diff --git a/RELEASES.md b/RELEASES.md
index 5c7a70522ab..652154ab709 100644
--- a/RELEASES.md
+++ b/RELEASES.md
@@ -4,6 +4,8 @@
 
 - Added support for fetching tables from Google Spreadsheets.
   ([#1976](https://github.com/enso-org/enso/pull/1976)).
+- Added support for certain statistical functions of table columns
+  ([#1990](https://github.com/enso-org/enso/pull/1990)).
 
 # Enso 0.2.28 (2021-09-02)
 
diff --git a/distribution/lib/Standard/Table/0.2.29-SNAPSHOT/src/Data/Column.enso b/distribution/lib/Standard/Table/0.2.29-SNAPSHOT/src/Data/Column.enso
index dfb2088d503..8846b4fecfd 100644
--- a/distribution/lib/Standard/Table/0.2.29-SNAPSHOT/src/Data/Column.enso
+++ b/distribution/lib/Standard/Table/0.2.29-SNAPSHOT/src/Data/Column.enso
@@ -920,6 +920,41 @@ type Column
             (Vector.Vector v).reduce (+) / v.length
         this.java_column.aggregate 'mean' vec_mean True
 
+    ## Computes the variance of the sample represented by this column.
+
+       Arguments:
+       - degrees_of_freedom_correction: a correction to account for the
+         missing degrees of freedom in the sample. The default value of `1`
+         computes a sample variance. Setting it to `0` will compute population
+         variance instead.
+    variance degrees_of_freedom_correction=1 =
+        mean = this.mean
+        shifted = this - mean
+        sq = shifted * shifted
+        sq.sum / (this.length - degrees_of_freedom_correction)
+
+    ## Computes the standard deviation of the sample represented by this column.
+
+       Arguments:
+       - degrees_of_freedom_correction: a correction to account for the
+         missing degrees of freedom in the sample. The default value of `1`
+         computes a sample standard deviation. Setting it to `0` will compute
+         population standard deviation instead.
+    standard_deviation degrees_of_freedom_correction=1 =
+        this.variance degrees_of_freedom_correction . sqrt
+
+    ## Computes the coefficient of determination of a given prediction column.
+
+       Arguments:
+       - predictions: the column predicting the values of this column.
+    r_squared predictions =
+        prediction_diff = this - predictions
+        ss_res = prediction_diff*prediction_diff . sum
+        ss_tot_lin = this - this.mean
+        ss_tot = ss_tot_lin*ss_tot_lin . sum
+        1 - ss_res / ss_tot
+
+
     ## UNSTABLE
 
        Sorts the column according to the specified rules.
diff --git a/test/Table_Tests/src/Column_Spec.enso b/test/Table_Tests/src/Column_Spec.enso
index e679693231d..7ed3095a4db 100644
--- a/test/Table_Tests/src/Column_Spec.enso
+++ b/test/Table_Tests/src/Column_Spec.enso
@@ -64,3 +64,23 @@ spec = Test.group "Columns" <|
         col = Examples.decimal_column.set_index Examples.integer_column
         col.index.to_vector . should_equal Examples.integer_column.to_vector
 
+    Test.specify "should allow computing variance and standard deviation" <|
+        const = Column.from_vector 'const' [1, 1, 1, 1, 1]
+        const.variance . should_equal 0
+        const.standard_deviation.should_equal 0
+
+        rand = Column.from_vector 'random' [10.0, 4.2, 6.8, 6.2, 7.2]
+        rand.variance . should_equal 4.372
+        rand.variance degrees_of_freedom_correction=0 . should_equal 3.4976
+        rand.standard_deviation . should_equal 2.090932806 epsilon=(10 ^ -6)
+        rand.standard_deviation degrees_of_freedom_correction=0 . should_equal 1.870187156 epsilon=(10 ^ -6)
+
+    Test.specify "should allow computing the R² score of a prediction" <|
+        sample = Column.from_vector 'sample' [1,2,3,4,5]
+        mean_pred = Column.from_vector 'mean' [3,3,3,3,3]
+        perfect_pred = Column.from_vector 'perfect' [1,2,3,4,5]
+        bad_pred = Column.from_vector 'bad' [5,4,3,2,1]
+
+        sample.r_squared mean_pred . should_equal 0
+        sample.r_squared perfect_pred . should_equal 1
+        sample.r_squared bad_pred . should_equal -3