Setting up different types of specifications

The function setup() seems rather simple, but it is in fact very versatile and allows to incorporate (almost) any analytic choice that one might engage in during data analysis. In this vignette, I am going to exemplify how this function can be used to create several analytical choices.

library(tidyverse)
library(specr)

Different independent and dependent variables

One analytic choice may refer to choosing different independent or dependent variables (e.g., different scales that measure the same construct) or different ways in how these variables are compute (e.g., mean vs. latent measurement, log-transformation, etc).

Simply providing different variable as vectors

In most cases, this simply means adding such variables to the data set and provide them as alternative choices in setup.

# Setup specs
specs <- setup(data = example_data,
               x = c("x1", "x2", "x3", "x4"),
               y = c("y1", "y2", "y3"),
               model = "lm")

# Summary of specifications
summary(specs, rows = 12)
## Setup for the Specification Curve Analysis
## -------------------------------------------
## Class:                      specr.setup -- version: 1.0.0 
## Number of specifications:   12 
## 
## Specifications:
## 
##   Independent variable:     x1, x2, x3, x4 
##   Dependent variable:       y1, y2, y3 
##   Models:                   lm 
##   Covariates:               no covariates 
##   Subsets analyses:         all 
## 
## Function used to extract parameters:
## 
##   function (x) 
## broom::tidy(x, conf.int = TRUE)
## <environment: 0x7fe4ecafa320>
## 
## 
## Head of specifications table (first 12 rows):
## # A tibble: 12 × 6
##    x     y     model controls      subsets formula    
##    <chr> <chr> <chr> <chr>         <chr>   <glue>     
##  1 x1    y1    lm    no covariates all     y1 ~ x1 + 1
##  2 x1    y2    lm    no covariates all     y2 ~ x1 + 1
##  3 x1    y3    lm    no covariates all     y3 ~ x1 + 1
##  4 x2    y1    lm    no covariates all     y1 ~ x2 + 1
##  5 x2    y2    lm    no covariates all     y2 ~ x2 + 1
##  6 x2    y3    lm    no covariates all     y3 ~ x2 + 1
##  7 x3    y1    lm    no covariates all     y1 ~ x3 + 1
##  8 x3    y2    lm    no covariates all     y2 ~ x3 + 1
##  9 x3    y3    lm    no covariates all     y3 ~ x3 + 1
## 10 x4    y1    lm    no covariates all     y1 ~ x4 + 1
## 11 x4    y2    lm    no covariates all     y2 ~ x4 + 1
## 12 x4    y3    lm    no covariates all     y3 ~ x4 + 1
# Run analysis and plot results
results <- specr(specs)
head(results$data)
## # A tibble: 6 × 26
##   x     y     model controls      subsets formula  model_function term  estimate
##   <chr> <chr> <chr> <chr>         <chr>   <glue>   <list>         <chr>    <dbl>
## 1 x1    y1    lm    no covariates all     y1 ~ x1… <fn>           x1      0.620 
## 2 x1    y2    lm    no covariates all     y2 ~ x1… <fn>           x1     -0.328 
## 3 x1    y3    lm    no covariates all     y3 ~ x1… <fn>           x1     -0.0821
## 4 x2    y1    lm    no covariates all     y1 ~ x2… <fn>           x2      0.272 
## 5 x2    y2    lm    no covariates all     y2 ~ x2… <fn>           x2     -0.102 
## 6 x2    y3    lm    no covariates all     y3 ~ x2… <fn>           x2     -0.0735
## # … with 17 more variables: std.error <dbl>, statistic <dbl>, p.value <dbl>,
## #   conf.low <dbl>, conf.high <dbl>, fit_r.squared <dbl>,
## #   fit_adj.r.squared <dbl>, fit_sigma <dbl>, fit_statistic <dbl>,
## #   fit_p.value <dbl>, fit_df <dbl>, fit_logLik <dbl>, fit_AIC <dbl>,
## #   fit_BIC <dbl>, fit_deviance <dbl>, fit_df.residual <int>, fit_nobs <int>

Specific variable/model combinations

However, in some cases, we may have different type of variables (e.g., a mean vs. a dichotomous variable). In some cases, for example when this refers to a dependent variable, such a choice of variables requires also different model estimation functions. So in our setup, we want to acknowledge for specific variable/model combinations. In the following, let’s imagine that next to our standard dependent variables (y1-y3), we also have a dichotomous item (y_dich; here only recoded from y1). In this case, we want to specify and add a custom function - here a logistic model (glm(formula, data, family = binomial())) - to the setup.

