One building's survey pool size was determined to be ~500 out of ~1500 total occupants. Assume survey forms are to delivered to every third person passing into the elevator vestibules (the only ingress). This is the only means of ingress to tenant areas and the point where the tenant flow into the building has become homogenous, regardless of their mode of transportation (note that stairways are locked in the direction of ingress causing all occupants to pass through the lobbies). Is this a sufficiently random sample? How does one account for those telecommuting or otherwise not traveling to work under this scenario?