1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
|
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/correlationBF.R
\name{correlationBF}
\alias{correlationBF}
\title{Function for Bayesian analysis of correlations}
\usage{
correlationBF(
y,
x,
rscale = "medium",
nullInterval = NULL,
posterior = FALSE,
callback = function(...) as.integer(0),
...
)
}
\arguments{
\item{y}{first continuous variable}
\item{x}{second continuous variable}
\item{rscale}{prior scale. A number of preset values can be given as
strings; see Details.}
\item{nullInterval}{optional vector of length 2 containing
lower and upper bounds of an interval hypothesis to test, in correlation units}
\item{posterior}{if \code{TRUE}, return samples from the posterior instead
of Bayes factor}
\item{callback}{callback function for third-party interfaces}
\item{...}{further arguments to be passed to or from methods.}
}
\value{
If \code{posterior} is \code{FALSE}, an object of class
\code{BFBayesFactor} containing the computed model comparisons is
returned. If \code{nullInterval} is defined, then two Bayes factors will
be computed: The Bayes factor for the interval against the null hypothesis
that the probability is 0, and the corresponding Bayes factor for
the complement of the interval.
If \code{posterior} is \code{TRUE}, an object of class \code{BFmcmc},
containing MCMC samples from the posterior is returned.
}
\description{
Bayes factors or posterior samples for correlations.
}
\details{
The Bayes factor provided by \code{ttestBF} tests the null hypothesis that
the true linear correlation \eqn{\rho}{rho} between two samples (\eqn{y}{y} and \eqn{x}{x})
of size \eqn{n}{n} from normal populations is equal to 0. The Bayes factor is based on Jeffreys (1961)
test for linear correlation. Noninformative priors are assumed for the population means and
variances of the two population; a shifted, scaled beta(1/rscale,1/rscale) prior distribution
is assumed for \eqn{\rho}{rho} (note that \code{rscale} is called \eqn{\kappa}{kappa} by
Ly et al. 2015; we call it \code{rscale} for consistency with other BayesFactor functions).
For the \code{rscale} argument, several named values are recognized:
"medium.narrow", "medium", "wide", and "ultrawide". These correspond
to \eqn{r} scale values of \eqn{1/\sqrt(27)}{1/sqrt(27)}, \eqn{1/3}{1/3},
\eqn{1/\sqrt(3)}{1/sqrt(3)} and 1, respectively.
The Bayes factor is computed via several different methods.
}
\examples{
bf = correlationBF(y = iris$Sepal.Length, x = iris$Sepal.Width)
bf
## Sample from the corresponding posterior distribution
samples = correlationBF(y = iris$Sepal.Length, x = iris$Sepal.Width,
posterior = TRUE, iterations = 10000)
plot(samples[,"rho"])
}
\references{
Ly, A., Verhagen, A. J. & Wagenmakers, E.-J. (2015).
Harold Jeffreys's Default Bayes Factor Hypothesis Tests: Explanation, Extension, and Application in Psychology.
Journal of Mathematical Psychology, Available online 28 August 2015, https://dx.doi.org/10.1016/j.jmp.2015.06.004.
Jeffreys, H. (1961). Theory of probability, 3rd edn. Oxford, UK: Oxford University Press.
}
\seealso{
\code{\link{cor.test}}
}
\author{
Richard D. Morey (\email{richarddmorey@gmail.com})
}
\keyword{htest}
|
