Tidycomm provides the test_icr() function to conveniently compute intercoder reliability tests for several variables and reliability estimates at the same time.
Test data has to be structured in a long format, with one column indicating the unit (e.g., article 1, article 2, etc.), one column indicating the coder (either by a string or a numeric ID), and one column each per coded variable to test.
For demonstration purposes, we will use the fbposts data included in Tidycomm that consists of 45 political Facebook posts (identified by post_id) coded by six coders (identified by coder_id) for various formal (post type, number of pictures used in post) and populism-related (attacks on elites, references to ‘the people’, othering) features:
fbposts
#> # A tibble: 270 x 7
#> post_id coder_id type n_pictures pop_elite pop_people pop_othering
#> <int> <int> <chr> <int> <int> <int> <int>
#> 1 1 1 photo 1 0 0 0
#> 2 1 2 photo 1 0 0 0
#> 3 1 3 photo 1 0 0 0
#> 4 1 4 photo 1 0 0 0
#> 5 1 5 photo 1 0 0 0
#> 6 1 6 photo 1 0 0 0
#> 7 2 1 photo 1 0 0 0
#> 8 2 2 photo 1 0 0 0
#> 9 2 3 photo 1 0 0 0
#> 10 2 4 photo 1 0 0 0
#> # ... with 260 more rowstest_icr() computes various intercoder reliability estimates for all specified variables. The first two arguments (in a pipe) are the unit-identifying variable and the coder-identifying variable, followed by the test variables:
fbposts %>%
test_icr(post_id, coder_id, pop_elite, pop_people, pop_othering)
#> # A tibble: 3 x 8
#> Variable n_Units n_Coders n_Categories Level Agreement Holstis_CR
#> <chr> <int> <int> <int> <chr> <dbl> <dbl>
#> 1 pop_eli~ 45 6 6 nomi~ 0.733 0.861
#> 2 pop_peo~ 45 6 2 nomi~ 0.778 0.916
#> 3 pop_oth~ 45 6 4 nomi~ 0.867 0.945
#> # ... with 1 more variable: Krippendorffs_Alpha <dbl>If no test variables are specified, all variables in the dataset (excluding the unit and coder variables) will be tested:
fbposts %>%
test_icr(post_id, coder_id)
#> # A tibble: 5 x 8
#> Variable n_Units n_Coders n_Categories Level Agreement Holstis_CR
#> <chr> <int> <int> <int> <chr> <dbl> <dbl>
#> 1 type 45 6 4 nomi~ 1 1
#> 2 n_pictu~ 45 6 7 nomi~ 0.822 0.930
#> 3 pop_eli~ 45 6 6 nomi~ 0.733 0.861
#> 4 pop_peo~ 45 6 2 nomi~ 0.778 0.916
#> 5 pop_oth~ 45 6 4 nomi~ 0.867 0.945
#> # ... with 1 more variable: Krippendorffs_Alpha <dbl>Currently, test_icr() supports the following reliability estimates:
agreement: Simple percent agreement.holsti: Holsti’s \(CR\) (mean pairwise percent agreement).kripp_alpha: Krippendorff’s \(\alpha\).cohens_kappa: Cohen’s \(\kappa\) (only available for two coders).fleiss_kappa: Fleiss’ \(\kappa\).brennan_prediger: Brennan & Prediger’s \(\kappa\) (for more than two coders, von Eye’s (2006) proposed extension to multiple coders is computed).By default, test_icr() will output simple percent agreement, Holsti’s \(CR\), and Krippendorff’s \(\alpha\) as reliability estimates. You can add other estimates by setting their name to TRUE in the function call (and remove the default ones by setting them to FALSE):
fbposts %>%
test_icr(post_id, coder_id, fleiss_kappa = TRUE, agreement = FALSE)
#> # A tibble: 5 x 8
#> Variable n_Units n_Coders n_Categories Level Holstis_CR Krippendorffs_A~
#> <chr> <int> <int> <int> <chr> <dbl> <dbl>
#> 1 type 45 6 4 nomi~ 1 1
#> 2 n_pictu~ 45 6 7 nomi~ 0.930 0.880
#> 3 pop_eli~ 45 6 6 nomi~ 0.861 0.339
#> 4 pop_peo~ 45 6 2 nomi~ 0.916 0.287
#> 5 pop_oth~ 45 6 4 nomi~ 0.945 0.566
#> # ... with 1 more variable: Fleiss_Kappa <dbl>By default, test_icr() assumes all test variables to be nominal. You can set other variable levels by passing a named vector of the form c(variable_name = "variable_level") to the levels argument.
