Mixed content is a very effective way of marking up textual data, the most obvious example being HTML.
However it should only be used when the data is appropriate (typically some kind of marked up textual data), as it complicates the parsing application, and without putting significant work into the parser its possible to create constructs that are not interpreted correctly or uniformly, again see HTML!
In simple terms an element may contain text and other elements
Sample
<description>character data <br/> more text <br/> and <strong>more data</strong></description>