# Dichotomous dependent variable
data <- example_data %>%
  mutate(y_dich = ifelse(y1 > mean(y1), 1, 0))

# Specific function
log_glm <- function(formula, data) {
  glm(formula, data, family = binomial())
}

# Setup specs
specs <- setup(data = data,
               x = c("x1", "x2"),
               y = c("y1", "y2", "y3", "y_dich"),
               model = c("lm", "log_glm"))
# Check
specs %>%
  as_tibble
## # A tibble: 16 × 7
##    x     y      model   controls      subsets formula         model_function
##    <chr> <chr>  <chr>   <chr>         <chr>   <glue>          <list>        
##  1 x1    y1     lm      no covariates all     y1 ~ x1 + 1     <fn>          
##  2 x1    y1     log_glm no covariates all     y1 ~ x1 + 1     <fn>          
##  3 x1    y2     lm      no covariates all     y2 ~ x1 + 1     <fn>          
##  4 x1    y2     log_glm no covariates all     y2 ~ x1 + 1     <fn>          
##  5 x1    y3     lm      no covariates all     y3 ~ x1 + 1     <fn>          
##  6 x1    y3     log_glm no covariates all     y3 ~ x1 + 1     <fn>          
##  7 x1    y_dich lm      no covariates all     y_dich ~ x1 + 1 <fn>          
##  8 x1    y_dich log_glm no covariates all     y_dich ~ x1 + 1 <fn>          
##  9 x2    y1     lm      no covariates all     y1 ~ x2 + 1     <fn>          
## 10 x2    y1     log_glm no covariates all     y1 ~ x2 + 1     <fn>          
## 11 x2    y2     lm      no covariates all     y2 ~ x2 + 1     <fn>          
## 12 x2    y2     log_glm no covariates all     y2 ~ x2 + 1     <fn>          
## 13 x2    y3     lm      no covariates all     y3 ~ x2 + 1     <fn>          
## 14 x2    y3     log_glm no covariates all     y3 ~ x2 + 1     <fn>          
## 15 x2    y_dich lm      no covariates all     y_dich ~ x2 + 1 <fn>          
## 16 x2    y_dich log_glm no covariates all     y_dich ~ x2 + 1 <fn>

As we can see, the setup() function produces all combinations between the variables and the model functions. This is not meaningful as the logistic model should only be used if the dependent variable is y_dich. In comparison to specr version 0.2.1, this is were the strength of version 1.0.0 comes in. Because we setup all specifications beforehand, we can simply remove those that we are not interested in or that we deem non-meaningful. Note: You won’t be able to filter or subset the specr.setup object (essentially a list). Instead, you want to filter/subset the first position of the list: the data frame that includes the specifications (here specs$specs or specs[[1]]).

# Filter out models that are not meaningful (here only keep log_glm, when y == "y4")
specs$specs <- specs$specs %>%
  filter(!(model == "log_glm" & y != "y_dich")) %>%
  filter(!(model == "lm" & y == "y_dich"))

