diff --git a/R/adjust-numeric-calibration.R b/R/adjust-numeric-calibration.R
index 402ddbc..8170961 100644
--- a/R/adjust-numeric-calibration.R
+++ b/R/adjust-numeric-calibration.R
@@ -84,7 +84,7 @@ print.numeric_calibration <- function(x, ...) {
 
 #' @export
 fit.numeric_calibration <- function(object, data, tailor = NULL, ...) {
-  method <- check_method(object$method, tailor$type)
+  method <- check_method(object$arguments$method, tailor$type)
   # todo: adjust_numeric_calibration() should take arguments to pass to
   # cal_estimate_* via dots
   fit <-
diff --git a/R/adjust-probability-calibration.R b/R/adjust-probability-calibration.R
index 089399b..38fc04e 100644
--- a/R/adjust-probability-calibration.R
+++ b/R/adjust-probability-calibration.R
@@ -87,7 +87,7 @@ print.probability_calibration <- function(x, ...) {
 
 #' @export
 fit.probability_calibration <- function(object, data, tailor = NULL, ...) {
-  method <- check_method(object$method, tailor$type)
+  method <- check_method(object$arguments$method, tailor$type)
   # todo: adjust_probability_calibration() should take arguments to pass to
   # cal_estimate_* via dots
   fit <-
diff --git a/tests/testthat/test-adjust-numeric-calibration.R b/tests/testthat/test-adjust-numeric-calibration.R
index 379907d..2e4c37d 100644
--- a/tests/testthat/test-adjust-numeric-calibration.R
+++ b/tests/testthat/test-adjust-numeric-calibration.R
@@ -32,7 +32,39 @@ test_that("basic adjust_numeric_calibration usage works", {
   # TODO: write out the probably code manually here
 })
 
-# TODO: test sensitivity to function arguments
+test_that("adjust_numeric_calibration() respects `method` argument", {
+  library(tibble)
+
+  set.seed(1)
+  d_calibration <- tibble(y = rnorm(100), y_pred = y/2 + rnorm(100))
+  d_test <- tibble(y = rnorm(100), y_pred = y/2 + rnorm(100))
+
+  expect_no_condition(
+    tlr <-
+      tailor() %>%
+      adjust_numeric_calibration(method = "isotonic")
+  )
+
+  # TODO: cannot be `expect_no_condition()` due to tidymodels/probably#157
+  expect_no_error(expect_no_warning(
+    tlr_fit <- fit(tlr, d_calibration, outcome = y, estimate = y_pred)
+  ))
+
+  expect_no_error(expect_no_warning(
+    tlr_pred <- predict(tlr_fit, d_test)
+  ))
+
+  # classes are as expected
+  expect_s3_class(tlr, "tailor")
+  expect_s3_class(tlr_fit, "tailor")
+  expect_s3_class(tlr_pred, "tbl_df")
+
+  # column names are as expected
+  expect_equal(colnames(d_test), colnames(tlr_pred))
+
+  # probably actually used an isotonic calibrator
+  expect_equal(tlr_fit$adjustments[[1]]$results$fit$method, "Isotonic regression")
+})
 
 test_that("adjustment printing", {
   expect_snapshot(tailor() %>% adjust_numeric_calibration())
diff --git a/tests/testthat/test-adjust-probability-calibration.R b/tests/testthat/test-adjust-probability-calibration.R
index 5704466..0ca01b7 100644
--- a/tests/testthat/test-adjust-probability-calibration.R
+++ b/tests/testthat/test-adjust-probability-calibration.R
@@ -41,6 +41,49 @@ test_that("basic adjust_probability_calibration() usage works", {
   # TODO: write out the manual code with probably
 })
 
+test_that("basic adjust_probability_calibration() usage works", {
+  skip_if_not_installed("modeldata")
+  library(modeldata)
+
+  # split example data
+  set.seed(1)
+  in_rows <- sample(c(TRUE, FALSE), nrow(two_class_example), replace = TRUE)
+  d_calibration <- two_class_example[in_rows, ]
+  d_test <- two_class_example[!in_rows, ]
+
+  # fitting and predicting happens without raising conditions
+  expect_no_condition(
+    tlr <-
+      tailor() %>%
+      adjust_probability_calibration(method = "isotonic")
+  )
+
+  # TODO: cannot be `expect_no_condition()` due to tidymodels/probably#157
+  expect_no_error(expect_no_warning(
+    tlr_fit <- fit(
+      tlr,
+      d_calibration,
+      outcome = c(truth),
+      estimate = c(predicted),
+      probabilities = c(Class1, Class2)
+    )
+  ))
+
+  expect_no_error(expect_no_warning(
+    tlr_pred <- predict(tlr_fit, d_test)
+  ))
+
+  # classes are as expected
+  expect_s3_class(tlr, "tailor")
+  expect_s3_class(tlr_fit, "tailor")
+  expect_s3_class(tlr_pred, "tbl_df")
+
+  # column names are as expected
+  expect_equal(colnames(d_test), colnames(tlr_pred))
+
+  # probably actually used an isotonic calibrator
+  expect_equal(tlr_fit$adjustments[[1]]$results$fit$method, "Isotonic regression")
+})
 
 test_that("adjustment printing", {
   expect_snapshot(tailor() %>% adjust_probability_calibration("logistic"))