Data modeling in Agile Projects
Data Models help business user and technical member agree on requirement. Model specify the data element that will be used and created by business process. Data model explictly determine structure of data.
There are 3 types of data model:
1. Conceptual Data Model
2. Logical Data Model (LDM)
3. Physical Data Model (PDM)
In Agile project - Business people and development team work together. Agile process value -
1. Individual and interaction over tools and process
2. Working software over documentation
If you are working on a large project (50 member team for 1 year), it is important that Data model is created at conceptual, logical, physical level. You do not want everyone in team are creating their own tables.
Generally large project with bigger team use waterfall method and you build data model in "Envision and Plan" phase.
If you are working on agile project which has 2 developer, 1 tester and 1 business member - You can create data model if it help to agree requirement. However most likely you do not need it, you can directly draw physical schema (physical data model).
The two principles given above - "Individual and interaction over tools" and "Working software over documentation" justify that make model that help otherwise directly focus on software.