Introduction to Segmented Univariate Analysis in Excel

In segmented univariate analysis, we segment the categorical variables and then perform univariate analysis across its categories. Let’s see how this simple technique can help you unveil powerful insights.

You have already done some segmented univariate analysis in the previous lectures, such as in the news popularity example, where you wanted to test the following three hypotheses:

1. On average, a higher number of articles are shared on weekdays than weekends.

2. Among the weekdays, articles published on Wednesdays get shared more than on any other weekday.

3. Articles of the type (or channel) 'lifestyle' and 'social media' are shared more than the other types on average.

In this case, the categorical variables used for grouping were ‘channel type’, ‘day of the week’, etc.  Across these categories or groups, you had then performed segmented univariate analysis to compare the average number of shares across days, channel types, etc.

In the next few lectures, we will take multiple examples and perform segmented univariate analysis in detail.