I often use a three-step model for writing abstracts: 1) Define territory; 2) Define niche; 3) Occupy niche. First, define the theoretical field or subfield of interest. Second, define a (new or unstudied) problematique within this field or subfield. Third, tell the reader that you will study that problematique. In the third step I usually include A) methods and data used and B) my findings. So, the structure looks like
1)
2)
3)
3A)
3B)
I hope that makes sense.
Best,
Kristian