# Check results (only meaningful specifications remain)
summary(specs, rows = 8)
## Setup for the Specification Curve Analysis
## -------------------------------------------
## Class:                      specr.setup -- version: 1.0.0 
## Number of specifications:   16 
## 
## Specifications:
## 
##   Independent variable:     x1, x2 
##   Dependent variable:       y1, y2, y3, y_dich 
##   Models:                   lm, log_glm 
##   Covariates:               no covariates 
##   Subsets analyses:         all 
## 
## Function used to extract parameters:
## 
##   function (x) 
## broom::tidy(x, conf.int = TRUE)
## <environment: 0x7fe4e7168210>
## 
## 
## Head of specifications table (first 8 rows):
## # A tibble: 8 × 6
##   x     y      model   controls      subsets formula        
##   <chr> <chr>  <chr>   <chr>         <chr>   <glue>         
## 1 x1    y1     lm      no covariates all     y1 ~ x1 + 1    
## 2 x1    y2     lm      no covariates all     y2 ~ x1 + 1    
## 3 x1    y3     lm      no covariates all     y3 ~ x1 + 1    
## 4 x1    y_dich log_glm no covariates all     y_dich ~ x1 + 1
## 5 x2    y1     lm      no covariates all     y1 ~ x2 + 1    
## 6 x2    y2     lm      no covariates all     y2 ~ x2 + 1    
## 7 x2    y3     lm      no covariates all     y3 ~ x2 + 1    
## 8 x2    y_dich log_glm no covariates all     y_dich ~ x2 + 1
# Run analysis and plot results
results <- specr(specs)
head(results$data)
## # A tibble: 6 × 28
##   x     y      model   controls    subsets formula model_function term  estimate
##   <chr> <chr>  <chr>   <chr>       <chr>   <glue>  <list>         <chr>    <dbl>
## 1 x1    y1     lm      no covaria… all     y1 ~ x… <fn>           x1      0.620 
## 2 x1    y2     lm      no covaria… all     y2 ~ x… <fn>           x1     -0.328 
## 3 x1    y3     lm      no covaria… all     y3 ~ x… <fn>           x1     -0.0821
## 4 x1    y_dich log_glm no covaria… all     y_dich… <fn>           x1      0.753 
## 5 x2    y1     lm      no covaria… all     y1 ~ x… <fn>           x2      0.272 
## 6 x2    y2     lm      no covaria… all     y2 ~ x… <fn>           x2     -0.102 
## # … with 19 more variables: std.error <dbl>, statistic <dbl>, p.value <dbl>,
## #   conf.low <dbl>, conf.high <dbl>, fit_r.squared <dbl>,
## #   fit_adj.r.squared <dbl>, fit_sigma <dbl>, fit_statistic <dbl>,
## #   fit_p.value <dbl>, fit_df <dbl>, fit_logLik <dbl>, fit_AIC <dbl>,
## #   fit_BIC <dbl>, fit_deviance <dbl>, fit_df.residual <int>, fit_nobs <int>,
## #   fit_null.deviance <dbl>, fit_df.null <int>

Adding composite and latent variables

In some cases, we may want to add mean indices or even latent measurement models of our variables of interest to explore how latent vs. manifest measurement affects our results. This requires a bit more upfront work as we need to work with lavaan for the latent measurement. Furthermore, we need to adjust the extract functions (fun1 and fun2 in setup()) so that the extraction of parameters aligns across different type of models. The beauty of broom::tidy() and broom::glance() (which are used by default) is that they adapt to the different model types that are passed to setup. Although the same parameters are extracted for most models, they sometimes differ for specific models (e.g., structural equation models resulting from lavaan::sem()). This means we need to adjust them to extract the same parameters across different models.

# Add mean (one choice)
data <- data %>%
  rowwise %>%
  mutate(x_mean = mean(x1, x2, x3, x4)) %>%
  ungroup

# Add custom function with latent measurement models to pass to "models" (another choice)
custom_sem <- function(formula, data) {
  
  # Make sure lavaan is loaded
  require(lavaan)
  
  # Add latent measurement as list
  latent <- list(latent_x12 = "latent_x12 =~ x1 + x2")
  
  # Remove +1 from formula as lavaan doesn't know how to process it 
  semformula <- str_remove_all(formula, "\\+ 1")
  
  # remove non-used latent measurement models from list by checking the formula
  valid <- purrr::keep(names(latent), ~ stringr::str_detect(formula, .x))
  
