Skip to content

Commit

Permalink
improve pd docs
Browse files Browse the repository at this point in the history
  • Loading branch information
mattansb committed Jul 30, 2023
1 parent 0d13d0c commit fecdef4
Show file tree
Hide file tree
Showing 4 changed files with 151 additions and 115 deletions.
134 changes: 72 additions & 62 deletions R/p_direction.R
Original file line number Diff line number Diff line change
@@ -1,81 +1,91 @@
#' Probability of Direction (pd)
#'
#' Compute the **Probability of Direction** (***pd***, also known
#' as the Maximum Probability of Effect - *MPE*). It varies between `50%`
#' and `100%` (*i.e.*, `0.5` and `1`) and can be interpreted as
#' the probability (expressed in percentage) that a parameter (described by its
#' posterior distribution) is strictly positive or negative (whichever is the
#' most probable). It is mathematically defined as the proportion of the
#' posterior distribution that is of the median's sign. Although differently
#' expressed, this index is fairly similar (*i.e.*, is strongly correlated)
#' to the frequentist **p-value**.
#' \cr\cr
#' Note that in some (rare) cases, especially when used with model averaged
#' posteriors (see [weighted_posteriors()] or
#' `brms::posterior_average`), `pd` can be smaller than `0.5`,
#' reflecting high credibility of `0`.
#' Compute the **Probability of Direction** (***pd***, also known as the Maximum
#' Probability of Effect - *MPE*). This can be interpreted as the probability
#' that a parameter (described by its posterior distribution) is strictly
#' positive or negative (whichever is the most probable). Although differently
#' expressed, this index is fairly similar (*i.e.*, is strongly correlated) to
#' the frequentist **p-value** (see details).
#'
#' @param x Vector representing a posterior distribution. Can also be a Bayesian model (`stanreg`, `brmsfit` or `BayesFactor`).
#' @param method Can be `"direct"` or one of methods of [density estimation][estimate_density], such as `"kernel"`, `"logspline"` or `"KernSmooth"`. If `"direct"` (default), the computation is based on the raw ratio of samples superior and inferior to 0. Else, the result is based on the [Area under the Curve (AUC)][auc] of the estimated [density][estimate_density] function.
#' @param null The value considered as a "null" effect. Traditionally 0, but could also be 1 in the case of ratios.
#' @param x A vector representing a posterior distribution, a data frame of
#' posterior draws (samples be parameter). Can also be a Bayesian model.
#' @param method Can be `"direct"` or one of methods of [`estimate_density()`],
#' such as `"kernel"`, `"logspline"` or `"KernSmooth"`. See details.
#' @param null The value considered as a "null" effect. Traditionally 0, but
#' could also be 1 in the case of ratios of change (OR, IRR, ...).
#' @inheritParams hdi
#'
#' @details
#' \subsection{What is the *pd*?}{
#' ## What is the *pd*?
#' The Probability of Direction (pd) is an index of effect existence, ranging
#' from `50%` to `100%`, representing the certainty with which an effect goes in
#' a particular direction (*i.e.*, is positive or negative). Beyond its
#' simplicity of interpretation, understanding and computation, this index also
#' presents other interesting properties:
#' \itemize{
#' \item It is independent from the model: It is solely based on the posterior
#' distributions and does not require any additional information from the data
#' or the model.
#' \item It is robust to the scale of both the response variable and the predictors.
#' \item It is strongly correlated with the frequentist p-value, and can thus
#' from 0 to 1, representing the certainty with which an effect goes in a
#' particular direction (*i.e.*, is positive or negative / has a sign). Beyond
#' its simplicity of interpretation, understanding and computation, this index
#' also presents other interesting properties:
#' - It is robust to the scale of both the response variable and the predictors.
#' - It is strongly correlated with the frequentist p-value, and can thus
#' be used to draw parallels and give some reference to readers non-familiar
#' with Bayesian statistics.
#' }
#' }
#' \subsection{Relationship with the p-value}{
#' In most cases, it seems that the *pd* has a direct correspondence with the frequentist one-sided *p*-value through the formula \ifelse{html}{\out{p<sub>one&nbsp;sided</sub>&nbsp;=&nbsp;1&nbsp;-&nbsp;<sup>p(<em>d</em>)</sup>/<sub>100</sub>}}{\eqn{p_{one sided}=1-\frac{p_{d}}{100}}} and to the two-sided p-value (the most commonly reported one) through the formula \ifelse{html}{\out{p<sub>two&nbsp;sided</sub>&nbsp;=&nbsp;2&nbsp;*&nbsp;(1&nbsp;-&nbsp;<sup>p(<em>d</em>)</sup>/<sub>100</sub>)}}{\eqn{p_{two sided}=2*(1-\frac{p_{d}}{100})}}. Thus, a two-sided p-value of respectively `.1`, `.05`, `.01` and `.001` would correspond approximately to a *pd* of `95%`, `97.5%`, `99.5%` and `99.95%`. See also [pd_to_p()].
#' }
#' \subsection{Methods of computation}{
#' The most simple and direct way to compute the *pd* is to 1) look at the
#' median's sign, 2) select the portion of the posterior of the same sign and
#' 3) compute the percentage that this portion represents. This "simple" method
#' is the most straightforward, but its precision is directly tied to the
#' number of posterior draws. The second approach relies on [density
#' estimation][estimate_density]. It starts by estimating the density function
#' (for which many methods are available), and then computing the [area under
#' the curve][area_under_curve] (AUC) of the density curve on the other side of
#' 0.
#' }
#' \subsection{Strengths and Limitations}{
#' **Strengths:** Straightforward computation and interpretation. Objective
#' property of the posterior distribution. 1:1 correspondence with the
#' frequentist p-value.
#' \cr \cr
#' **Limitations:** Limited information favoring the null hypothesis.
#' }
#' with Bayesian statistics (Makowski et al., 2019). See also [`pd_to_p()`].
#'
#' @return
#' Values between 0.5 and 1 corresponding to the probability of direction (pd).
#' ## Possible Range of Values
#' The largest value *pd* can take is 1 - the posterior is strictly directional.
#' However, the smallest value *pd* can take depends on the parameter space
#' represented by the posterior.
#' \cr\cr
#' Note that in some (rare) cases, especially when used with model averaged
#' posteriors (see [weighted_posteriors()] or
#' `brms::posterior_average`), `pd` can be smaller than `0.5`,
#' reflecting high credibility of `0`. To detect such cases, the
#' `method = "direct"` must be used.
#' **For a continuous parameter space**, exact values of 0 (or any point null
#' value) are not possible, and so 100% of the posterior has _some_ sign, some
#' positive, some negative. Therefore, the smallest the *pd* can be is 0.5 -
#' with an equal posterior mass of positive and negative values. Values close to
#' 0.5 _cannot_ be used to support the null hypothesis (that the parameter does
#' _not_ have a direction) is a similar why to how large p-values cannot be used
#' to support the null hypothesis (see [`pd_tp_p()`]; Makowski et al., 2019).
#' \cr\cr
#' **For a discrete parameter space or a parameter space that is a mixture
#' between discrete and continuous spaces**, exact values of 0 (or any point
#' null value) _are_ possible! Therefore, the smallest the *pd* can be is 0 -
#' with 100% of the posterior mass on 0. Thus values close to 0 can be used to
#' support the null hypothesis (see van den Bergh et al., 2021).
#' \cr\cr
#' Examples of posteriors representing discrete parameter space:
#' - When a parameter can only take discrete values.
#' - When a mixture prior/posterior is used (such as the spike-and-slab prior;
#' see van den Bergh et al., 2021).
#' - When conducting Bayesian model averaging (e.g., [weighted_posteriors()] or
#' `brms::posterior_average`).
#'
#' ## Methods of computation
#' The *pd* is defined as:
#' \deqn{p_d = max({Pr(\hat{\theta} < \theta_{null}), Pr(\hat{\theta} > \theta_{null})})}{pd = max(mean(x < null), mean(x > null))}

