Thread: Data - rounding percentages.

Hello all, I'm just trying to figure this out.

If we have a sample of 135
133 yes
2 no

then my probability of "yes" is 133 over 135 or 98.5185%
"no" would be 100-98.5185 or 1.48148%

If I would try to apply this result to the next 100 samples
since I can't say half a yes or half a no....I would round up or down?
or 2 "no's" out of 100 and 98 "yes's"

I'm just rusty at this. Any help would be appreciated.

Thanks,
Carl

98.5 rounds to 99. So you would have 99 yes and 1 no.

This may not be a math question per say, but would the application have any bearing on which way you round. Imagine instead of yes and no samples it was number of bicycle crashes or something with a social context. Would that matter?

Originally Posted by CTolbert56

Hello all, I'm just trying to figure this out.

If we have a sample of 135
133 yes
2 no

then my probability of "yes" is 133 over 135 or 98.5185%
"no" would be 100-98.5185 or 1.48148%

If I would try to apply this result to the next 100 samples
since I can't say half a yes or half a no....I would round up or down?
or 2 "no's" out of 100 and 98 "yes's"

I'm just rusty at this. Any help would be appreciated.

Thanks,
Carl

Your sample size is too small to give an estimate any better than about 0-4 "no"s on a subsequent random sample from the same population.

CB