  # Create new formula that includes latent measurement models
  formula <- paste(formula, "\n", paste(latent[valid], collapse = " \n "))
  
  # Pass formula to `sem()`
  sem(formula, data)
}

# Create custom tidy function that extracts the same parameters from different models!
tidy_new <- function(x) {
  if(class(x) == "lavaan") {
    broom::tidy(x, conf.int = TRUE) %>% 
    select(term, estimate, conf.low, conf.high) %>%   # select parameters you want to keep
    filter(grepl(" ~ ", term)) %>%                    # term needs to be adjusted
    separate(term, c("dv", "term"), sep = " ~ ") %>%  # extract independent variable
    select(-dv)                                       # remove dependent variable
  } else {
    broom::tidy(x, conf.int = TRUE) %>% 
    select(term, estimate, conf.low, conf.high)      # same parameters as above
  }
}

# Setup specs with new custom function
specs <- setup(data = data,
               x = c("x1", "x2", "x3", "x4", "x_mean", "latent_x12"),
               y = c("y1", "y2"),
               model = c("lm", "custom_sem"),
               fun1 = tidy_new, # We pass the new extract function
               fun2 = NULL)     # switch off "glance" as it produces different fit indices and wouldn't work

# Quick check (still includes non-meaningful specifications)
summary(specs, rows = 12)
## Setup for the Specification Curve Analysis
## -------------------------------------------
## Class:                      specr.setup -- version: 1.0.0 
## Number of specifications:   24 
## 
## Specifications:
## 
##   Independent variable:     x1, x2, x3, x4, x_mean, latent_x12 
##   Dependent variable:       y1, y2 
##   Models:                   lm, custom_sem 
##   Covariates:               no covariates 
##   Subsets analyses:         all 
## 
## Function used to extract parameters:
## 
##   function(x) {
##   if(class(x) == "lavaan") {
##     broom::tidy(x, conf.int = TRUE) %>% 
##     select(term, estimate, conf.low, conf.high) %>%   # select parameters you want to keep
##     filter(grepl(" ~ ", term)) %>%                    # term needs to be adjusted
##     separate(term, c("dv", "term"), sep = " ~ ") %>%  # extract independent variable
##     select(-dv)                                       # remove dependent variable
##   } else {
##     broom::tidy(x, conf.int = TRUE) %>% 
##     select(term, estimate, conf.low, conf.high)      # same parameters as above
##   }
## }
## 
## 
## Head of specifications table (first 12 rows):
## # A tibble: 12 × 6
##    x     y     model      controls      subsets formula    
##    <chr> <chr> <chr>      <chr>         <chr>   <glue>     
##  1 x1    y1    lm         no covariates all     y1 ~ x1 + 1
##  2 x1    y1    custom_sem no covariates all     y1 ~ x1 + 1
##  3 x1    y2    lm         no covariates all     y2 ~ x1 + 1
##  4 x1    y2    custom_sem no covariates all     y2 ~ x1 + 1
##  5 x2    y1    lm         no covariates all     y1 ~ x2 + 1
##  6 x2    y1    custom_sem no covariates all     y1 ~ x2 + 1
##  7 x2    y2    lm         no covariates all     y2 ~ x2 + 1
##  8 x2    y2    custom_sem no covariates all     y2 ~ x2 + 1
##  9 x3    y1    lm         no covariates all     y1 ~ x3 + 1
## 10 x3    y1    custom_sem no covariates all     y1 ~ x3 + 1
## 11 x3    y2    lm         no covariates all     y2 ~ x3 + 1
## 12 x3    y2    custom_sem no covariates all     y2 ~ x3 + 1
# Filter out non-meaningful specifications
specs$specs <- specs$specs %>%
  filter(!(model == "custom_sem" & !grepl("latent", x))) %>%
  filter(!(model == "lm" & grepl("latent", x)))

# Check again
summary(specs, rows = 12)
## Setup for the Specification Curve Analysis
## -------------------------------------------
## Class:                      specr.setup -- version: 1.0.0 
## Number of specifications:   24 
## 
## Specifications:
## 
##   Independent variable:     x1, x2, x3, x4, x_mean, latent_x12 
##   Dependent variable:       y1, y2 
##   Models:                   lm, custom_sem 
##   Covariates:               no covariates 
##   Subsets analyses:         all 
## 
## Function used to extract parameters:
## 
##   function(x) {
##   if(class(x) == "lavaan") {
##     broom::tidy(x, conf.int = TRUE) %>% 
##     select(term, estimate, conf.low, conf.high) %>%   # select parameters you want to keep
##     filter(grepl(" ~ ", term)) %>%                    # term needs to be adjusted
##     separate(term, c("dv", "term"), sep = " ~ ") %>%  # extract independent variable
##     select(-dv)                                       # remove dependent variable
##   } else {
##     broom::tidy(x, conf.int = TRUE) %>% 
##     select(term, estimate, conf.low, conf.high)      # same parameters as above
##   }
## }
## 
## 
## Head of specifications table (first 12 rows):
## # A tibble: 12 × 6
##    x          y     model      controls      subsets formula            
##    <chr>      <chr> <chr>      <chr>         <chr>   <glue>             
##  1 x1         y1    lm         no covariates all     y1 ~ x1 + 1        
##  2 x1         y2    lm         no covariates all     y2 ~ x1 + 1        
##  3 x2         y1    lm         no covariates all     y1 ~ x2 + 1        
##  4 x2         y2    lm         no covariates all     y2 ~ x2 + 1        
##  5 x3         y1    lm         no covariates all     y1 ~ x3 + 1        
##  6 x3         y2    lm         no covariates all     y2 ~ x3 + 1        
##  7 x4         y1    lm         no covariates all     y1 ~ x4 + 1        
##  8 x4         y2    lm         no covariates all     y2 ~ x4 + 1        
##  9 x_mean     y1    lm         no covariates all     y1 ~ x_mean + 1    
## 10 x_mean     y2    lm         no covariates all     y2 ~ x_mean + 1    
## 11 latent_x12 y1    custom_sem no covariates all     y1 ~ latent_x12 + 1
## 12 latent_x12 y2    custom_sem no covariates all     y2 ~ latent_x12 + 1
# Run analysis and plot results
results <- specr(specs)
plot(results, choices = c("x", "y"))

As we can see, we can now investigate the manifest and latent measures next to the individual items.

Control variables

Another analytical choice may refer to the inclusion of covariates. Here, the setup() function is again very versatile and allows to include different combinations of covariates.

Adding a bunch of control variables

By providing just a vector of covariates, setup() produces all combinations of these covariates plus a specification without any covariates. If we add the argument simplify = TRUE, not all combinations between covariates are created. Instead, only no covariates, each individually, and all together are included.

# Setup specification that include all combinations of covariates
specs1 <- setup(data = example_data,
               x = c("x1", "x2"),
               y = c("y1", "y2"),
               model = "lm",
               controls = c("c1", "c2", "c3", "c4"))  # simply providing a vector of control variables

# Setup secifications that include only no covariates, each individually, and all together
specs2 <- setup(data = example_data,
               x = c("x1", "x2"),
               y = c("y1", "y2"),
               model = "lm",
               controls = c("c1", "c2", "c3", "c4"),
               simplify = TRUE)   # Difference to specs1!

# Check
distinct(specs1$specs, controls)
## # A tibble: 16 × 1
##    controls         
##    <chr>            
##  1 no covariates    
##  2 c1               
##  3 c2               
##  4 c3               
##  5 c4               
##  6 c1 + c2          
##  7 c1 + c3          
##  8 c1 + c4          
##  9 c2 + c3          
## 10 c2 + c4          
## 11 c3 + c4          
## 12 c1 + c2 + c3     
## 13 c1 + c2 + c4     
## 14 c1 + c3 + c4     
## 15 c2 + c3 + c4     
## 16 c1 + c2 + c3 + c4
distinct(specs2$specs, controls)
## # A tibble: 6 × 1
##   controls      
##   <chr>         
## 1 no covariates 
## 2 c1            
## 3 c2            
## 4 c3            
## 5 c4            
## 6 all covariates

Sets of covariates

We can also add groups of covariates. The way setup creates combinations between such groups remains the same.

# Add groups of covariates
specs3 <- setup(data = example_data,
               x = c("x1", "x2"),
               y = c("y1", "y2"),
               model = "lm",
               controls = c("c1 + c2", "c3 + c4"))
# Check
distinct(specs3$specs, controls)
## # A tibble: 4 × 1
##   controls         
##   <chr>            
## 1 no covariates    
## 2 c1 + c2          
## 3 c3 + c4          
## 4 c1 + c2 + c3 + c4

The setup() function further has an argument add_to_formula, which allows to add covariates (or any other formula-relevant aspects, e.g., random effect structures) to all model functions.

# Add some control variables to all models
specs4 <- setup(data = example_data,
               x = c("x1", "x2"),
               y = c("y1", "y2"),
               model = "lm",
               controls = c("c1", "c2"),
               add_to_formula = "c3")

# Check (see how `c3` is added to each formula, but is not part of controls)
specs4$specs[1:6,]
## # A tibble: 6 × 7
##   x     y     model controls      subsets formula                model_function
##   <chr> <chr> <chr> <chr>         <chr>   <glue>                 <list>        
## 1 x1    y1    lm    no covariates all     y1 ~ x1 + 1 + c3       <fn>          
## 2 x1    y1    lm    c1            all     y1 ~ x1 + c1 + c3      <fn>          
## 3 x1    y1    lm    c2            all     y1 ~ x1 + c2 + c3      <fn>          
## 4 x1    y1    lm    c1 + c2       all     y1 ~ x1 + c1 + c2 + c3 <fn>          
## 5 x1    y2    lm    no covariates all     y2 ~ x1 + 1 + c3       <fn>          
## 6 x1    y2    lm    c1            all     y2 ~ x1 + c1 + c3      <fn>

Dealing with duplicates in specifications

Finally, we may sometimes decide to add one of the independent or dependent variables as covariates as well. In this case, it would not make sense to add e.g., x1 as covariate if it is already the independent variable. In this new version of specr, the setup() function automatically detects such duplications and deletes such specifications.

# Adding a covariate that is also a independent or dependent variable
specs5 <- setup(data = example_data,
               x = c("x1", "x2"),
               y = c("y1", "y2"),
               model = "lm",
               controls = c("x1", "y1"))

# Check (see how only 9 specifications are kept)
specs5$specs
## # A tibble: 9 × 7
##   x     y     model controls      subsets formula           model_function
##   <chr> <chr> <chr> <chr>         <chr>   <glue>            <list>        
## 1 x1    y1    lm    no covariates all     y1 ~ x1 + 1       <fn>          
## 2 x1    y2    lm    no covariates all     y2 ~ x1 + 1       <fn>          
## 3 x1    y2    lm    y1            all     y2 ~ x1 + y1      <fn>          
## 4 x2    y1    lm    no covariates all     y1 ~ x2 + 1       <fn>          
## 5 x2    y1    lm    x1            all     y1 ~ x2 + x1      <fn>          
## 6 x2    y2    lm    no covariates all     y2 ~ x2 + 1       <fn>          
## 7 x2    y2    lm    x1            all     y2 ~ x2 + x1      <fn>          
## 8 x2    y2    lm    y1            all     y2 ~ x2 + y1      <fn>          
## 9 x2    y2    lm    x1 + y1       all     y2 ~ x2 + x1 + y1 <fn>

Subset analyses

One of the most powerful arguments within the setup() framework is subsets, which can be used for so-called “subset analyses”. With this argument, a variety of analytical choices can be included.

Simple subset analysis based on grouping variables

In simple cases, we may want to investigate whether a relationship differs across certain subgroups. In the example data, we can for example investigate whether the relationship between x and y differs across age groups (group1) and gender (group2). We need to can add such subset analyses by specifying a list in two alternative ways. The easiest way refers to simply providing named vectors that refer to the respective variables in the data set and their unique values:

list(group1 = c("young", "middle", "old),
      group2 = c("female", "male"))

Alternatively, we can achieve the same thing by using the function unique(), which extracts the unique values from each variable in the data set:

list(group1 = unique(example_data$group1),
     group2 = unique(example_data$group2))

These lists are then passed as combinatorial factors to the expand function within setup(). If two or more grouping variables are provided in this way (you can provide as many as you want, but it quickly becomes messy), setup() automatically adds them as “choice” columns to the specification setup data frame. Note how it automatically adds the value NA to make sure that also the simple subsets and not only the combinations between the two groups are included in the specifications. It further creates a new column “subsets”, which represents the combination of the subsetting factors.

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

# Summary of specifications
summary(specs)
## Setup for the Specification Curve Analysis
## -------------------------------------------
## Class:                      specr.setup -- version: 1.0.0 
## Number of specifications:   96 
## 
## Specifications:
## 
##   Independent variable:     x1, x2 
##   Dependent variable:       y1, y2 
##   Models:                   lm 
##   Covariates:               no covariates, c1 
##   Subsets analyses:         middle & female, old & female, young & female, female, middle & male, old & male, young & male, male, middle, old, young, all 
## 
## Function used to extract parameters:
## 
##   function (x) 
## broom::tidy(x, conf.int = TRUE)
## <environment: 0x7fe4dcc30aa0>
## 
## 
## Head of specifications table (first 6 rows):
## # A tibble: 6 × 8
##   x     y     model controls      subsets         group1 group2 formula    
##   <chr> <chr> <chr> <chr>         <chr>           <fct>  <fct>  <glue>     
## 1 x1    y1    lm    no covariates middle & female middle female y1 ~ x1 + 1
## 2 x1    y1    lm    no covariates old & female    old    female y1 ~ x1 + 1
## 3 x1    y1    lm    no covariates young & female  young  female y1 ~ x1 + 1
## 4 x1    y1    lm    no covariates female          <NA>   female y1 ~ x1 + 1
## 5 x1    y1    lm    no covariates middle & male   middle male   y1 ~ x1 + 1
## 6 x1    y1    lm    no covariates old & male      old    male   y1 ~ x1 + 1
# Check subsets (in this case, 12 different types of subset analysis,
# including using "all" subjects)
distinct(specs$specs, subsets)
## # A tibble: 12 × 1
##    subsets        
##    <chr>          
##  1 middle & female
##  2 old & female   
##  3 young & female 
##  4 female         
##  5 middle & male  
##  6 old & male     
##  7 young & male   
##  8 male           
##  9 middle         
## 10 old            
## 11 young          
## 12 all
# Run analysis and plot results
results <- specr(specs)
plot(results, choices = c("x", "y", "subsets"))

More complex analysis using the subsets functionality

The true potential of this type of subsetting can only be grasped with another example. Let’s imagine we want to include the removal of outliers as an analytical choice in our specification curve analysis. Perhaps we are unsure about the cut-off value and want to investigate different (arbitrary) ones. This can again be seen as a subset analysis.

We first create variables that denote who is an outlier and who is not according to the rule we want to use. We then pass these variables as subsets to the setup() function. This again produces non-meaningful specifications (e.g., a subset that includes outliers, a subset that includes people who align with different outlier rules, etc.). So we want to keep only the meaningful specifications. In this case, these are just 7 different types of specifications.

