Hello all.

I am new here. Not really a math or a statistics buff, but I do have a question.

I work for a local utility agency and we are looking at US Census Bureau data. Right now we're looking at the number of mobile homes in a particular census tract and then comparing this to the percent of the same throughout the county. By way of example:

In one particular census tract there are 1,000 housing units and of these 100 are mobile homes, which means that 10% of the homes in that tract are mobile homes.

In the entire county there are 100,000 housing units and of these, 10,000 are mobile homes, which is (again) 10% of the units.

So far, no problem. But here's the rub. One of the people working on this data has labeled the county level units as a 'median' number (as in 'County Median 10%'). I keep arguing that this is not a 'median' but simply a percentage that reflects a 'share' - if you will - of the whole, so one in ten units (a.k.a 10%) - at both the tract level and the county-wide level - are mobile homes. And that a 'median' number doesn't apply here, nor does it fit or make sense.

I contend that we not looking at or defining a 'median' which would be: 'relating to a value or quantity lying at the midpoint of a frequency distribution of observed values or quantities.' There are no 'midpoints' here, only numbers that are being represented as a percent of a whole.

So, can someone confirm what is correct here (and if I, in fact have to eat a bit of humble pie!)

Thanks, D.

I am new here. Not really a math or a statistics buff, but I do have a question.

I work for a local utility agency and we are looking at US Census Bureau data. Right now we're looking at the number of mobile homes in a particular census tract and then comparing this to the percent of the same throughout the county. By way of example:

In one particular census tract there are 1,000 housing units and of these 100 are mobile homes, which means that 10% of the homes in that tract are mobile homes.

In the entire county there are 100,000 housing units and of these, 10,000 are mobile homes, which is (again) 10% of the units.

So far, no problem. But here's the rub. One of the people working on this data has labeled the county level units as a 'median' number (as in 'County Median 10%'). I keep arguing that this is not a 'median' but simply a percentage that reflects a 'share' - if you will - of the whole, so one in ten units (a.k.a 10%) - at both the tract level and the county-wide level - are mobile homes. And that a 'median' number doesn't apply here, nor does it fit or make sense.

I contend that we not looking at or defining a 'median' which would be: 'relating to a value or quantity lying at the midpoint of a frequency distribution of observed values or quantities.' There are no 'midpoints' here, only numbers that are being represented as a percent of a whole.

So, can someone confirm what is correct here (and if I, in fact have to eat a bit of humble pie!)

Thanks, D.

Last edited: