Data Explorer
The most useful single node for a first look. It gives, per column: type, missing count, unique values, min/max/mean, and a distribution plot.
Run it straight after reading a file. Two minutes here saves an hour of confusion later — the same habit as df.info() and df.describe() in pandas.
Statistics
More detailed numeric statistics — quartiles, variance, skewness — plus a nominal-value table showing the distinct values of text columns and their counts.
That nominal table is how you spot TX, Tx and Texas all present in one field before they become three groups in a report.
Visual nodes
Bar Chart, Line Plot, Scatter Plot, Box Plot, Heatmap and others all produce interactive views. Selections made in one view can propagate to others when the nodes are combined in a component.
These are for checking your work rather than for final reporting — for that, output to a BI tool. But a scatter plot mid-workflow is often the fastest way to see that a join went wrong.
Checking as you go
After every node that can change the row count — Joiner, Row Filter, GroupBy, Concatenate — check the count.
A workflow that produces wrong numbers almost always does so because a count changed somewhere unnoticed, and finding it at the end means bisecting eighty nodes.
Reporting
KNIME can generate reports directly, and the Table View node renders a formatted table in a component's view. For anything an audience consumes regularly, writing to a database or extract and connecting a BI tool is usually the better arrangement.