Login

I’m trying to design a study on how neighborhood watch programs actually impact residents' perception of safety, not just crime statistics. My concern is that my survey questions might be inadvertently measuring general community satisfaction instead of the specific construct of perceived safety, which would threaten the study’s internal validity.

I did a small field test last year. People talked about safety but quickly tied it to how they felt about neighbors and pride in the block. It felt like the room you walk into is already deciding how safe it is.

In one draft I used items about fear of crime, comfort walking alone, and also general satisfaction with the neighborhood. The results blurred together.

We conducted cognitive interviewing with 6 residents, and several misunderstood perceived safety as police effectiveness or how loud the street is at night.

I think a practical move is to separate exposure to the watch program from outcomes, maybe with a natural experiment or a simple control, but it's messy in real streets.

We learned to drop long sentences and ask direct, concrete items, but still it's hard to avoid people answering through their mood that day.

Would you focus on a single neighborhood or compare several to see if perceived safety changes with watch program presence?

Login
Username:
Password:	Lost Password?
	Remember me