A propeller accelerates air, and the stream tube going through the propeller disc can be idealized as something which is initially wider than the propeller disc and contracts while being accelerated (Froude's propeller hypothesis). Now two types of obstruction are possible:
- Obstructions to the flow near the propeller axis, and
- Obstructions to the flow near the stream tube.
It is quite obvious that something which blocks the flow will reduce the propeller's efficiency, and since flow speed is higher behind the propeller, the same obstruction will reduce performance more if placed behind the propeller disc. This is why pusher designs show better performance with the same engine power.
But also objects near the outer edge of the propeller can reduce efficiency. When a wall on one side ahead of the propeller prevents air from flowing towards the disc, the flow though the propeller disc is disturbed and efficiency will suffer. Similarly, a wall past the propeller will prevent a symmetric outflow, and the stream tube will attach to the wall (Coanda effect). This can be used to the advantage of high lift devices on wings, but will generally reduce efficiency.
For experimental data, I recommend to read NACA reports like the one for engine cowling development or this one on fuselage-propeller interference.