diff --git a/R/measures.R b/R/measures.R
index 2b3bba441a..b27abbd11b 100644
--- a/R/measures.R
+++ b/R/measures.R
@@ -974,7 +974,7 @@ measureFN = function(truth, response, negative) {
 tpr = makeMeasure(id = "tpr", minimize = FALSE, best = 1, worst = 0,
   properties = c("classif", "req.pred", "req.truth"),
   name = "True positive rate",
-  note = "Percentage of correctly classified observations in the positive class. Also called hit rate or recall.",
+  note = "Percentage of correctly classified observations in the positive class. Also called hit rate or recall or sensitivity.",
   fun = function(task, model, pred, feats, extra.args) {
     measureTPR(pred$data$truth, pred$data$response, pred$task.desc$positive)
   }
@@ -1050,7 +1050,7 @@ measureFNR = function(truth, response, negative, positive) {
 ppv = makeMeasure(id = "ppv", minimize = FALSE, best = 1, worst = 0,
   properties = c("classif", "req.pred", "req.truth"),
   name = "Positive predictive value",
-  note = "Defined as: tp / (tp + number of fp). Also called precision. If the denominator is 0, PPV is set to be either 1 or 0 depending on whether the highest probability prediction is positive (1) or negative (0).",
+  note = "Defined as: tp / (tp + fp). Also called precision. If the denominator is 0, PPV is set to be either 1 or 0 depending on whether the highest probability prediction is positive (1) or negative (0).",
   fun = function(task, model, pred, feats, extra.args) {
     if (pred$predict.type == "prob") {
       prob = getPredictionProbabilities(pred)
@@ -1086,7 +1086,7 @@ measureEdgeCase = function(truth, positive, prob) {
 npv = makeMeasure(id = "npv", minimize = FALSE, best = 1, worst = 0,
   properties = c("classif", "req.pred", "req.truth"),
   name = "Negative predictive value",
-  note = "Defined as: (tn) / (tn + fn).",
+  note = "Defined as: tn / (tn + fn).",
   fun = function(task, model, pred, feats, extra.args) {
     measureNPV(pred$data$truth, pred$data$response, pred$task.desc$negative)
   }
@@ -1105,7 +1105,7 @@ measureNPV = function(truth, response, negative) {
 fdr = makeMeasure(id = "fdr", minimize = TRUE, best = 0, worst = 1,
   properties = c("classif", "req.pred", "req.truth"),
   name = "False discovery rate",
-  note = "Defined as: (fp) / (tn + fn).",
+  note = "Defined as: fp / (tp + fp).",
   fun = function(task, model, pred, feats, extra.args) {
     measureFDR(pred$data$truth, pred$data$response, pred$task.desc$positive)
   }
@@ -1124,7 +1124,7 @@ measureFDR = function(truth, response, positive) {
 mcc = makeMeasure(id = "mcc", minimize = FALSE,
   properties = c("classif", "req.pred", "req.truth"), best = 1, worst = -1,
   name = "Matthews correlation coefficient",
-  note = "Defined as sqrt((tp + fp) * (tp + fn) * (tn + fp) * (tn + fn)), denominator set to 1 if 0",
+  note = "Defined as (tp * tn - fp * fn) / sqrt((tp + fp) * (tp + fn) * (tn + fp) * (tn + fn)), denominator set to 1 if 0",
   fun = function(task, model, pred, feats, extra.args) {
     measureMCC(pred$data$truth, pred$data$response, pred$task.desc$negative, pred$task.desc$positive)
   }
diff --git a/tests/testthat/test_base_measures.R b/tests/testthat/test_base_measures.R
index b8441125a6..344cb99219 100644
--- a/tests/testthat/test_base_measures.R
+++ b/tests/testthat/test_base_measures.R
@@ -356,7 +356,7 @@ test_that("check measure calculations", {
   #test multiclass measures
 
   #mmce
-  mmce.test = mean(c(1L != 1L, 2L != 0L, 0L != 0L, 1L != 2L))
+  mmce.test = mean(c(1L != 1L, 2L != 1L, 0L != 0L, 1L != 2L))
   mmce.perf = performance(pred.classif, measures = mmce, model = mod.classif)
   expect_equal(mmce.test, mmce$fun(pred = pred.classif))
   expect_equal(mmce.test, as.numeric(mmce.perf))