Plot specification curve and analytic choices — plot.specr.object • specr

This function plots visualizations of the specification curve analysis. The function requires an object of class specr.object, usually the results of calling specr() to create a standard visualization of the specification curve analysis. Several types of visualizations are possible.

# S3 method for specr.object
plot(
  x,
  type = "default",
  var = .data$estimate,
  group = NULL,
  choices = c("x", "y", "model", "controls", "subsets"),
  labels = c("A", "B"),
  rel_heights = c(2, 3),
  desc = FALSE,
  null = 0,
  ci = TRUE,
  ribbon = FALSE,
  formula = NULL,
  print = TRUE,
  ...
)

Arguments

x: A specr.object object, usually resulting from calling specr().
type: What type of figure should be plotted? If type = "default", the standard specification curve analysis plot (the specification curve as the upper panel and an overview of the relevant choices as the lower panel) is created. If type = "curve", only the specification curve (upper panel of the default plot) is plotted. If type = "choices", only the choice panel (lower part of the default plot) is plotted. If type = "boxplot", an alternative visualization of differences between choices is plotted that summarizes results per choice using box-and-whisker plot(s). If type = "samplesizes", a barplot of sample sizes per specification is plotted. See examples for more information.
var: Which parameter should be plotted in the curve? Defaults to estimate, but other parameters (e.g., p.value, fit_r.squared,...) can be plotted too.
group: Should the arrangement of the curve be grouped by a particular choice? Defaults to NULL, but can be any of the present choices (e.g., x, y, controls...)
choices: A vector specifying which analytic choices should be plotted. By default, all choices (x, y, model, controls, subsets) are plotted.
labels: Labels for the two parts of the plot
rel_heights: vector indicating the relative heights of the plot.
desc: Logical value indicating whether the curve should the arranged in a descending order. Defaults to FALSE.
null: Indicate what value represents the 'null' hypothesis (defaults to zero).
ci: Logical value indicating whether confidence intervals should be plotted.
ribbon: Logical value indicating whether a ribbon instead should be plotted
formula: In combination with type = "variance", you can provide a specific formula to extract specific variance components. The syntax of the formula is based on lme4::lmer() and thus looks something like, e.g.: "estimate ~ 1 + (1|x) + (1|y)" (to estimate the amount of variance explained by different independent x and dependent variables y). All other choices are then subsumed under residual variance. By no formula is provided, all choices (x, y, model, controls, and subsets) that have more than one alternative are included. See examples for further details.
print: In combination with type = "variance", logical value indicating whether the intra-class correlations (i.e., percentages of variance explained by analstical choices) should be printed or not. Defaults to TRUE.
...: further arguments passed to or from other methods (currently ignored).

Value

A ggplot object that can be customized further.

Examples

if (FALSE) {
# Specification Curve analysis ----
# Setup specifications
specs <- setup(data = example_data,
   y = c("y1", "y2"),
   x = c("x1", "x2"),
   model = "lm",
   controls = c("c1", "c2"),
   subsets = list(group1 = unique(example_data$group1),
                  group2 = unique(example_data$group2)))

# Run analysis
results <- specr(specs)

# Resulting data frame with estimates
as_tibble(results)  # This will be used for plotting


# Visualizations ---
# Plot results in various ways
plot(results)                            # default
plot(results, choices = c("x", "y"))     # specific choices
plot(results, ci = FALSE, ribbon = TRUE) # exclude CI and add ribbon instead
plot(results, type = "curve")
plot(results, type = "choices")
plot(results, type = "samplesizes")
plot(results, type = "boxplot")


# Grouped plot
plot(results, group = controls)

# Alternative and specific visualizations ----
# Other variables in the resulting data set can be plotted too
plot(results,
     type = "curve",
     var = fit_r.squared,   # extract "r-square" instead of "estimate"
     ci = FALSE)

# Such a plot can also be extended (e.g., by again adding the estimates with
# confidence intervals)
library(ggplot2)
plot(results, type = "curve", var = fit_r.squared) +
  geom_point(aes(y = estimate), shape = 5) +
  labs(x = "specifications", y = "r-squared | estimate")

# We can also investigate how much variance is explained by each analytical choice
plot(results, type = "variance")

# By providing a specific formula in `lme4::lmer()`-style, we can extract specific choices
# and also include interactions between chocies
plot(results,
     type = "variance",
     formula = "estimate ~ 1 + (1|x) + (1|y) + (1|group1) + (1|x:y)")

## Combining several plots ----
# `specr` also exports the function `plot_grid()` from the package `cowplot`, which
# can be used to combine plots meaningfully
a <- plot(results, "curve")
b <- plot(results, "choices", choices = c("x", "y", "controls"))
c <- plot(results, "samplesizes")
plot_grid(a, b, c,
          align = "v",
          axis = "rbl",
          rel_heights = c(2, 3, 1),
          ncol = 1)
}