First, one must have a research question and aim. Secondly, how one defines mixed methods determines presentation. While we do not need to get hung up on labels, sequential/parallel and triangulation-embedded-explanatory-exploratory provide guidance on model construction. Third, what does integration look like? During data collection, analysis, and/or results? Harvard Catalyst recommends visuals for planning and the results. See https://catalyst.harvard.edu/community-engagement/mmr/
Michael Fetters "workbook" on mixed methods from Sage has several very useful examples of joint displays.
In general, I think integration is relatively straightforward for sequential designs, so joint displays are most important for so-called convergent (previously known as parallel) designs.