Check warning on line 58 in R/p_direction.R

View workflow job for this annotation

GitHub Actions / lint-changed-files / lint-changed-files

file=R/p_direction.R,line=58,col=121,[line_length_linter] Lines should not be more than 120 characters.
#' \cr\cr
#' The most simple and direct way to compute the *pd* is to compute the
#' proportion of positive (or larger than `null`) posterior samples, the
#' proportion of negative (or smaller than `null`) posterior samples, and take
#' the larger of the two. This "simple" method is the most straightforward, but
#' its precision is directly tied to the number of posterior draws.
#' \cr\cr
#' The second approach relies on [`density estimation()`]: It starts by
#' estimating the continuous-smooth density function (for which many methods are
#' available), and then computing the [area under the curve][area_under_curve]
#' (AUC) of the density curve on either side of `null` and taking the maximum
#' between them. Note the this approach assumes a continuous density function,
#' and so **when the posterior represents a (partially) discrete parameter
#' space, only the direct method _must_ be used** (see above).
#'
#' @return
#' Values between 0.5 and 1 *or* between 0 and 1 (see above) corresponding to
#' the probability of direction (pd).
#'
#' @seealso [pd_to_p()] to convert between Probability of Direction (pd) and p-value.
#'
#' @note There is also a [`plot()`-method](https://easystats.github.io/see/articles/bayestestR.html) implemented in the \href{https://easystats.github.io/see/}{\pkg{see}-package}.

Check warning on line 80 in R/p_direction.R

View workflow job for this annotation

GitHub Actions / lint-changed-files / lint-changed-files

file=R/p_direction.R,line=80,col=121,[line_length_linter] Lines should not be more than 120 characters.
#'
#' @references
#' Makowski D, Ben-Shachar MS, Chen SHA, Lüdecke D (2019) Indices of Effect
#' Existence and Significance in the Bayesian Framework. Frontiers in Psychology
#' 2019;10:2767. \doi{10.3389/fpsyg.2019.02767}
#' - Makowski, D., Ben-Shachar, M. S., Chen, S. A., & Lüdecke, D. (2019).
#' Indices of effect existence and significance in the Bayesian framework.
#' Frontiers in psychology, 10, 2767. \doi{10.3389/fpsyg.2019.02767}
#' - van den Bergh, D., Haaf, J. M., Ly, A., Rouder, J. N., & Wagenmakers, E. J.
#' (2021). A cautionary note on estimating effect size. Advances in Methods
#' and Practices in Psychological Science, 4(1). \doi{10.1177/2515245921992035}
#'
#' @examples
#' library(bayestestR)
Expand Down
126 changes: 75 additions & 51 deletions man/p_direction.Rd

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

Loading

0 comments on commit fecdef4

Please sign in to comment.