# Create variables that denote outliers (here with a range of arbitrary thresholds)
data <- data %>%
  mutate(outlier1 = ifelse(y1 < mean(y1) - 2*sd(y1) | y1 > mean(y1) + 2*sd(y1), "outlier", "2.0*SD"),
         outlier2 = ifelse(y1 < mean(y1) - 2.1*sd(y1) | y1 > mean(y1) + 2.1*sd(y1), "outlier", "2.1*SD"),
         outlier3 = ifelse(y1 < mean(y1) - 2.2*sd(y1) | y1 > mean(y1) + 2.2*sd(y1), "outlier", "2.2*SD"),
         outlier4 = ifelse(y1 < mean(y1) - 2.3*sd(y1) | y1 > mean(y1) + 2.3*sd(y1), "outlier", "2.3*SD"),
         outlier5 = ifelse(y1 < mean(y1) - 2.4*sd(y1) | y1 > mean(y1) + 2.4*sd(y1), "outlier", "2.4*SD"),
         outlier6 = ifelse(y1 < mean(y1) - 2.5*sd(y1) | y1 > mean(y1) + 2.5*sd(y1), "outlier", "2.5*SD"))

# Setup specs
specs <- setup(data = data,
               x = c("x1", "x2"),
               y = c("y1", "y2"),
               model = "lm",
               controls = c("c1", "c2"),
               subsets = list(outlier1 = c("2.0*SD"),
                              outlier2 = c("2.1*SD"),
                              outlier3 = c("2.2*SD"),
                              outlier4 = c("2.3*SD"),
                              outlier5 = c("2.4*SD"),
                              outlier6 = c("2.5*SD")))

# Remove unnecessary combinations
specs$specs <- specs$specs %>%
  filter(subsets == "2.0*SD" | subsets == "2.1*SD" | 
         subsets == "2.2*SD" | subsets == "2.3*SD" |
         subsets == "2.4*SD" | subsets == "2.5*SD" |  
         subsets == "all")

# Check specifications (see how it contains only meaningful subsets?)
summary(specs, rows = 7)
## Setup for the Specification Curve Analysis
## -------------------------------------------
## Class:                      specr.setup -- version: 1.0.0 
## Number of specifications:   1024 
## 
## Specifications:
## 
##   Independent variable:     x1, x2 
##   Dependent variable:       y1, y2 
##   Models:                   lm 
##   Covariates:               no covariates, c1, c2, c1 + c2 
##   Subsets analyses:         2.5*SD, 2.4*SD, 2.3*SD, 2.2*SD, 2.1*SD, 2.0*SD, all 
## 
## Function used to extract parameters:
## 
##   function (x) 
## broom::tidy(x, conf.int = TRUE)
## <environment: 0x7fe4d5e50378>
## 
## 
## Head of specifications table (first 7 rows):
## # A tibble: 7 × 12
##   x     y     model controls      subsets outlier1 outlier2 outlier3 outlier4
##   <chr> <chr> <chr> <chr>         <chr>   <fct>    <fct>    <fct>    <fct>   
## 1 x1    y1    lm    no covariates 2.5*SD  <NA>     <NA>     <NA>     <NA>    
## 2 x1    y1    lm    no covariates 2.4*SD  <NA>     <NA>     <NA>     <NA>    
## 3 x1    y1    lm    no covariates 2.3*SD  <NA>     <NA>     <NA>     2.3*SD  
## 4 x1    y1    lm    no covariates 2.2*SD  <NA>     <NA>     2.2*SD   <NA>    
## 5 x1    y1    lm    no covariates 2.1*SD  <NA>     2.1*SD   <NA>     <NA>    
## 6 x1    y1    lm    no covariates 2.0*SD  2.0*SD   <NA>     <NA>     <NA>    
## 7 x1    y1    lm    no covariates all     <NA>     <NA>     <NA>     <NA>    
## # … with 3 more variables: outlier5 <fct>, outlier6 <fct>, formula <glue>
# Run analysis and plot results
results <- specr(specs)
plot(results, choices = c("x", "y", "subsets"))

There are of course many more types of analytical decisions. For some more examples, see some of the other vignettes. If you have a specific type of analytical decision that you don’t know how to include, feel free to open an issue on github.