Setup: A radioactive source is placed at alternating distances from a Geiger Counter. Counts per second for 10 different distances (each spaced 1 cm apart) were taken for preset times ranging from 200 to 2000 seconds (depending on how far the source is from the counter).
Objective and Question: We are plotting counts per second vs source-to-counter distance for our sample, which emits beta radiation. We expect it to be $\frac{1}{r^2}$ dependance by the inverse square law, however we are getting less than $\frac{1}{r^2}$ dependance for larger distances (i.e) we are getting fewer counts than predicted by the inverse square law. At first we thought it was the air particles interacting with the beta particles, but that shouldn't have anything to do with it (according to the instructor), and the lab question hints it should have something to do with the geometry of the experiment. Can someone please help me understand this problem?