bayes factor is computed by comparing the prior
and posterior odds of the parameter falling within or outside the null interval
(Morey & Rouder, 2011; Liao et al., 2020); When the null is a point, a Savage-Dickey
density ratio is computed, which is also an approximation of a Bayes factor comparing
the marginal likelihoods of the model against a model in which the tested parameter
has been restricted to the point null (Wagenmakers et al., 2010; Heck, 2019).
\cr \cr
Note that the \code{logspline} package is used for estimating densities and probabilities,
and must be installed for the function to work.
\cr \cr
\code{bayesfactor_pointnull()} and \code{bayesfactor_rope()} are wrappers around
\code{bayesfactor_parameters} with different defaults for the null to be tested against
(a point and a range, respectively). Aliases of the main functions are prefixed
with \code{bf_*}, like \code{bf_parameters()} or \code{bf_pointnull()}
\cr \cr
\strong{For more info, in particular on specifying correct priors for factors with more than 2 levels, see \href{https://easystats.github.io/bayestestR/articles/bayes_factors.html}{the Bayes factors vignette}.}
}
\details{
This method is used to compute Bayes factors based on prior and posterior distributions.
\cr\cr
For the computation of Bayes factors, the model priors must be proper priors (at the very least
they should be \emph{not flat}, and it is preferable that they be \emph{informative}); As the priors for
the alternative get wider, the likelihood of the null value(s) increases, to the extreme that for completely
flat priors the null is infinitely more favorable than the alternative (this is called \emph{the Jeffreys-Lindley-Bartlett
paradox}). Thus, you should only ever try (or want) to compute a Bayes factor when you have an informed prior.
\cr\cr
(Note that by default, \code{brms::brm()} uses flat priors for fixed-effects; See example below.)
\subsection{Setting the correct \code{prior}}{
It is important to provide the correct \code{prior} for meaningful results.
\itemize{
\item When \code{posterior} is a numerical vector, \code{prior} should also be a numerical vector.
\item When \code{posterior} is a \code{data.frame}, \code{prior} should also be a \code{data.frame}, with matching column order.
\item When \code{posterior} is a \code{stanreg} or \code{brmsfit} model: \itemize{
\item \code{prior} can be set to \code{NULL}, in which case prior samples are drawn internally.
\item \code{prior} can also be a model equvilant to \code{posterior} but with samples from the priors \emph{only}.
}
\item When \code{posterior} is an \code{emmGrid} object: \itemize{
\item \code{prior} should be the \code{stanreg} or \code{brmsfit} model used to create the \code{emmGrid} objects.
\item \code{prior} can also be an \code{emmGrid} object equvilant to \code{posterior} but created with a model of priors samples \emph{only}.
}
}}
\subsection{One-sided Tests (setting an order restriction)}{
One sided tests (controlled by \code{direction}) are conducted by restricting the prior and
posterior of the non-null values (the "alternative") to one side of the null only
(\cite{Morey & Wagenmakers, 2014}). For example, if we have a prior hypothesis that the
parameter should be positive, the alternative will be restricted to the region to the right
of the null (point or interval).
}
\subsection{Interpreting Bayes Factors}{
A Bayes factor greater than 1 can be interpereted as evidence against the null,
at which one convention is that a Bayes factor greater than 3 can be considered
as "substantial" evidence against the null (and vice versa, a Bayes factor
smaller than 1/3 indicates substantial evidence in favor of the null-model)
(\cite{Wetzels et al. 2011}).
}
}
\examples{
library(bayestestR)
prior <- distribution_normal(1000, mean = 0, sd = 1)
posterior <- distribution_normal(1000, mean = .5, sd = .3)
bayesfactor_parameters(posterior, prior)
\dontrun{
# rstanarm models
# ---------------
if (require("rstanarm") && require("emmeans")) {
contrasts(sleep$group) <- contr.bayes # see vingette
stan_model <- stan_lmer(extra ~ group + (1 | ID), data = sleep)
bayesfactor_parameters(stan_model)
bayesfactor_parameters(stan_model, null = rope_range(stan_model))
# emmGrid objects
# ---------------
group_diff <- pairs(emmeans(stan_model, ~group))
bayesfactor_parameters(group_diff, prior = stan_model)
}
# brms models
# -----------
if (require("brms")) {
contrasts(sleep$group) <- contr.bayes # see vingette
my_custom_priors <-
set_prior("student_t(3, 0, 1)", class = "b") +
set_prior("student_t(3, 0, 1)", class = "sd", group = "ID")
brms_model <- brm(extra ~ group + (1 | ID),
data = sleep,
prior = my_custom_priors
)
bayesfactor_parameters(brms_model)
}
}
}
\references{
\itemize{
\item Wagenmakers, E. J., Lodewyckx, T., Kuriyal, H., and Grasman, R. (2010). Bayesian hypothesis testing for psychologists: A tutorial on the Savage-Dickey method. Cognitive psychology, 60(3), 158-189.
\item Heck, D. W. (2019). A caveat on the Savage–Dickey density ratio: The case of computing Bayes factors for regression parameters. British Journal of Mathematical and Statistical Psychology, 72(2), 316-333.
\item Morey, R. D., & Wagenmakers, E. J. (2014). Simple relation between Bayesian order-restricted and point-null hypothesis tests. Statistics & Probability Letters, 92, 121-124.
\item Morey, R. D., & Rouder, J. N. (2011). Bayes factor approaches for testing interval null hypotheses. Psychological methods, 16(4), 406.
\item Liao, J. G., Midya, V., & Berg, A. (2020). Connecting and contrasting the Bayes factor and a modified ROPE procedure for testing interval null hypotheses. The American Statistician, 1-19.
\item Wetzels, R., Matzke, D., Lee, M. D., Rouder, J. N., Iverson, G. J., and Wagenmakers, E.-J. (2011). Statistical Evidence in Experimental Psychology: An Empirical Comparison Using 855 t Tests. Perspectives on Psychological Science, 6(3), 291–298. \doi{10.1177/1745691611406923}
}
}
\author{
Mattan S. Ben-Shachar
}