Thanks @xgaouva ! I made some adjustments. Couple of responses to your comments:
They should check primary keys before doing anything to the dataset. That is part of determining what the structure of their dataset is. If they don't understand exactly what is in their data and how its structure, they can't tidy it effectively.
Checking for duplicate rows is part of checking for a primary key (i.e. there is not primary key if there are duplicates).
Joining is not part of data tidying. Data tidying ends with separate dibbles for different types of data. At the end of tidying, they should be able to describe how different data sources CAN be joined, but the actual joining would be part of a specific analysis, not the data tidying process.
Check slide "Overview tidying steps"