fbposts %>%
test_icr(post_id, coder_id, levels = c(n_pictures = "ordinal"))
#> # A tibble: 5 x 8
#> Variable n_Units n_Coders n_Categories Level Agreement Holstis_CR
#> <chr> <int> <int> <int> <chr> <dbl> <dbl>
#> 1 type 45 6 4 nomi~ 1 1
#> 2 n_pictu~ 45 6 7 ordi~ 0.822 0.930
#> 3 pop_eli~ 45 6 6 nomi~ 0.733 0.861
#> 4 pop_peo~ 45 6 2 nomi~ 0.778 0.916
#> 5 pop_oth~ 45 6 4 nomi~ 0.867 0.945
#> # ... with 1 more variable: Krippendorffs_Alpha <dbl>Nominal test variables can be represented by either integer codes or string labels, whereas ordinal variables must be represented by integer codes, and interval/ratio variables must be numeric (integer or float).
Please note that currently only the computation of Krippendorff’s \(\alpha\) is influenced by the variable level.
Missing values in intercoder reliability tests can be ambiguous (did the coder forget to code this variable for this unit, or does the missing value indicate that none of the categories was deemed fitting?) and present an obstacle to several reliability estimates (of the currently implemented estimates, only Krippendorff’s \(\alpha\) can deal with missing values).
Thus, test_icr() will by default respond with a warning when NA values are present in the test variables and output NA for all reliability estimates but Krippendorff’s \(\alpha\):
# Introduce some missing values
fbposts$type[1] <- NA
fbposts$type[2] <- NA
fbposts$pop_elite[5] <- NA
fbposts %>%
test_icr(post_id, coder_id)
#> Warning: Variable 'type' contains missing values. Consider setting na.omit
#> = TRUE or recoding missing values
#> Warning: Variable 'pop_elite' contains missing values. Consider setting
#> na.omit = TRUE or recoding missing values
#> # A tibble: 5 x 8
#> Variable n_Units n_Coders n_Categories Level Agreement Holstis_CR
#> <chr> <int> <int> <int> <chr> <dbl> <dbl>
#> 1 type 45 6 4 nomi~ NA NA
#> 2 n_pictu~ 45 6 7 nomi~ 0.822 0.930
#> 3 pop_eli~ 45 6 6 nomi~ NA NA
#> 4 pop_peo~ 45 6 2 nomi~ 0.778 0.916
#> 5 pop_oth~ 45 6 4 nomi~ 0.867 0.945
#> # ... with 1 more variable: Krippendorffs_Alpha <dbl>You can set na.omit = TRUE to exclude all units with NA values for a specific test variable from the computation for this variable:
fbposts %>%
test_icr(post_id, coder_id, na.omit = TRUE)
#> # A tibble: 5 x 8
#> Variable n_Units n_Coders n_Categories Level Agreement Holstis_CR
#> <chr> <int> <int> <int> <chr> <dbl> <dbl>
#> 1 type 44 6 4 nomi~ 1 1
#> 2 n_pictu~ 45 6 7 nomi~ 0.822 0.930
#> 3 pop_eli~ 44 6 6 nomi~ 0.727 0.858
#> 4 pop_peo~ 45 6 2 nomi~ 0.778 0.916
#> 5 pop_oth~ 45 6 4 nomi~ 0.867 0.945
#> # ... with 1 more variable: Krippendorffs_Alpha <dbl>