Testing assumptions when running deal numbers

[Scott Polly]

Hi everyone. I'm still in the learning stage of investing, gathering knowledge and working on building my "deal sense".  One basic question I have is this: How do I test my assumptions, particularly in terms of rent amounts? 

I imagine that one question can break down into a few more:

• When considering comps, I try to stay in the same relative area (or at least a similar area in the same town). In particular, I try to match schools and general neighborhood "vibe". Is there a less subjective way to do this?

• In terms of rental data, I can see rates for a lot of apartments in my area of interest, but not too many duplexes or single-family rentals. Is there a general rule of thumb for comparing apartments to houses? For example, a 2 BR house should rent for x more than a 2BR apartment?

• What's a good way to test the accuracy of tools like the BP rent estimator and others? Does anyone have a good source of long-term rental data? 

One of my favorite things about this site and others is that often responses suggest an entirely new way to think about the problem. Should I be thinking about all of this in a different way?

Thanks in advance for any advice.