Blackbaud Data Warehouse databases use a modified star schema or constellation. The significant modification to the schema is that fact tables reference other fact tables. In a pure star schema, fact tables do not reference other fact tables.
According to Kimball, factless fact tables contain no measures. Factless fact tables facilitate many-to-many relationships. For example, there are many Constituencies with many Constituents and many Constituents with many Constituencies. Microsoft's Analysis Services model uses intermediate tables (Measure Groups or Facts) to accomplish this. So factless fact tables ensure that there is a record of every constituent and constituency that exists. For an explanation of many-to-many relationships with respect to extending the BBDW OLAP cube, see BBDW OLAP Extensions: Advanced Dimensional Usage.
Stars contained in a BBDW database include a revenue star and an interaction star, among others. However, there are other stars and remember, the database is extensible. Views added in versions above 2.91 abstract snowflake and constellation patterns to remove the snowflakes and constellations from the schema. So, there are snowflake and constellation patterns when considered from the table level. But when considered from views, the star schema holds true.