Statistical help

  • Thread starter Thread starter maarkr
  • Start date Start date
M

maarkr

I have a group of hazards (say 25) with a percentage of
relative risk for each hazard. The risks generally range
from 10% to 70%. My statistics brain has gone alzheimers,
so what I want to do is come up with a cutoff percentage
that is significant, so I can focus on my analysis on the
greatest hazards.

Is there a function that can help me? or should I just
pick an arbitrary percentage (say 50%), or pick the top
20% of each group to base my reports on?
 
That's really something that will depend on your situation, and isn't
strictly an XL question. Are your relative risk data weighted for
severity of outcome?

Relative risk is normally measured relative to a known (or estimated)
probability (e.g., of mortality or morbidity) - and hazards usually
produce a relative risk >100% (that's one reason they're hazards). Are
your data weighted or scaled?

If they are scaled, addressing a 90% relative risk of getting a paper
cut while ignoring a 2% risk of a nuclear detonation is a decision that
XL is ill-equipped to handle via formula.
 
Back
Top