Split: Minor changes

Author: janezd

source/widgets/data/split.md

@@ -11,12 +11,12 @@ Split text or categorical variables into indicator variables.
 
 - Data: data table with added columns
 
-The widget splits a user-defined text or categorical variable into separate variables. The newly-created variables can be categorical (with Yes and No values), numeric (with 0 and 1 values), or counted (if a given value appears more than once). The widget is typically used with survey data.
+The widget splits the chosen text or categorical variable into separate variables. The newly-created variables can be categorical (Yes or No) or numeric (0 or 1) indicators, or appearance counts. The widget is typically used with survey data.
 
 ![](images/Split-stamped.png){width=40%}
 
-1. Select the variable to split on. Define the delimiter.
-2. Set the type of output variables. Categorical returns Yes/No values, numerical 0/1, while Counts returns counts (the same as Counter function in Python).
+1. Select the variable to split. Define the delimiter.
+2. Set the type of output variables. Categorical produces categorical indicator with values Yes and No, Numerical produces a dummy numerical variable with values 0 and 1, while Counts creates a variable that contains the number of appearances of each term. The latter two options are equivalent if each value can appear only once.
 3. If *Apply Automatically* is ticked, changes are communicated automatically. Alternatively, press *Apply*.
 
 Example
@@ -30,6 +30,6 @@ Say, we wish to observe and count the reasons users like Orange. We split on *Re
 
 ![](images/Split-Example.png)
 
-The advantage of Split is that now, we can count each response individually. Say, we observe how the responses differ on the position of the respondent in [Box Plot](../visualize/boxplot.md). It seem the professors especially appreciate Orange for its visual programming approach.
+The advantage of Split is that now, we can count each response individually. Say, we observe how the responses differ on the position of the respondent in [Box Plot](../visualize/boxplot.md). It seems the professors especially appreciate Orange for its visual programming approach.
 
 ![](images/Split-BoxPlot.png){width=500px}